Warning devices

Investigation summary

What happened

On the afternoon of 10 September 2025, a Tasman Cargo Airlines Boeing 767-300 (B767), registered VH-XQU and operating as Tasman 22, was approaching Sydney Airport, New South Wales, at the conclusion of an air transport freight‑only flight from Hong Kong International Airport, Hong Kong. While intercepting the glideslope for the instrument landing system approach for runway 16R, the aircraft’s autopilot pitched the nose down. Subsequently, the approach was conducted with a high rate of descent while descending away from the glideslope.

The descent triggered an air traffic control minimum safe altitude warning (MSAW) before the autopilot was disconnected and the aircraft levelled out at 1,700 ft, approximately 1,000 ft below the glideslope. It then descended a further 150 ft before a missed approach was commenced. On the subsequent approach, the crew elected to conduct a localiser approach and the aircraft landed without further incident.

What the ATSB found

The ATSB found that 2 Airbus A380s on the ground at Sydney Airport taxied through the instrument landing system critical area and in front of the glideslope antenna, causing interference to the glideslope signal. As a result, after detecting the interference, the Boeing 767's autopilot established the aircraft on a flight path that deviated away from the glideslope, before alerting the crew that it was operating in a degraded mode.

It was also determined that the pilot flying continued the approach with the autopilot in a degraded mode. As a result, the aircraft’s high descent rate triggered an air traffic control minimum safe altitude warning. After disconnecting the autopilot, the pilot flying delayed the initiation of a missed approach and the aircraft descended below the localiser segment minimum safe altitude. In addition, the pilot monitoring did not effectively monitor the aircraft's flight path during the approach and did not call out deviations or advise the pilot flying to conduct a missed approach.

Furthermore, it was found that Tasman Cargo Airlines allowed the practice of flight crew exchanging flying and monitoring roles prior to 1,500 ft when conducting a practice autoland. It was also identified that Tasman Cargo's training did not inform flight crew of the conditions under which instrument landing system critical areas were protected. Consequently, the flight crew believed that the critical area was being protected, and the risk of glideslope interference had been mitigated.

What has been done as a result

Tasman Cargo Airlines has conducted the following proactive safety action:

• Removed the ‘glideslope out’ procedure from its policy and procedures manual (PPM), replacing it with the requirement to conduct a missed approach and notify air traffic control.
• Added a note in the PPM referencing the flight crew operating manual glideslope interference bulletin.
• Published an operational alert regarding Sydney Airport runway 16R glideslope interference with guidance information.
• Updated computer-based training materials to better highlight the ILS critical area and requirements during normal and low visibility operations.
• Introduced a requirement for the pilot flying to take control of the aircraft prior to the commencement of the approach when conducting an autoland.

In addition, while not in direct response to this occurrence, Boeing advised it is in the process of updating the flight control software for B767 aircraft. Planned for release in 2027, the update will include the following changes:

• The flight director pitch bar will remain biased out of view if the autopilot is disconnected while in attitude stabilising mode.
• Improved glideslope capture logic to reduce the occurrence of false captures leading to attitude stabilising mode.
• Limiting the flight path angle while in attitude stabilising mode to be between 0‍–3.25° of descent.
• Display of NO AUTOLAND after being in attitude stabilising mode for 15 seconds if the aircraft is still above alert height.
• Automatic autopilot disconnect 10 seconds after the display of NO AUTOLAND if the aircraft is still above 500 ft AGL.

Safety message

Flight crew of Boeing 747-400, 747-8, 757, 767, 777 and 787 aircraft should familiarise themselves with their aircraft’s flight crew operating manual bulletin for ILS signal interference and be prepared to conduct prescribed procedures when encountering cockpit indications and autoflight behaviour consistent with the issue.

Automation can reduce pilot workload and enhance flight safety. However, it is critical that flight crew maintain constant awareness of the performance of the autoflight system. When under the control of automation, flight crew should monitor the flight path and verify that aircraft behaviour is consistent with the automation modes and parameters selected. If a discrepancy is identified, automation should be disconnected or the level of automation reduced until control is re-established.

The occurrence

On the afternoon of 10 September 2025, a Tasman Cargo Airlines Boeing 767-3JHF (B767), registered VH-XQU and operating as Tasman 22, was approaching Sydney Airport, New South Wales, at the conclusion of an air transport freight-only flight from Hong Kong International Airport, Hong Kong. On board were 3 flight crew, consisting of a captain, who occupied the left seat and was the pilot monitoring,¹ a first officer who occupied the right seat and was the pilot flying, and a relief pilot who occupied a jump seat.

The crew planned to conduct a practice autoland approach (see the section titled Practice autoland) to perform a scheduled check of the aircraft’s systems. As company procedures required that the captain was the pilot flying for an autoland approach, the crew had briefed that the captain and first officer would exchange flying and monitoring roles at 1,500 ft above ground level (AGL). The first officer also recalled that the approach briefing conducted by the crew included items specific to an autoland (see the section titled Instrument approach briefing).

At 1454, when the aircraft was at its cruising level of flight level 350,² the crew advised air traffic control (ATC) of their intention to conduct a practice autoland at Sydney, which the controller acknowledged. Shortly afterwards, the aircraft commenced descent. At 1510, the crew were transferred to the approach controller, who issued a clearance to continue descent. The crew read back the clearance and confirmed that the controller was aware they were conducting a practice autoland.

The approach controller subsequently advised the tower controller at Sydney Airport that the aircraft would be conducting a practice autoland, before transferring them to a second approach controller. At 1516, the aircraft was cleared to descend to an altitude of 3,000 ft and conduct the instrument landing system (ILS) approach for runway 16R.³ Two minutes later, at an altitude of 5,300 ft, the aircraft intercepted the localiser course for the approach. At the time, the autopilot was controlling the aircraft’s flight path and the autothrottle was engaged in speed mode (see the section titled Autoflight system).

Concurrently, on the ground at Sydney Airport, an Airbus A380 was awaiting departure at holding point A1, adjacent to runway 16R (see the section titled Glideslope critical area). A second A380 was also waiting for departure in sequence on taxiway A. At 1519, the first A380 was cleared to line up on runway 16R and began moving onto the runway (Figure 1). The second A380 followed in sequence to hold at A1.

Figure 1: Ground movement of A380s

At 1519:18, when slightly below and flying to intercept the glideslope, the crew of VH‑XQU selected approach mode on the mode control panel and the autoflight system indicated that it had captured the glideslope (Figure 2). Two seconds later, the aircraft pitched down, adopting a nose down pitch of approximately 2°. As a result, the aircraft’s descent rate increased, and it began to diverge below the glideslope. At 1519:35, 15 seconds after the commencement of the pitch down and when the aircraft was descending through 3,730 ft, the crew were presented with several alerts in the cockpit:

• a line through the pitch (G/S) flight mode annunciation (FMA)
• removal of the flight director pitch bar from the primary flight display
• an autopilot caution message.

Figure 2: Flight path relative to runway 16R glideslope

The first officer recalled seeing the autopilot caution and associated alerts. They also reported that they recognised that the aircraft’s descent rate had increased and that it was no longer following the glideslope. The captain also recalled identifying that the aircraft increased its descent rate, however they believed that this occurred later in the approach and that initially the aircraft followed the glideslope. 

The aircraft continued to descend away from the glideslope, maintaining an average descent rate of 1,650 feet per minute (ft/min) for approximately another minute. During this time, the autopilot flight director system (AFDS) remained in approach mode with 3 autopilots engaged, and the alerts continued to be displayed to the crew. Both flight crew reported that, after being alerted, they discussed changing to conduct a localiser approach (see the section titled Instrument approach and landing chart). The first officer further recalled that, in addition, they began briefing and preparing for the change of approach. Both crew members advised that they considered they had time to have this discussion as the aircraft was still at a relatively high altitude and they had not yet passed the final approach point. The aircraft’s airspeed remained above the autothrottle selected airspeed, however the flight crew did not recall being required to manage excess airspeed during the approach.

The relief pilot, seated behind the captain and first officer, recalled that once the autopilot caution was observed, they assisted by calling out distances and required altitudes from the instrument approach chart (see the section titled Instrument approach and landing chart). The relief pilot further recalled that while doing this, they advised the other crew members that the aircraft was below the glideslope and called for them to slow the rate of descent. 

At 1520:33, ATC radar recorded the aircraft at an altitude of 2,200 ft with a descent rate of 1,731 ft/min approximately 10 NM (19 km) from the airport. This descent rate and low altitude triggered an ATC minimum safe altitude warning (MSAW),⁴ which was presented on the approach controller’s display (Figure 3). The approach controller had just instructed the flight crew to change radio frequency and contact the tower controller, therefore they asked the tower controller to issue a safety alert to the crew. Given the frequency change instruction had just been issued, the approach controller also attempted to contact the flight crew directly on the approach frequency. The crew responded they were still on that frequency. The approach controller gave the crew a low altitude alert, advising them to check their altitude and the crew responded that they were going around. 

Figure 3: Approach controller display at 1520:39

Just prior to the approach controller advising the crew of the low altitude warning, as the aircraft descended through 1,850 ft, the first officer disconnected the autopilot, re‑engaged the AFDS in vertical speed mode and raised the nose of the aircraft. As a result, the descent stopped at 1,700 ft, 8.6 NM (16 km) from the runway and approximately 1,000 ft below the glideslope. The aircraft then flew level for about 10 seconds before it began to descend again. At 1521:08, a go‑around was initiated from an altitude of 1,550 ft, as the low altitude alert was being given by ATC.

The aircraft climbed to 3,000 ft and was re-sequenced by ATC for another approach. This time, the crew elected to conduct a localiser approach, which they briefed while repositioning the aircraft to rejoin the approach. During this subsequent approach, the crew observed that the glideslope appeared to be functioning normally, and the aircraft landed without further incident.

After landing, the flight crew discussed the incident between themselves, and with the incoming crew for the next flight. They recalled that the B767 flight crew operations manual (FCOM) contained a bulletin that described the AFDS behaviour that was experienced (see the section titled Manufacturer signal interference bulletin) and concluded that glideslope interference had likely occurred.

Both the first officer and the relief pilot advised that they believed the glideslope critical area would have been protected because they were conducting an autoland; the captain also advised that they thought the glideslope critical area would have been protected, but due to the weather conditions at the time. The captain further reported that, after landing, they called ATC to enquire if there were any vehicles or aircraft that may have interfered with the glideslope signal and was advised that there were not. The aircraft’s ILS system was tested by maintenance personnel after landing and was assessed as serviceable.

Context

Pilot information

The captain held an Air Transport Pilot Licence (Aeroplane) and class 1 aviation medical certificate. They had 15,296 hours of flying experience, of which 2,567 were on the Boeing 767 (B767) aircraft type, with 102 hours accrued in the previous 90 days.

The first officer held a Commercial Pilot Licence (Aeroplane) and class 1 aviation medical certificate. They had 7,362 hours of flying experience, of which 440 were on the B767 aircraft type, with 84 hours accrued in the previous 90 days.

The relief pilot held an Air Transport Pilot Licence (Aeroplane) and class 1 aviation medical certificate. They had 20,599 hours of flying experience, of which 102 hours were on the B767 aircraft type, all accrued in the previous 90 days.

The flight crew had 2 days free of work prior to the flight and all pilots reported that they had slept adequately prior to, and had rested during, the flight.

Aircraft information

General information

VH-XQU was a Boeing 767-3JHF, manufactured in 2009 and first registered in Australia with the operator in 2022. The aircraft was fitted with 2 General Electric CF6‑80C2B7F turbofan engines and had accumulated 56,678 flight hours.

Autoflight system

Overview

The aircraft’s automatic-flight system consists of an autopilot flight director system (AFDS), and an autothrottle system. The AFDS provides pitch (vertical) and roll (lateral) guidance via flight director command bars overlaid on the attitude direction indicator (ADI) display (Figure 4).

Figure 4: Attitude director indicator display

Flight crew controlled the mode and associated flight parameters of the AFDS via the mode control panel. Available modes included:

• Approach mode (APP): The flight director provided guidance to track the localiser laterally, and the glideslope vertically. This mode was used during an instrument landing system (ILS) approach.
• Vertical speed mode (V/S): The flight director pitch bar indicated the aircraft pitch required to maintain the selected vertical speed. This speed was adjusted by flight crew using the vertical speed selector.

The aircraft was equipped with 3 autopilots. When engaged, each autopilot activated flight controls required for the aircraft to follow the flight director command bars. Only one autopilot was required to be engaged to provide autoflight capability, however multiple autopilots could be engaged to provide redundancy. This was required when conducting certain operations such as an autoland. When no autopilot was engaged the pilot flying was required to manually control the aircraft to follow the flight director.

Automatic control of engine thrust was provided by the aircraft’s autothrottle system. This system was also controlled via the mode control panel. When speed mode (SPD) was active, the autothrottle would control engine thrust, between idle and maximum thrust, to achieve the selected airspeed.

The flight mode annunciations (FMA) displayed (green) active and (white) armed modes of the AFDS and the (green) active mode of the autothrottle system to the flight crew. 

The flight crew operations manual (FCOM) contained information about engaging and disengaging the autoflight system and advised that:

After localizer and glideslope capture, the localizer and glideslope modes can only be deactivated by disengaging the autopilot and turning both flight directors off or by selecting GA [go-around] mode.

The FCOM also advised that:

If unwanted operation is noticed or when an autopilot failure is annunciated, the autopilot should be disconnected and the airplane flown manually.

Meteorology

The automatic terminal information service (ATIS)⁵ at Sydney Airport reported weather conditions at the time of the occurrence that included:

• cloud layers of few at 500 ft, scattered at 1,000 ft and broken at 3,000 ft⁶
• wind from 140° M at 15 kt
• visibility of 5,000 m in rain
• temperature of 16°C 
• QNH⁷ of 1004
• advice that turbulence was expected in the circuit area
• an aerodrome warning that thunderstorms were expected from the west until 1845.

The flight crew advised that weather conditions were consistent with the ATIS, and that the aircraft was in instrument meteorological conditions⁸ throughout the approach and subsequent missed approach.

Recorded data

The ATSB was provided with data from the aircraft’s quick access recorder, which captured the incident flight. The data was also sent to the manufacturer for review (see the section titled Manufacturer review of recorded data). The cockpit voice recorder was not available due to the time that had elapsed since the occurrence. 

A review of the data identified that at 1519:00, the aircraft, with 3 autopilots engaged, was at an altitude of 4,000 ft with the localiser captured (LOC) and descending slightly while intercepting the glideslope from below (Figure 5). At 1519:18, the AFDS changed to approach mode and the associated FMA indicated that the glideslope mode (G/S) was active.

Two seconds later, the aircraft pitch lowered from 4° nose up to approximately 2° nose down, and away from the flight director pitch command bar. The glideslope deviation then recorded a period of variations inconsistent with the aircraft’s position relative to the glideslope. The aircraft’s descent rate subsequently increased, and it began descending away from the glideslope. 

Figure 5: Graphical representation of recorded flight modes and alerts

The data recorded the activation of a pitch fault FMA, an autopilot caution, and a master caution light at 1519:35. Approximately 15 seconds later, the flight director pitch command bar began showing values alternating between 0° and 45°, which the manufacturer advised was consistent with the flight director pitch bar being removed from the primary flight display (biased out of view).

Over the next minute, the aircraft maintained a descent rate of approximately 1,650 ft/min. During this time, the 3 autopilots remained engaged. Also during this time, the aircraft’s airspeed remained above the selected airspeed (Figure 6) while the engines remained at idle thrust. At 1520:05 the aircraft passed the initial approach fix at an altitude of approximately 2,900 ft, with the glideslope deviation at 2 dots, or full-scale deflection. From 1520:08, the speedbrake was deployed in varying positions, before being armed for landing at 1520:29.

Figure 6: Graphical representation of recorded airspeed and configuration

At 1520:42, the 3 autopilots were disconnected and the FMA fault indication and autopilot caution ceased. The data also recorded a change in AFDS mode from approach mode to vertical speed mode at this time, initially with a descent rate of 1,600 ft/min selected, consistent with the aircraft’s vertical speed at the time.⁹ At 1520:50, the aircraft’s altitude stabilised at 1,700 ft as the aircraft’s pitch raised to 5° nose up. The vertical speed selected was then reduced to −800 ft/min, then to 0.

At 1520:59, the flight director pitch command bar progressively raised over 7 seconds. The aircraft pitch also increased over this time. Also during this period, the airspeed decreased to the selected airspeed of 183 kt and engine thrust increased above idle. At 1521:08 go‑around mode was selected and the aircraft commenced a climb to 3,000 ft.

Instrument landing system

Overview

An instrument landing system (ILS) is an instrument approach that uses lateral (localiser) and vertical (glideslope) position information, using angular deviation signals from the localiser antennas (located past the upwind end of the runway) and the glideslope antennas (located to the side of the runway). Aircraft systems detect these radio signals and provide instrument indications that enable an aircraft to be manoeuvred along a precise final approach path.

Instrument approach and landing chart

The flight crew used a Jeppesen instrument approach and landing (IAL) chart to brief and conduct the approach (Figure 7). The chart contained information required to conduct either an ILS approach, with the glideslope providing vertical guidance, or a localiser approach that provided distance and altitude information for the flight crew to manage the vertical path, when vertical guidance from the glideslope was not available.

Figure 7: Instrument approach and landing chart runway 16R ILS/LOC

The chart specified the 25 NM (46 km) minimum sector altitude as 2,700 ft. This was the minimum altitude an aircraft could descend to while maintaining terrain and obstacle clearance when between 10–25 NM (19–46 km) of the airport and not in visual conditions or conducting an instrument approach. The minimum sector altitude reduced to 2,100 ft when within 10 NM (19 km) of the airport.

The chart also included a caution message that advised of the risk of glideslope interference (see the section titled Glideslope interference) stating that:

GP [glidepath] false indications due acft [aircraft] near Twy [taxiway] A1.

Information on the chart specific to conducting the localiser approach included a segment minimum safe altitude associated with each segment of the approach. Flight crew were required to ensure the aircraft remained at or above these altitudes when within the corresponding segment to maintain minimum obstacle and terrain clearance. The chart also included table of distance/altitude relationships to assist in maintaining a 3° descent profile when glideslope guidance was not available.

Missed approach requirements

The Aeronautical Information Package (AIP) stated that when conducting an ILS approach, a missed approach was required to be executed under certain conditions, including:

During an instrument approach and below MSA (as specified on the IAL chart) the performance of the radio aid becomes suspect, or the radio aid fails.

Glideslope interference

Glideslope critical area

Disturbances to the ILS glideslope signal may occur when vehicles or aircraft are operated near the antenna. For this reason, an ILS critical area is defined around the glideslope antenna location, within which aircraft and vehicle movement is restricted under certain conditions.

The glideslope critical area for runway 16R at Sydney Airport was an area 600 m long and 145 m wide and included the A1 holding point (Figure 8). When protection of this critical area was required, air traffic control (ATC) directed aircraft to hold at an alternate holding point on taxiway A, instead of A1.

Figure 8: Glideslope critical area for runway 16R

The conditions under which the ILS critical areas were required to be protected were defined in the Manual of Air Traffic Standards (MATS), which stated:

When the ceiling¹⁰ is at or below 600 ft or the visibility is 2000 m or less, ensure no aircraft enter the glide path or localiser critical areas when an arriving aircraft is within:
a) the outer marker; or
b) 4 NM from the threshold if no outer marker exists.

Manufacturer signal interference bulletin

The FCOM contained a bulletin, issued in November 2021 and titled Erroneous Autopilot Flight Director System (AFDS) Guidance when Instrument Landing System (ILS) Signal Interference Occurs. The bulletin stated that:

Boeing has received several reports of unexpected pitch guidance when capturing or tracking the glideslope during an instrument landing system (ILS) approach. In each event for which data was provided, Boeing has determined that glideslope signal interference occurred at the time of the unexpected pitch guidance and, in most of these events, the unexpected pitch guidance occurred during glideslope capture. ILS signal interference can occur when vehicles, aircraft, or other factors affect the localizer or glideslope signal.

The bulletin further stated that:

The AFDS can detect the degradation or instability of radio signals that support specific autopilot modes. When the AFDS detects a degraded or unstable signal during an ILS approach with the autopilot engaged, the affected AFDS mode changes to an attitude stabilizing mode based on inertial data at the time of the signal degradation or instability. The purpose of the attitude stabilizing mode is to prevent large and abrupt pitch and roll changes during short periods of localizer or glideslope signal interference. When the localizer or glideslope signal stabilizes and the airplane is within parameters for capture, the AFDS returns to tracking the localizer or glideslope. Alternatively, if the localizer or glideslope signal does not stabilize or the airplane is not within parameters for capture, the attitude stabilizing mode remains active. In this case, the AFDS continues to provide guidance in the attitude stabilizing mode, with possible high rates of descent and significant deviation from the localizer or glideslope.

The bulletin also advised that:

There is no direct indication to the pilot that the attitude stabilizing mode is active if the airplane is above 200 feet radio altitude and either:

• the localizer attitude stabilizing mode is active for less than 20 seconds or

• the glideslope attitude stabilizing mode is active for less than 15 seconds

If the airplane is above 200 feet radio altitude and the attitude stabilizing mode remains active for 20 seconds or more (for localizer) or 15 seconds or more (for glideslope):

• the AUTOPILOT message shows (if autopilot is engaged) (Figure 9) and

• the flight director roll or pitch bar is removed (if flight director is on) and

• an amber line shows through the affected flight mode annunciation (FMA)

Figure 9: Indications following extended time in attitude stabilising mode

Operating instructions were given for flight crew conducting an ILS approach, which advised that:

While on an ILS approach, monitor localizer and glideslope raw data and call out any significant deviations. Perform an immediate go-around if not within the criteria to continue the approach.

It is essential to crosscheck altitude at the FAF [final approach fix] and monitor pitch attitude and descent rate throughout the approach.

If a glideslope anomaly is suspected, an abnormal altitude range-distance relationship may exist. This can be identified by crosschecking distance to the runway with altitude or crosschecking the airplane position with waypoints indicated on the navigation display. The altitude should be approximately 300 feet height above touchdown per NM of distance to the runway for a 3° glideslope.

Manufacturer review of recorded data

At the request of the ATSB, the manufacturer reviewed the recorded data from the occurrence flight. The review included analysis of glideslope beam variation at the time of glideslope capture and throughout the approach. It identified that:

At a longitudinal distance of approximately 15.5 nautical miles [29 km] prior to the runway threshold, the observed glideslope beam began to diverge from the ideal beam, moving below the ideal beam. This deviation between the ideal and observed beam is characteristic of a beam anomaly.

The observed glideslope beam reached approximately 300 feet below the ideal beam. The divergence then began to decrease as the observed beam began to converge towards the ideal beam.

The observed glideslope beam diverged from the ideal beam again, varying by up to 350 feet below and then 400 feet above the ideal beam over the next 1.5 nautical miles.

A second period of divergence between the ideal and observed glideslope commenced at approximately the same time as the second A380 entered the critical area.

Based on this analysis, the manufacturer concluded that:

The airplane experienced a beam anomaly while attempting to capture the glideslope from below the beam. The observed glideslope beam diverged below the ideal 3-degree beam by a vertical offset of 300 feet, causing the airplane to attempt glideslope capture prematurely.

The review also referred to the FCOM bulletin for glideslope interference and advised that:

In this event, the airplane behaviour was consistent with a glideslope beam anomaly, causing the AFDS to enter attitude stabilizing mode. The airplane encountered a beam anomaly while intercepting the glideslope in G/S mode before entering into a steady descent with an average calculated vertical speed of approximately -1,700 feet per minute. The AFDS did not return to tracking the glideslope after the glideslope signal stabilized, probably because the glideslope deviation was no longer within the parameters for capture. For approximately the first 15 seconds, there were no indications in the QAR [quick access recorder] data that attitude stabilizing mode was active until the Autopilot Caution and FMA Pitch Fault activated. An additional 15 seconds passed before the flight director pitch guidance became BOV [biased out of view], consistent with the removal of the flight director pitch bar. This state remained until the flight crew switched the pitch mode to V/S.

Noting the operational guidance in the bulletin, the review concluded that:

Earlier recognition of the airplane state and degraded autopilot performance may have reduced the magnitude of the deviation below the glideslope.

The manufacturer also advised that there was no evidence to show whether the glideslope deviation pointer would have been displayed during the period that the alerts were active.

Practice autoland

Overview

An autoland approach is an approach during which the aircraft is fully controlled by the autoflight system through to landing. During an autoland, pilots assume a monitoring role and intervene only in the event of a system failure or other abnormal event. The procedure is primarily intended to be used in low-visibility conditions, however an autoland can also be conducted in better conditions to meet aircraft or flight crew recency and currency requirements. An autoland conducted under these conditions is termed a practice autoland.

Regulatory requirements and guidance

Flight crew were required to advise ATC of an intention to conduct a practice autoland when arriving at an aerodrome and the AIP stated that:

In weather conditions where the ceiling and/or visibility are above CAT I minima, pilots should inform ATC about any intention to conduct:

a. an approach with minima less than standard CAT I; or

b. an autoland procedure.

This information must not be taken as a request for or expectation of the protection of the ILS but to enable ATC to inform the flight crew of any known or anticipated disturbance.

The AIP also stated that when receiving this advice, and the ILS critical area was not protected, the controller was required to report ‘ILS critical area not protected’. Airservices Australia advised that only a tower controller provided advice of any known or anticipated disturbances. They also confirmed that an intention to conduct a practice autoland procedure did not change ATC’s requirements with regards to protecting the ILS critical areas.

The Civil Aviation Safety Authority published Advisory Circular (AC) 91‑12 Conduct of practice autoland operations. The AC described the problems and potential risks when conducting an autoland and stated:

Multiple factors may influence the accuracy of the ILS signal when the ILS autoland system is to be used:

• in conditions where ATC is not protecting ILS critical and/or sensitive areas.

These factors include:

• interference of the ILS signal due to an intrusion within the ILS critical and sensitive areas by:
  - taxiing aircraft
  - ground vehicles
  - over-flight of the ILS localiser.

The AC also identified a number of considerations, procedures and instructions when conducting practice autoland operations, which included:

ATC should be informed about the crew’s intention to conduct an autoland. Pilots should not expect the protection of the ILS, but on receiving advice from the crew of their intention to conduct a practice autoland, ATC may inform the flight crew of any known or anticipated disturbance.

Operator procedures and training

Flight crew roles and responsibilities

The operator’s policy and procedures manual (PPM) defined the roles of the pilot flying and the pilot monitoring as:

The PF [pilot flying] will control and monitor the aircraft, regardless of the level of automation being used. The duties of the PF are:

• flight control:
  - flight path and airspeed control
  - navigation
  - aircraft configuration

• conducting normal procedures in compliance with company manuals

• monitoring flight status and condition.

The PM [pilot monitoring] will monitor the aircraft and action of the PF. The duties of the PM are:

• general monitoring and crosschecking the procedures for all flight phases including take-off, cruise, approach and landing, aircraft system status, and condition

• make callouts to PF for any deviation from the flight path or system malfunction or failure

• give advice to the PF to execute a missed approach when the approach becomes unstable or unable to continue the approach safely

• ATC communications, checklist reading, FMS [flight management system] CDU [control display unit] operation

• complete the PF directions.

A callout must be made by the PM for the other crew if a deviation occurs from the normal procedures, or intended flight path (attitude, speed, altitude, and direction). If there is a failure for the PF to respond to callouts of the PM, appropriate corrective action shall be made for the safety of the flight including taking over the aircraft control.

The PPM defined tolerances for certain flight parameters, including a tolerance of one dot when tracking the glideslope, and advised that:

These parameters are to be observed and sustained flight should not be allowed to continue without corrective action by the flight crew. Any deviation outside of approach tolerances in IMC must result in a GA [go-around].

Use of automation

The PPM contained an automation policy which advised:

Automation is provided to enhance safety, reduce pilot workload and improve operational capabilities.

Automation can be as beneficial as mentioned above for flight operations, if automation is used appropriately. However, at the same time, if flight crew members fail to maintain proficiency in the use of automation or to properly correspond to recognised degradation of automation performance, it could be an obstruction for the flight safety.

In addition, over-reliance on automation system may lead to flight crews accepting whatever the aircraft was doing without proper monitoring or may result in deterioration of flight crew members’ manual flying skills. Therefore, for safe flight operation, proper use of automation system, maintaining manual flying skills, and practical use of CRM [crew resource management] skills are required.

Following elements are imperative to use automation at the most appropriate level:

• full understanding and knowledge of automation system

• proficiency in the use of all levels of automation

• monitoring and cross-check of automation

• skills to recognize degraded performance of automation system

• skills to shift between all levels of automation including manual flying.

The PPM also contained guidelines for flight crew when using automation, which included that:

• The flight crew must compare the performance of autoflight system (operation status) with the flight path of the aircraft.

• It is the responsibility of the flight crew to maintain flight mode awareness and control to ensure the flight director provides the guidance required.

• If any autoflight system is not operating as expected or any doubt exists, disengage it.

Instrument approach briefing

The PPM required flight crew to conduct a briefing of the threats, plans and considerations of the expected approach after the approach had been loaded. Considerations to be included in the brief were:

• approach title and chart effective date

• manual or AUTOLAND Landing

• nomination of required navigation aids

• approach entry/holding pattern

i• nitial approach altitude

• approach track(s)

• limiting and check altitudes

• CAT I, CAT II, CAT III MDA/DH [minimum descent altitude / decision height] and visibility minimum

• aerodrome elevation

• circling procedures (if applicable)

• significant terrain

• missed approach procedure

• monitoring of approach aids.

In addition, the crew were required to brief specific items prior to commencing an autoland approach which included the following non-normal situations:

• persistent localiser or glideslope deviation alert between 500 ft RA [radio altimeter] & 200 ft RA or any deviation alert below 200 ft RA

• ASA [autoland status annunciator] changes to NO AUTOLAND

• ILS Localiser and/or Glideslope Transmitter Failure

• autopilot disconnect.

Glideslope out procedure

The PPM contained procedures for when the glideslope failed during an ILS approach. The procedures stated that when in IMC:

If the glideslope fails when the aircraft is established on the ILS procedure (or being radar vectored) after the IAF [initial approach fix] ATC must be notified immediately.

An approach may be continued if the following items can be completed prior to FAF [final approach fix]:

• complete briefing for a LOC [localiser] (GS [glideslope] out) approach

• mode change and operation for LOC approach

• set the next fix altitude or MDA on altitude window

• barometric altimeter bug set

• completion of landing configuration and landing checklist

A missed approach must be conducted for the following situations

• failure to complete the steps above, prior to the FAF

• failure of situational awareness for the LOC (GS out) procedure

• when directed by ATC.

Low visibility operations

Autoland procedures

The PPM defined procedures for taxi, take-off and landing in conditions where visual reference was limited by weather. This included procedures for conducting a low visibility operations (LVO) approach and landing including an autoland. It was stated that:

LVO approach and landings shall be carried out using the highest level of automation available to an autoland.

The left seat pilot is PF and the right seat pilot is PM for the approach, landing or missed approach. 

In addition, crew responsibilities during an LVO landing were stated as:

Pilot Flying Duties

The PF shall guard the control column, rudder pedals and thrust levers throughout the approach, landing, or missed approach, make standard callouts and responses as per the LVO landing standard callouts and actions, and take manual control in the event of autopilot disconnect.

Pilot Monitoring Duties

The PM must monitor the aircraft flight path and flight instruments throughout the approach, landing and rollout, or missed approach. They are to remain ‘heads down’ and make no attempt to seek visual reference. Any deviation or system failure must be called to alert the PF.

Additionally, specific standard callouts (by the PM) and responses (by the PF) were required to be used when conducting an LVO landing. These specific callouts began at 1,500 ft AGL.

The PPM did not contain procedures specific to conducting a practice autoland. The operator later advised that when conducting LVO operations, the left seat pilot was required to be the pilot flying for the entire approach however, when conducting a practice autoland, the pilots could switch roles so long as the left seat pilot was the pilot flying prior to 1,500 ft AGL.

The PPM contained information about ILS critical areas that was relevant to the conduct of an autoland and advised that:

When the weather ceiling is above 600 ft and 2,000 m visibility the ILS critical and sensitive areas are not protected by ATC and a number of factors may influence the accuracy of the ILS signal.

These include intrusions of the ILS critical and sensitive areas by the following factors:

• taxiing aircraft

• ground vehicles

• over-flight of the ILS localiser.

The flight crew shall therefore notify ATC at the commencement of an autoland approach when the ceiling and visibility are above the parameters listed above. e.g. ‘We are conducting an autoland approach’.

When an approach condition develops where the glide slope or the localizer becomes unreliable, due to signal interference, the aircraft departs from the approach path, or for other causes where manual flying is required, the flight crew must transition to manual flying immediately.

Flight crew training

The operator provided training to flight crew for LVO via a combination of online computer-based training (CBT), in-person training and training and assessment in a flight simulator.

The CBT included information about the purpose of an ILS critical area, and of alternate holding point requirements when protection was in force. Presentation materials used during in-person training identified the loss of glideslope or localiser signal as a possible non-normal situation. The indications expected in this situation were:

• pointer disappears

• line through associated FMA

• FD [flight director] bar for failed mode is removed

• autopilot EICAS [Engine indicating and crew alerting system]

The operator advised that simulator training during both induction and ongoing training covered failures during approaches, including the glideslope. In addition, they advised this training covered the cockpit indications during a glideslope out scenario. They also advised that the FCOM signal interference bulletin was discussed during training with focus on the potential high rates of descent associated with the issue.

Reviewed training materials did not contain information about the conditions under which the ILS critical area would be protected.

Related occurrences

ATSB investigation AO-2015-144

The ATSB investigated a flight below minimum altitude of a Boeing 787 conducting an ILS approach to Perth Airport on 4 December 2015. It was found that interference to the glideslope signal likely occurred due to a Boeing 737 aircraft taxiing from the holding point to the runway at the time.

ATSB investigation AO-2017-023

The ATSB also investigated a descent below lowest safe altitude of a Boeing 747‑400 aircraft conducting an ILS approach to Sydney Airport’s runway 16R on 12 February 2017. It was found that glideslope interference likely occurred due to a Boeing 787 aircraft holding at A1 during the approach. The report also identified 2 similar occurrences in March 2017 of Boeing 747 aircraft experiencing high descent rates during an ILS approach. In both instances an Airbus A380 was occupying taxiway A1 at the time.

Other occurrences

The ATSB occurrence database did not contain any additional occurrences after 2017 that could be identified as glideslope interference. However, as glideslope interference is not a reportable matter under the Transport Safety Investigation Regulations, it is likely an occurrence would only be reported to the ATSB if a consequential event required it and may not include detail to identify glideslope interference.

Following a notice to flight crew to report instances of glideslope interference, the operator was subsequently made aware of 9 additional instances of suspected glideslope interference when approaching runway 16R at Sydney Airport. For each, the ATSB reviewed publicly available flight data and identified an aircraft positioned either at holding point A1 or moving from the holding point onto the runway at the time of the reported interference.

Table 1: Subsequent occurrences of glideslope interference reported to operator

In response to a request for reports of glideslope interference at Sydney Airport, Airservices Australia advised that several of the subsequent events listed by the operator were received and reviewed by technical surveillance specialists. All events were determined to be due to aircraft in the glideslope critical area with the effect on the approaching aircraft as expected. No reports of glideslope interference were received from other operators between January 2025 and January 2026.

Safety analysis

Introduction

This analysis will discuss the factors leading to the performance of the autoflight system and the protection requirements of the instrument landing system critical area. In addition, it will examine the actions of the flight crew in response. Finally, the analysis will consider the operator’s procedures and training for conducting precision approaches, low visibility operations and practice autoland approaches.

Glideslope interference

As the aircraft was intercepting the glideslope prior to the commencement of a practice autoland approach, an A380 aircraft holding at A1, within the ILS critical area, began moving onto the runway. A second A380 then entered the critical area as the first vacated. The movements of both aircraft coincided with anomalies observed in the glideslope signal received by the aircraft.

The weather conditions at Sydney Airport at the time were better than those for which protection of the ILS critical area was required. Therefore, ATC was not required to protect the area for a practice autoland. Furthermore, while the tower controller was required to inform the flight crew that the ILS was not being protected, the flight crew had not yet transferred to this controller. Therefore, there was no opportunity for the flight crew to be given this advice.

After initially attempting to capture the glideslope from below, the aircraft's descent rate increased away from the flight director pitch guidance, and it began deviating away from the glideslope. As described in the flight crew operations manual (FCOM) bulletin for ILS signal interference, when the autopilot flight director system (AFDS) detected an unstable glideslope signal it changed to an attitude stabilising mode. Cockpit alerts were subsequently displayed to alert the flight crew to the degraded performance of the autopilot. These alerts were consistent with those expected when the AFDS had been in attitude stabilising mode for 15 seconds. Therefore, the movement of the first A380 through the ILS critical area caused an interference to the glideslope signal, which was subsequently detected by the AFDS. As a result, after attempting to capture the glideslope prematurely, the AFDS entered and remained in an attitude stabilising mode before alerting the crew. 

Descent below glideslope

Pilot flying

After the interference stopped, it is likely that the flight path deviation had increased beyond the threshold at which the AFDS could re-capture the glideslope. Therefore, the AFDS remained in attitude stabilising mode while the annunciations indicating the degraded performance of the autopilot continued to be displayed to the flight crew. 

The pilot flying recognised that the autopilot was no longer following the glideslope, and that the aircraft was descending at a high rate. However, believing that they had sufficient time and altitude, they did not disengage the autopilot and commenced discussion and preparation for a change to a localiser approach. Both operator and manufacturer procedures required that if automation was not operating as expected then it should be disengaged and the aircraft flown manually. However, the pilot flying allowed flight below the glideslope to continue beyond the initial approach fix and below the approach commencement altitude.

After the aircraft’s flight path triggered a minimum safe altitude warning (MSAW), but prior to receiving the corresponding low altitude alert from ATC, the pilot flying did disconnect the autopilot. By this time, the aircraft’s position was significantly below the minimum sector altitude, below which a missed approach was required to be conducted when experiencing a failure of the glideslope. Furthermore, the manufacturer’s procedure specific to glideslope interference required a missed approach to be conducted when corresponding failure annunciations were displayed. These annunciations were continuously displayed throughout the approach. Following the disconnection of the autopilot, the pilot further delayed the initiation of a missed approach. During this period of manual flight, the aircraft descended a further 150 ft and below the localiser segment minimum safe altitude before a missed approach was commenced.

The ATSB considered the extended time between the MSAW being triggered and the flight crew receiving a low altitude alert. As the MSAW coincided with an instruction to the crew to contact the tower controller, the low altitude alert could not be given immediately because the approach controller was required to establish which frequency the crew were listening on. However, as the crew had already disconnected the autopilot and arrested the initial descent, this was not considered to be contributory. 

Pilot monitoring

During the approach, the pilot monitoring was required to monitor the aircraft’s flight path and advise the pilot flying of any deviations. Specifically, when conducting an ILS approach, they were required to monitor the glideslope for deviation. If glideslope tracking deviated beyond one dot, a callout was required, and the flight path was required to be corrected by the pilot flying. When in instrument meteorological conditions (IMC) further deviation outside of this required a missed approach. At the commencement of the approach, the glideslope deviation was recorded as already at 2 dots (full scale) and remained so throughout the approach. However, both the pilot flying and the pilot monitoring could not recall that the glideslope deviation pointer was displayed after the autopilot alerts activated. Furthermore, the manufacturer could not determine whether the pointer was displayed to the flight crew after the AFDS entered attitude stabilising mode.

Notwithstanding this, other flight instruments were available to monitor the aircraft’s flight path. The aircraft’s pitch attitude was more nose down than expected during an ILS approach. In addition, the vertical speed indicator would have been indicating a descent rate greater than expected. The FCOM bulletin advised flight crew to monitor both pitch attitude and descent rate during an ILS approach. Furthermore, the relief pilot called out required altitudes and distances during the approach, providing additional information regarding the aircraft’s position below the glideslope. As no deviation calls or requests to conduct a missed approach were reported as being made by the pilot monitoring, it is likely that they were not effectively monitoring the aircraft’s flight path throughout the approach. 

Operator procedures and training

Glideslope out procedure

The operator had a ‘glideslope out’ procedure that allowed the crew the flexibility to transition to localiser approach provided specific preparatory actions were able to be completed prior to the final approach fix. Recognising that on this occasion the crew experienced glideslope interference rather than glideslope out, the cockpit indications were similar. As such, consideration of transitioning to a localiser approach was understandable.

However, given the significant deviation from the expected flightpath due to the aircraft’s sustained high rate of descent, safe transition to a localiser approach was considered highly unlikely on this occasion. This was ultimately recognised by the crew and a missed approach initiated.

The operator advised that the procedure allowing the transition from an instrument landing system approach to a localiser approach was withdrawn following this occurrence.

Practice autoland procedure

As approved by the operator, the flight crew briefed that while the first officer was initially the pilot flying, they would exchange flying and monitoring roles at 1,500 ft above ground level (AGL), after which point the captain would be the pilot flying. However, the operator’s procedures did not differentiate between a practice autoland and one conducted in low visibility conditions. Therefore, the captain was required to be the pilot flying from the commencement of the approach. 

In addition, while callouts required for an autoland did not commence until 1,500 ft AGL, there were specific responsibilities assigned to the flying and monitoring roles during an autoland approach that applied to the entire approach.

A practice autoland offered an opportunity to develop familiarity with, and proficiency in, those roles. Furthermore, the Civil Aviation Safety Authority guidance material highlighted the importance of flight crew responsibilities during transition to visual conditions, in deteriorating visibility or when failures occur. As such, exchanging roles introduced the potential for confusion or ambiguity as to the roles and responsibilities of each crew member during the approach, particularly if the handover coincided with a transition to visual conditions, or the need to respond to an abnormal event.

Low visibility operations training

The operator provided training to flight crew for low visibility operations, which included information about the ILS critical area. However, the training did not include information about the conditions under which the critical area would be protected. In addition, while the operator's procedures informed flight crew that they should notify ATC of their intent to conduct a practice autoland, they did not specify that this notification did not change ATC protection requirements, implying that the notification would result in the critical area being protected.

The captain reported that they considered that the weather conditions would necessitate protection of the critical area. Additionally, both the first officer and the relief pilot reported that they thought the ILS was being protected as they had advised ATC of their intention to conduct an autoland. Therefore, while it likely did not influence the flight crew’s response to the glideslope interference, they mistakenly believed that the risk of such an event had been mitigated.

Findings

From the evidence available, the following findings are made with respect to the descent below glideslope involving Boeing 767, VH-XQU, near Sydney Airport, New South Wales, on 10 September 2025. 

Contributing factors

• Two Airbus A380s on the ground at Sydney Airport taxied through the instrument landing system critical area and in front of the glideslope antenna, causing interference to the glideslope signal. As a result, after detecting the interference, the Boeing 767's autopilot established the aircraft on a flight path that deviated away from the glideslope, before alerting the crew that it was operating in a degraded mode.
• The pilot flying continued the approach with the autopilot in a degraded mode. As a result, the aircraft’s high descent rate triggered an air traffic control minimum safe altitude warning. After disconnecting the autopilot, the pilot flying delayed the initiation of a missed approach and the aircraft descended below the localiser segment minimum safe altitude.
• The pilot monitoring did not effectively monitor the aircraft's flight path during the approach and did not call out deviations or advise the pilot flying to conduct a missed approach.

Other factors that increased risk

• Tasman Cargo Airlines allowed the practice of flight crew exchanging flying and monitoring roles prior to 1,500 ft when conducting a practice autoland. This increased the risk of role confusion or ambiguity during a high workload activity.
• Tasman Cargo Airlines’ training did not inform flight crew of the conditions under which instrument landing system critical areas were protected. Consequently, the flight crew believed that the critical area was being protected, and the risk of glideslope interference had been mitigated.

Safety actions

Safety action not associated with an identified safety issue

Safety action by Tasman Cargo Airlines

Tasman Cargo Airlines has taken the following proactive safety action:

• Removed the ‘glideslope out’ procedure from its policy and procedures manual (PPM), replacing it with the requirement to conduct a missed approach and notify air traffic control.
• Added a note in the PPM referencing the glideslope interference bulletin from the flight crew operating manual.
• Published an operational alert regarding Sydney Airport runway 16R glideslope interference with guidance information.
• Updated computer-based training materials to better highlight the ILS critical area and requirements during normal and low visibility operations.
• Introduced a requirement for the pilot flying to take control of the aircraft prior to the commencement of the approach when conducting an autoland.

Safety action by Boeing

While not in direct response to this occurrence, Boeing is in the process of updating the flight control software for B767 aircraft. Planned for release in 2027, the update will include the following changes:

• The flight director pitch bar will remain biased out of view if the autopilot is disconnected while in attitude stabilising mode.
• Improved glideslope capture logic to reduce the occurrence of false captures leading to attitude stabilising mode.
• Limiting the flight path angle while in attitude stabilising mode to be between 0–3.25° of descent.
• Display of NO AUTOLAND after being in attitude stabilising mode for 15 seconds if the aircraft is still above alert height.
• Automatic autopilot disconnect 10 seconds after the display of NO AUTOLAND if the aircraft is still above 500 ft AGL.

Glossary

Sources and submissions

Sources of information

The sources of information during the investigation included:

• the flight crew
• the operator
• The Boeing Company
• Airservices Australia
• recorded data from the aircraft quick access recorder.

References

Civil Aviation Safety Authority (2024). Conduct of practice autoland operations (advisory circular AC 91-12 v1.1), https://www.casa.gov.au/sites/default/files/2021-12/advisory-circular-91-12-conduct-of-practice-autoland-operations.pdf, CASA, accessed 1 December 2025.

Submissions

Under section 26 of the Transport Safety Investigation Act 2003, the ATSB may provide a draft report, on a confidential basis, to any person whom the ATSB considers appropriate. That section allows a person receiving a draft report to make submissions to the ATSB about the draft report. 

A draft of this report was provided to the following directly involved parties:

• the flight crew
• the operator
• The Boeing Company
• Airservices Australia
• Civil Aviation Safety Authority
• United States National Transportation Safety Board.

Submissions were received from:

• the operator
• The Boeing Company
• Civil Aviation Safety Authority.

The submissions were reviewed and, where considered appropriate, the text of the report was amended accordingly.

Purpose of safety investigations The objective of a safety investigation is to enhance transport safety. This is done through:  identifying safety issues and facilitating safety action to address those issues providing information about occurrences and their associated safety factors to facilitate learning within the transport industry. It is not a function of the ATSB to apportion blame or provide a means for determining liability. At the same time, an investigation report must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. The ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action. About ATSB reports ATSB investigation reports are organised with regard to international standards or instruments, as applicable, and with ATSB procedures and guidelines. Reports must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. An explanation of terminology used in ATSB investigation reports is available here. This includes terms such as occurrence, contributing factor, other factor that increased risk, and safety issue. Publishing information Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2026 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Commonwealth Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this report is licensed under a Creative Commons Attribution 4.0 International licence. The CC BY 4.0 licence enables you to distribute, remix, adapt, and build upon our material in any medium or format, so long as attribution is given to the Australian Transport Safety Bureau.  Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

1. ^    Pilot flying (PF) and pilot monitoring (PM): procedurally assigned roles with specifically assigned duties at specific stages of a flight. The PF does most of the flying, except in defined circumstances, such as planning for descent, approach and landing. The PM carries out support duties and monitors the PF’s actions and the aircraft’s flight path.
2. ^    Flight level: at altitudes above 10,000 ft in Australia, an aircraft’s height above mean sea level is referred to as a flight level (FL). FL 350 equates to 35,000 ft pressure altitude.
3. ^    Runway number: the number represents the magnetic heading of the runway. The runway identification may include L, or R as required for left or right when there are parallel runways. 
4. ^    Minimum safe altitude warning (MSAW): an automated warning for air traffic controllers to draw attention to an aircraft that at its current descent rate is projected to conflict with terrain.
5. ^    Automatic terminal information service: the provision of current, routine information to arriving and departing aircraft by means of continuous and repetitive broadcasts. ATIS information is prefixed with a unique letter identifier and is updated either routinely or when there is a significant change to weather and/or operations. 
6. ^    Cloud cover: in aviation, cloud cover is reported using words that denote the extent of the cover – ‘few’ indicates that up to a quarter of the sky is covered, ‘scattered’ indicates that cloud is covering between a quarter and a half of the sky, ‘broken’ indicates that more than half to almost all the sky is covered, and ‘overcast’ indicates that all the sky is covered. 
7. ^    QNH: the altimeter barometric pressure subscale setting used to indicate the height above mean sea level.
8. ^    Instrument meteorological conditions (IMC): weather conditions that require pilots to fly primarily by reference to instruments, and therefore under instrument flight rules (IFR), rather than by outside visual reference. Typically, this means flying in cloud or limited visibility. 
9. ^    When vertical speed mode is engaged, it will initially target the vertical speed current at the time of engagement. 
10. ^   Cloud ceiling: The height above the ground of the base of the lowest layer of cloud covering more than one-half the sky.

Investigation summary

What happened

On 13 August 2025, a Cessna 560, registered VH-OHE, was being operated by Avcair on a medical transport flight from Learmonth Airport to Perth Airport, Western Australia, with 2 flight crew and 3 passengers on board.

As the aircraft neared Perth Airport at night, air traffic control cleared the flight crew to descend to 3,000 ft and conduct the instrument landing system (ILS) approach for runway 24. The aircraft was being flown using the autopilot and, while it was still descending to the assigned altitude, the crew agreed to set the approach decision altitude in the altitude selector. 

On reaching 3,000 ft, which was also the minimum sector altitude, the aircraft continued descending and did not capture the ILS glideslope. The crew attempted to correct the flight path using the automation system during which time the aircraft continued to descend, the autopilot disconnected, the aircraft entered an uncommanded left bank, and the enhanced ground proximity warning system (EGPWS) issued 2 alerts.

At about 500 ft above ground level, the captain took control and initiated a climb. Shortly after, air traffic control issued a low altitude warning safety alert to the crew. During the recovery, the aircraft reached a bank angle of about 50° and was manoeuvred back towards the ILS approach. In the late stage of the approach, the aircraft’s airspeed increased beyond the operator’s stabilised approach criteria. The flight crew continued the approach, landing a short time later.

What the ATSB found

The ATSB found the flight crew incorrectly set the approach decision altitude in the altitude selector before the autopilot had captured the ILS glideslope. The selection was contrary to the operator’s procedures and, in combination with the selected flight director modes, resulted in the autopilot commanding a continued descent towards terrain.

When the flight crew attempted to arrest the descent, they encountered difficulty locating the required flight director mode button on the first officer’s selector panel. The aircraft’s flight director mode selector panels had a fault that resulted in the intermittent illumination of mode selector buttons. This altered the visual presentation of the row of buttons at night and impacted the flight crew’s ability to locate the desired mode. A similar defect had been observed during a flight in the preceding days, however no endorsement of a defect was recorded in the aircraft’s maintenance release, and no maintenance action had been undertaken. 

The flight crew became preoccupied with the flight director mode selection, and their attention was diverted from monitoring the aircraft state. As a result, the aircraft entered an uncommanded descending left turn after the autopilot was disengaged, and the crew did not appropriately respond to the EGPWS alerts.

At 500 ft above ground level, the captain initiated a recovery climb. However, during this time the aircraft exceeded the maximum angle of bank, which resulted in an undesired aircraft state and increased the risk of a loss of control at a low altitude. The flight crew re‑established the aircraft on the approach, but as the aircraft descended through 400 ft the airspeed exceeded the operator’s stabilised approach criteria. However, rather than conducting the required go-around, the flight crew elected to continue the approach and landed the aircraft. 

Air traffic control issued a safety alert to the flight crew 35 seconds after the controller received an initial minimum safe altitude warning (MSAW) about the aircraft’s descent. This delay was a missed opportunity to alert the flight crew of the detected terrain proximity at a time when the aircraft was still descending. The delay was likely influenced by the controller’s exposure to previous nuisance MSAWs within the same area. While changes had been implemented to address nuisance MSAW alerts 2 months prior, and communicated to Perth controllers, the controller was not aware of these changes.

What has been done as a result

The operator of VH-OHE advised the ATSB that all company pilots were provided with additional ground-based training relating to the use of aircraft automation and crew resource management. The organisation also communicated to pilots the importance of reporting incidents in accordance with established operator procedures.

Safety message

This incident highlights how an unexpected disruption during an instrument approach can rapidly lead to a significant deviation from the intended flight path and result in an undesired aircraft state. In these scenarios, effective crew resource management, where each crewmember performs their procedurally assigned roles of flying and monitoring, is essential to ensure the continued safety of flight while the disruption is investigated and appropriately managed. If safe flight tolerances cannot be adequately monitored and maintained while the disruption is investigated, discontinuing the approach is the safest response. 

The International Civil Aviation Organization’s State of Global Aviation Safety report identified controlled flight into terrain (CFIT) as a global high risk that accounted for 10% of total fatal accidents worldwide by occurrence category in 2024. The adoption of the enhanced ground proximity warning system (EGPWS) has contributed to a reduction in the number of CFIT accidents, however the system serves as a last line of defence by alerting pilots to the hazardous proximity of terrain. Consequently, flight crews should respond immediately and appropriately to EGPWS alerts, as a delay can allow a developing terrain proximity event to result in a CFIT. 

Summary video

The occurrence

On the night of 13 August 2025, a Cessna 560, registered VH-OHE, was being operated by Avcair on a medical transport flight under the instrument flight rules¹ from Learmonth Airport to Perth Airport, Western Australia (WA). On board was the captain, first officer (FO), 2 medical personnel and a patient. The FO was the pilot flying (PF)² and the captain was the pilot monitoring (PM). 

This was the flight crew’s second sector for the day, having positioned VH-OHE from Karratha Airport, WA, to Learmonth Airport that afternoon for patient retrieval. The aircraft departed Learmonth Airport at 1912³ and proceeded to Perth Airport at flight level (FL)⁴ 410. 

At 2001, as the aircraft neared Perth Airport, the flight crew received air traffic control (ATC) clearance to conduct the JULIM 6A standard instrument arrival (STAR)⁵ for runway 24.⁶ At about 2025, the flight crew initiated a descent using the aircraft’s autopilot, which was coupled to the flight director (FD) with navigation and vertical navigation (VNAV) modes active (see Flight guidance and automation).

At 2037, as the aircraft descended through FL170, the flight crew were transferred to the Perth ATC approach controller (approach controller) who issued further descent and heading clearances. 

At 2042:45, VH-OHE was maintaining a heading of 160° when the approach controller advised the flight crew that the aircraft was 3 NM (5.6 km) right of the localiser.⁷ The flight crew were instructed to turn onto a heading of 210°, descend to 3,000 ft, and were cleared to conduct the instrument landing system (ILS)⁸ approach for runway 24 (Figure 1). The forward-facing camera installed in VH‑OHE’s cockpit captured video and audio of the flight crew’s actions and cockpit instrumentation. The video from the incident flight recorded the captain acknowledging the clearance while the FO set 3,000 ft in the aircraft’s altitude preselector (ASEL) (see Altitude preselector). 

Figure 1: VH-OHE runway 24 approach flight path

At 2043:56, as the aircraft neared the runway 24 ILS localiser, the FO selected the approach FD mode, which armed the localiser and glideslope (see Flight director mode selector). The FD immediately captured the localiser, and the autopilot initiated a right bank to follow the FD guidance while the aircraft continued descending. The FO then announced, ‘localiser captured.’ At this time, the aircraft was 17 NM (31 km) from the runway threshold and below the ILS glideslope.

At 2044:52, the altitude alert⁹ activated passing 4,000 ft, and the captain announced ‘1,000 [ft] to go,’¹⁰ which was acknowledged by the FO. The captain then asked the FO if they would like to set ‘260 [ft] as the minima’ in the ASEL, which the FO agreed to. 

On passing 3,380 ft, the captain extended the landing gear and selected the ‘take off and approach’ flap setting of 15°. At 2045:25, the captain initially set 0 ft, then 300 ft in the ASEL, which they later recalled was the approach decision altitude.¹¹

At 2045:38, the aircraft passed through the assigned altitude of 3,000 ft, which was also the applicable minimum sector altitude (MSA) (see Minimum sector altitude), and the FO enquired why the aircraft had not maintained 3,000 ft or captured the glideslope. The captain responded that the aircraft was below the glideslope, with pitch mode engaged and queried why the aircraft was continuing to descend. The FO responded they were unsure of the reason. 

At 2046:11, the Perth Airport approach controller position underwent a shift swap, with the incoming approach controller taking over as VH‑OHE descended through 2,700 ft.

At 2046:15, the captain instructed the FO to engage the vertical speed (VS) FD mode to reduce the aircraft’s rate of descent. At this time the autopilot disconnected, which was accompanied by an audible tone and the ‘AUTOPILOT OFF’ light which illuminated for one second (see Autopilot control panel). This occurred while the FO was attempting to locate and activate the VS mode button on the FD mode selector panel, however, neither pilot recalled disengaging the autopilot at this time. The FO later recalled encountering difficulty locating the desired button due to a fault with their FD mode selector panel back-lighting (see Flight director mode selector panel lighting fault). The captain attempted to assist the FO to locate the VS mode button and, at 2046:20, the aircraft began to bank to the left while the rate of descent increased.

At 2046:28, the aircraft’s rate of descent reached 2,000 fpm as the aircraft descended through 2,150 ft with 20° left angle of bank and 8° nose down pitch attitude (Figure 2 and Figure 3). During this time the captain and FO were still attempting to engage the VS button on the FO side FD mode selector. 

Figure 2: Captain’s side primary flight display at 2046:28

Figure 3: VH-OHE flight path below minimum sector altitude

At 2046:32, the aircraft descended through 1,960 ft (1,000 ft above ground level [AGL]) with a rate of descent of about 2,000 fpm and an angle of bank of 22°. At this time, an enhanced ground proximity warning system (EGPWS) alert was generated which presented to the flight crew as a visual alert on their electronic attitude direction indicators (EADI),¹² and an accompanying aural ‘caution terrain, caution terrain,’ alert sounded in the cockpit (see Enhanced ground proximity warning system). Neither of the flight crew recalled observing either the visual annunciations or the audible alert. 

At the same time the flight crew received the EGPWS alert (2046:32), a minimum safe altitude warning (MSAW) alert was triggered for VH‑OHE on the approach controller’s air situation display (see Perth Airport air traffic control). In response, the approach controller assessed that VH‑OHE was above the minimum safe altitude and cancelled the audible alert, however the visual alert remained on the situation display. They later recalled the MSAW alert occurred in an area where they had observed numerous erroneous MSAWs in the year prior to the incident (see Nuisance MSAW alerts).

At 2046:37, the captain re-engaged the autopilot while the FO began experiencing issues with their headset boom microphone that prevented them communicating over the cockpit interphone system. The captain then advised the FO, ‘now the autopilot is engaged,’ which the FO did not recall hearing at the time. The FO persisted attempting to hand-fly the aircraft while it continued descending with the autopilot engaged. 

At 2046:48, as the aircraft descended through 1,650 ft (650 ft AGL), an EGPWS alert appeared on the EADI accompanied by a ‘terrain terrain, pull up’ aural alert, which the captain later recalled observing. At this time, the captain was directing the FO to ‘go right to the glideslope.’  

At 2046:55, the autopilot disengaged a second time with an accompanied audible tone and ‘AUTOPILOT OFF’ light, which remained illuminated for the remainder of the flight (see Autopilot control panel). This occurred while the FO was still attempting to hand-fly the aircraft. At about 500 ft AGL, the captain announced they were taking control of the aircraft. The FO recalled first observing an EGPWS alert after the captain took over, while neither pilot recalled disengaging the autopilot. Shortly after, the captain advanced both thrust levers and established the aircraft in a climb on a westerly heading at 2047:06. 

Around this time, the approach controller recalled observing a second MSAW alert for VH‑OHE, that was triggered at 2046:46, and recalled observing the aircraft was between 1,500‍–‍1,700 ft. At 2047:07, the approach controller broadcast, ‘oscar hotel echo, low altitude warning, check your altitude immediately on QNH 1012, lowest safe altitude is 2,800 [ft]’. The FO, who was now PM, initially attempted to acknowledge the warning, but they were still experiencing issues with their headset boom microphone and could not make themselves heard by the controller. 

At 2047:24, the aircraft reached a maximum climb rate of 4,000 fpm and was 0.2 NM (0.4 km) right of the runway 24 centreline and below the glideslope. In response, the captain initiated a left turn and progressively increased the angle of bank. In doing so the aircraft reached a maximum angle of bank of 50° and 13° nose up pitch at an altitude of 2,200 ft (1,250 ft AGL) (Figure 4). The angle of bank was not verbalised by either the captain or the FO (see Flight tolerances and approach call outs).

Figure 4: Captain’s side primary flight display at 2047:27 

The captain subsequently reduced the angle of bank as the aircraft became established on a south-westerly heading and continued climbing at over 2,000 fpm to regain the desired glidepath. Up to this point, the flight crew had not discussed their intentions following the recovery.

At 2047:37, the FO resolved the issue with their headset and established contact with the approach controller who asked ‘oscar hotel echo, just confirm you’re established on the localiser … we had you a little bit low.’ The FO then advised the approach controller ‘affirm … we corrected that.’ During this time, the aircraft was beyond half‑scale deflection to the right of the localiser abeam the initial approach fix (IAF)¹³ for the ILS approach. 

At 2047:46, the approach controller asked the flight crew if they intended to continue with the approach, to which the FO responded that they would, and they were subsequently transferred to Perth ATC tower controller (tower controller).   

At 2048:07, the FO contacted the tower controller and advised that they were ‘established.’¹⁴ At this time the aircraft was 7.5 NM (13.9 km) from the runway 24 threshold, established on the localiser and was beyond half-scale deflection above the glideslope. At 2048:39, the captain requested ‘more flap,’¹⁵ which was actioned by the FO, followed shortly by the landing checklist. 

At 2050:07, the tower controller advised the flight crew that the wind direction at the airport was from 250° with a velocity of 27 kt and cleared VH‑OHE to land on runway 24. As the aircraft passed 1,000 ft, the flight crew acknowledged the altitude with a ‘checked’ call, which coincided with the operator’s stabilisation height during an instrument approach (see Stable approach criteria). Both the captain and FO sighted the runway as the aircraft descended through 800 ft.

At 2051:38, the aircraft descended through 430 ft, and the airspeed increased above the operator’s stable approach criteria, peaking at 133 kt. The airspeed remained above the maximum allowable airspeed until the aircraft passed 200 ft at 2052:09 (see Stable approach criteria). Neither the captain nor the FO verbalised the airspeed exceedance, and they continued with the approach. The aircraft touched down on runway 24 at 2052:28. 

Context

Personnel information

Captain

The captain held an Air Transport Pilot Licence (ATPL) (Aeroplane) and a class 1 aviation medical certificate. They had a total of 7,490 hours total flying time, with 470 hours on the Cessna 560 (560). They held a flight examiner rating for the 560 and were employed as a training captain with Avcair. On 5 May 2025, the captain was assessed as competent during an operator proficiency check (OPC) conducted in an approved 560 flight simulator.

First officer 

The first officer (FO) held an ATPL (Aeroplane) and a class 1 aviation medical certificate. They had a total time of 9,374 hours, with 50.6 hours on the Cessna 560. 

All the FO’s 560 flying time had been accumulated during their initial type rating and subsequent line training with Avcair. For the initial 560 type rating, Avcair engaged a third-party training organisation, which conducted the training in an approved 560 flight simulator. The captain involved in the incident flight had also conducted part of the FO’s subsequent line training in the 560, which accounted for 12.6 hours. After successfully completing their line training, the FO was assessed as competent during an OPC conducted in the 560 on 21 July 2025. The incident operation on 13 August 2025 was the first flying the FO had conducted since completing their line training.

Fatigue

Flight and duty records indicated that both flight crew members had been on standby duties for the preceding 5 days in Karratha, WA, and had signed on for duty at 1423 local time on the day of the incident. Both pilots reported obtaining at least 8 hours of sleep in the 24-hour period preceding the incident. At the time of the incident, the captain reported feeling ‘somewhat fresh,’ while the FO reported feeling ‘responsive but not at peak.’

The ATSB considered that fatigue was unlikely to have affected the flight crew’s performance at the time of the incident.  

Aircraft information

General information 

VH-OHE (Figure 5) was a Cessna 560 equipped with 2 Pratt & Whitney Canada turbofan engines. The aircraft was manufactured in 1995 and was first registered in Australia in 1999. Avcair became the registered operator of the aircraft in 2021. 

The aircraft was configured for medical transport flights. The last periodic inspection was completed on 14 July 2025. At this time, the aircraft had accumulated 10,589 hours total time in service. At the time of the incident, the aircraft had accumulated an additional 24 hours since the last periodic inspection. 

Figure 5: VH-OHE

Operational information

Flight guidance and automation 

General information

The Cessna 560 is equipped with a flight director (FD) and autopilot system capable of providing flight guidance and automatic flight control. The aircraft can be flown manually using the guidance provided by the modes selected on the FD. The aircraft may also be flown using the autopilot. When coupled to the flight director, the autopilot will control the aircraft using the commands generated by the FD computer. The autopilot can also be used without any inputs from the FD system, in pitch, roll and heading hold.

Autopilot control panel

The autopilot control panel (ACP), mounted on the cockpit pedestal, provides the means of engaging the autopilot (Figure 6). The push-on/push-off autopilot (AP) button engages the autopilot and yaw damper (YD) simultaneously. 

Figure 6: Autopilot control panel

The autopilot is normally disengaged by depressing the AP/TRIM DISC switch on either the captain’s or FO’s yoke, electrically trimming the elevator trim system, or depressing the go-around button on either throttle. In these scenarios, a warning tone will sound for one second and the amber AUTOPILOT OFF light located adjacent to each pilot’s primary flight display (PFD) will illuminate for one second. Any other method of autopilot disconnection, such as pressing the AP button on the ACP, will cause the warning horn to sound for one second and the amber AUTOPILOT OFF light to stay illuminated until extinguished by the flight crew. The autopilot will also disengage if a sustained force is applied to the vertical or horizontal axis. 

Flight director mode selector panel 

The flight director can be activated by either pilot selecting an FD mode using the FD mode selector panel located above each pilot’s PFD. The panel incorporates 7 push‑on/push-off buttons that engage the selected FD mode(s) (Figure 7).

Figure 7: Flight director mode selector panel

Each button incorporates white background lighting (when selected on) and a green activation light when the corresponding mode is armed or captured. The FD mode status is also annunciated on both pilot’s PFDs, with white text indicating an armed mode and green text indicating a captured or active mode.

Each FD mode button arms or activates a lateral or vertical guidance mode, or both in the case of the approach (APR) mode. The selectable flight director modes relevant to the incident flight are summarised in Table 1.

Table 1: Relevant FD modes and functions 

If the autopilot is engaged after FD activation, it is automatically coupled to the FD and follows the FD computer generated flight path. In scenarios where an active vertical FD mode is disengaged, while the autopilot is engaged, the autopilot will hold the pitch attitude of the aircraft at the moment of disengagement.  

Altitude preselector 

The altitude preselector (ASEL) enables flight crew to set a preselected altitude, in 100 ft increments, in the altitude window at the lower right corner of the multifunction display (MFD).¹⁷ The preselected altitude is also presented on the pilot’s PFD. 

As the aircraft approaches the preselected altitude, the FD provides pitch commands to the autopilot (when engaged) to level off and maintain the altitude. While in a descent, if a lower altitude is selected before levelling off, the FD will continue descent in the active vertical mode until the aircraft reaches the preselected altitude or captures the glideslope when the APR mode is armed.

As required by the Avcair Operations Manual (AOM), during the descent, the flight crew would use the ASEL to set the desired altitude. When conducting an instrument landing system (ILS) approach the Avcair Aircraft Operating Procedures (AOP) – Approach and Landing Callouts stated: 

Once the glide slope has been captured, the altitude pre-selector is set to the missed approach altitude.

The missed approach altitude for the Perth runway 24 ILS approach is 3,000 ft (see Instrument landing system approach).

A review of the flight crew’s records between 1 April 2025 and the incident indicated the majority of flights conducted were into airports that were only equipped with non‑precision approaches. When conducting a non‑precision approach, the AOP stated:

The crew must set MDA +50 feet as a company DA for any occasion an MDA must be set. The amended minima must be entered in the ASEL.

Flight director mode selector panel lighting fault 

The cockpit video recording of the incident flight indicated that the during the descent and approach, the altitude hold (ALT) FD mode selector panel button on the captain’s side was not illuminated although it was on the FO’s side. Similarly, the FO’s flight level change (FLC) FD mode selector panel button was not illuminated but was on the captain’s side (Figure 8). 

Figure 8: FD mode selector panel lighting taken at 2048:57

The captain and FO recalled that the backlighting on the FD mode selector panel had operated intermittently prior to the incident flight. Both pilots reported being aware of the practice of rubbing or tapping the affected button to rectify the fault. During the incident flight, the cockpit video recording also captured a similar fault with the FO side APR mode selector button, but it was rectified in flight by the FO when they tapped the button.

An image supplied to the ATSB for a flight in VH-OHE, that took place several days prior to the incident flight, documented the same illumination issue with the ALT, VS and FLC buttons on the FO side FD mode selector panel (Figure 9). ‘Nil defects’ was recorded in the notes section of the corresponding flight log, and no defect had been recorded in the aircraft maintenance release (MR)¹⁸ following this flight. 

Figure 9: Supplied image of VH-OHE cockpit taken on 7 August 2025

As required by the AOM, in the event that a defect was identified following the completion of a flight, the captain was responsible for recording this in the notes section of the aircraft flight log and reporting the issue to the head of flight operations. 

The Avcair Maintenance Control Manual (MCM) described the procedures for reporting of defects which stated:

Defects shall be entered in Part 2 of the Maintenance Release to enable an evaluation and appropriate remedial action by Maintenance Staff and the HAAMC¹⁹ and MC.²⁰

The MR document used for VH-OHE was a standard Civil Aviation Safety Authority (CASA) form 918. Part 2 of the MR was used to record defects that occurred during the period of validity. Regulation 50 of the Civil Aviation Regulations 1988 outlined the defects and major damage to be endorsed on the MR and stated:

(1) This regulation applies to each of the following persons:

       (a) the holder of the certificate of registration for an Australian aircraft;

       (b) the operator of an Australian aircraft;

       (c) a flight crew member of an Australian aircraft.

 (2) If:

       (a) there is a defect in the aircraft; or

       (b) the aircraft has suffered major damage;

a person mentioned in subregulation (1), who becomes aware of the defect or damage, must endorse the maintenance release of the aircraft or other document approved for use as an alternative for the purposes of this regulation, setting out the particulars of the defect or damage, as the case may be, and sign the endorsement.

The Avcair internal investigation report into the incident found that the crew did not report the incident to the relevant personnel after landing at Perth, or prior to the subsequent sector in accordance with the AOM procedure for ‘reporting and recording defects and incidents.’ As a result, the aircraft’s cockpit voice recorder was overwritten before the event was reported and so was not available for the ATSB investigation.

The operator also advised the ATSB that no maintenance action had been carried out on the FD mode selector button lighting in the 28-day period prior to the incident. A review of the aircraft MR that was valid for the incident flight did not contain endorsements for defects relating to the flight director mode selector panel lighting.

Enhanced ground proximity warning system 

The Cessna 560 is fitted with a Honeywell enhanced ground proximity warning system (EGPWS), which uses the aircraft geographic position, aircraft altitude and a terrain database to predict potential conflicts between the aircraft flight path and terrain. The system provides pilots with aural alert messages, and visual annunciations which are displayed on each pilot’s PFD.

The Aircraft Flight Manual Supplement for the EGPWS prescribed the flight crew’s initial terrain avoidance actions following activation of a visual annunciator or aural terrain alert. 

For activation of the aural 'caution, terrain' alert or amber GND PROX annunciator on the PFDs, the initial action for the pilot flying (PF) was to take immediate positive corrective action to adjust the flight path, or correct the aircraft configuration, until the alert ceased.

For activation of the aural 'terrain, terrain, pull up' alert or red PULL UP annunciator on the PFDs, the initial actions for the PF were to disconnect the autopilot, aggressively apply maximum thrust, roll the wings level, and rotate at 3° per second to an initial pitch attitude of 20°. 

The Avcair Training and Checking Manual required flight crew undertake initial EGPWS ground training on commencement with the organisation, and recurrent training every 2 years. The captain and first officer last completed the EGPWS ground training on 4 December 2023 and 19 March 2025 respectively.   

Flight tolerances and approach call outs

The AOP contained call outs and challenges to be performed by the pilot monitoring (PM) when flight tolerances were exceeded and the required responses by the PF (Table 2).

Table 2: Relevant flight path deviations and standard calls 

Standard deviation calls also play an important role in the early identification of a developing undesired aircraft state (UAS). A UAS may result from a pilot induced aircraft position or speed deviations, misapplication of flight controls, or incorrect systems configurations associated with a reduced margin of safety. For safe flight, pilots must quickly recognise and recover from an undesired aircraft state before it leads to a loss of control or uncontrolled flight into terrain (Civil Aviation Safety Authority, 2025).

Stable approach criteria 

A stabilised approach is one where an aircraft maintains a constant descent to the runway while other key flight parameters such as airspeed and aircraft configuration are maintained within defined flight tolerances.

The AOM required all approaches conducted in instrument meteorological conditions be stabilised no later than 1,000 ft above ground level (AGL), and 500 ft AGL in visual meteorological conditions. An approach complied with the operator’s stabilised approach criteria when the following conditions were met: 

• All briefings completed and checklists have been actioned... 

• The landing gear shall be down and locked, and the landing flap should be selected.

• The aircraft speed (excluding momentary excursions defined as a deviation lasting only a few seconds and where every indication is that it will return within the stabilised criteria) is not less than VREF²¹ and not more than the lower of (Exception 3)

          ◦ Structural limitations for the configuration; and 

          ◦ Instrument approach handling speeds for the relevant performance category; and 

          ◦ Speed as per below to ensure a controlled and progressive reduction in speed:

                    - HAT²² + 1000’ at VREF + 60 KTS 

                    - HAT + 900’ at VREF + 50 KTS 

                    - HAT + 800’ at VREF + 40 KTS 

                    - HAT + 700’ at VREF + 30 KTS 

                    - HAT + 600’’ at VREF + 20 KTS 

                    - …HAT + 500’ at VREF + 10 KTS 

• The aircraft is on the correct flight path within the approach path tolerances laid down in the AIP [Aeronautical Information Publication] for the type of approach being flown and only minor corrections are required to be made. 

• Vertical speed is no more than 1,000 feet per minute. 

• Power setting is appropriate for the aircraft configuration. 

500’ HAT and Below Stabilised Approach Requirements 

At or below 500 ft HAT, if any of the Stable Approach Criteria are no longer met or destabilise (except for momentary excursions of speed), then the PIC in single pilot operations or the PM shall announce “GO AROUND” and an immediate go-around shall be flown.

The AOM also contained the recommended method for calculating the final approach and threshold speeds. It stated:

The Company recommended method of approach speed correction is to add one half of the reported steady headwind component plus the full gust increment above the steady wind to the reference speed. When adjusting for wind velocity, the maximum command speed should not exceed VREF + 20 knots or landing flap placard speed²³ minus 5 knots, whichever is lower.

The reported wind component of 250° at 27 kt provided by the Perth tower controller prior to landing likely resulted in a headwind component of 27 kt during the final approach. Based on the operator’s recommended method for calculating the approach speed, the applicable airspeed for the incident approach was 14 kt above the VREF of 103 kt, which equates to 117 kt.

In response to the draft investigation report, the captain advised the ATSB it was their understanding that a wind correction would result in a higher VREF, and assessed that the HAT 500 ft stabilised criteria of VREF + 10 kt during the incident approach equated to 127 kt.

Meteorological information

Automatic traffic information service

The Perth Airport terminal information that was valid during the incident advised the following weather conditions:

• expect instrument landing system approach for runway 24 
• whole runway wet
• wind direction 240–320° at 20 kt (maximum crosswind 20 kt)
• 10 km visibility reducing to 3 km
• thunderstorms and associated rain showers
• few²⁴ cloud at 600 ft, broken²⁵ cloud at 1,200 ft and 2,300 ft.

Airport information

Perth Airport

Perth Airport had 2 runway strips oriented 016°/196° and 061°/241° magnetic (03/21 and 06/24). Runway 24 was equipped with a category 1²⁶ ILS and provided the typical 3° glideslope to the runway. 

Instrument landing system approach

The Perth runway 24 ILS approach commenced at KOORI (initial approach fix) located 8.2 NM (15.2 km) from the runway threshold, which had an elevation of 58 ft. Aircraft conducting the ILS approach could descend to the decision altitude of 260 ft, which was 202 ft above the runway threshold. If a missed approach was required after the approach had commenced, the published procedure was to track 240° and climb to 3,000 ft (Figure 10).  

Figure 10: Perth runway 24 ILS approach chart

Minimum sector altitude

The minimum sector altitude (MSA) is the lowest usable altitude that provides at least 300 m (or 1,000 ft) clearance above all objects within a circle or a sector of a circle of radius 46 km (25 NM) or 18.5 km (10 NM) centred on a significant point. If a flight crew encounters an issue that requires immediate terrain avoidance, climbing to the MSA provides safe separation from obstacles within that sector.

The MSA for the Perth runway 24 ILS approach was 3,000 ft within a 46 km (25 NM) radius of the Perth very high frequency omnidirectional range station (VOR).²⁷

Perth Airport air traffic control

General information

The Perth terminal control unit, located at Perth Airport, provided air traffic services, including approach control, for aircraft arriving at Perth Airport. Approach controllers were responsible for adjusting the sequence of arriving aircraft to achieve the correct priority as defined in the Airservices Australia Manual of Air Traffic Services (MATS).²⁸

Approach controllers relied on an air situational display, which depicted an aircraft’s radar derived position overlaid on a composite of the Airservices Australia terminal area chart. The air situation display also included the aircraft’s velocity vectors, history dots as well as an aircraft label which presented the aircraft’s callsign, type and present altitude (Figure 11).

Figure 11: Air situational display

Minimum safe altitude warning

The minimum safe altitude warning (MSAW) function was a software enhancement that relied on existing radar hardware to alert air traffic controllers to an increased risk of an aircraft collision with terrain. The system provided the controller with a warning time of 60 seconds. When an MSAW was triggered, controllers were presented with an aural and visual alert (Figure 12).

Figure 12: MSAW alert for VH-OHE triggered at 2046:32 

The Airservices Australia National Air Traffic Services Procedure Manual (NAPM) categorised an MSAW as a ‘priority 1’ situation display alert, which required immediate intervention by the controller. The NAPM stated that:

On receipt of an MSAW, controllers must:

1. assess its integrity 

2. issue a 'Safety Alert' when appropriate.

The phraseology to be used by ATC when providing safety alerts and avoiding action was contained in the AIP. An example of a generic ‘low altitude warning’ alert was:

SAFETY ALERT, LOW ALTITUDE WARNING, CHECK YOUR ALTITUDE IMMEDIATELY, QNH IS (number) [(units)]. [THE MINIMUM SAFE ALTITUDE IS (altitude)].

Nuisance MSAW alerts 

In late 2023, Airservices Australia increased the trigger level for MSAW alerts within the vicinity of Perth Airport. The change resulted in a significant number of nuisance MSAW alerts for aircraft conducting the approach for runway 24. In December 2024, Airservices Australia finalised a safety case assessment and reporting determination (SCARD) to evaluate further changes intended to address the nuisance alerts. The SCARD stated:

About a year ago, the MSAW levels around Perth were raised (only marginally by about 30 feet or so) but it was enough to generate MSAW alerts for almost every aircraft arriving on RWY24, particularly when these aircraft are on a 4-5 NM final approach at approximately 1700 FT which is a normal level for that distance to touch. This issue is creating a situation where controller[s] are becoming desensitised to the alarms.

The SCARD assessed a proposal for a revised MSAW inhibit area which encompassed 1 NM either side of the runway 24 centreline and extended to 6 NM from the runway threshold and would be activated 24 hours a day. The proposal also stated:  

​No training required for new MSAW area, as the system is not currently functioning as expected. Local Instructions will be updated to desc[r]ibe the new area and method of operation.

The new area and method of operation of the MSAW inhibition area were incorporated into the Perth Terminal Control Unit (TCU) local instructions on 12 June 2025. Airservices Australia reported that information relating to the MSAW changes were also included in a corresponding New Information Register, which air traffic controllers were responsible for reviewing and acknowledging prior to commencing duties. The ATSB was unable to obtain a copy of the applicable New Information Register due to the MATS records retention procedures, which required records to be retained for a period of 30 days only.

At the time of the incident flight (August 2025), the approach controller reported they were not aware of any changes that had been implemented to address nuisance MSAW alerts at Perth Airport and did not recall if either the local instructions or New Information Register contained any related information.  

Recorded information

Airservices Australia recorded data

Airservices Australia provided the ATSB with radar and local audio recordings of the approach controller’s console during the incident on 13 August 2025. Air traffic control (ATC) audio recordings and radar data records were also provided. 

Cockpit video camera

Cockpit video recordings from the incident flight were obtained and reviewed by the ATSB. The recordings captured pilot actions, flight instruments and aural alerts during the approach and subsequent recovery. 

Flight data recorder

Flight data from the aircraft’s flight data recorder (FDR) was downloaded and analysed by the ATSB. The information sourced from the flight data recorder was synchronised and combined with the cockpit video recordings using common events present in both sources.

Related occurrences

ATSB investigation AO-2024-030

On 9 May 2024 an Embraer ERJ 190-100 IGW aircraft, registered VH-UZI, was being operated on a flight from Cairns to Brisbane, Queensland. During the approach, the pilot flying disconnected the autopilot and shortly after, the aircraft’s automated ILS flight mode unexpectedly disengaged. Following the unexpected change to the aircraft’s flight mode, the flight crew diverted their attention to recapturing the ILS approach mode and did not effectively monitor the aircraft's flight path. 

Consequently, the aircraft exceeded the glideslope limit requirement of the stabilised approach criteria. This was not detected by the flight crew. The aircraft continued to descend below the glideslope, resulting in the EGPWS glideslope alert activating. Subsequently, the flight crew did not perform the required terrain avoidance manoeuvre and instead continued the approach.

Safety analysis

Altitude preselector selection

While the first officer (FO) was flying the aircraft using the autopilot in heading and vertical navigation flight director (FD) modes, Perth Airport air traffic control (ATC) provided a clearance to descend to 3,000 ft and conduct the runway 24 instrument landing system (ILS) approach. In response, the FO set 3,000 ft in the altitude preselector (ASEL) and armed the localiser (LOC) and glideslope (GS) FD modes.

The operator’s procedures required the ASEL to be set to the ATC assigned altitude then the missed approach altitude after glideslope capture, which was also 3,000 ft for the runway 24 ILS. This meant no further ASEL selections were required for the approach. However, the flight crew agreed to set the approach decision altitude of 260 ft in the ASEL, which could only be set in 100 ft increments, resulting in an assigned altitude of 300 ft. This was actioned by the captain before the autopilot captured 3,000 ft or the ILS glideslope and, as a result, the aircraft continued descending below 3,000 ft (towards 300 ft). 

The selection of the decision altitude in the ASEL was similar to the operator’s procedure for conducting a non-precision approach, which required the flight crew to set the minimum descent altitude +50 ft in the ASEL. In the 19 weeks leading up to the incident, the majority of the airports the flight crew operated to were only equipped with non‑precision approaches. Consequently, it is possible that the incorrect selection of the ILS DA in the ASEL was influenced by the flight crew’s recent repeated exposure to non‑precision approach procedures.

When the aircraft did not level off at 3,000 ft as expected by the flight crew, the continued descent was queried by the FO. In response, the captain identified that the aircraft was below the glideslope in the pitch hold FD mode, which would activate when an active vertical FD mode is disengaged while the autopilot was engaged. The ATSB was unable to determine why the pitch hold FD mode activated at that time.

The captain then directed the FO to select the vertical speed (VS) FD mode to arrest the descent. However, the FO encountered difficulty locating the VS button on the FD mode selector panel, at which point the captain also attempted to assist the FO without success. Both the captain’s and FO’s FD mode selector panels had a lighting fault. On the FO’s side, the VS mode button was illuminated, but the adjacent button to the right was not. This altered the visual presentation of the row of buttons at night and impacted the flight crew’s ability to locate the desired mode. Additionally, the FO’s limited experience flying the 560 may have further impacted their ability to locate the VS FD mode button.

Shortly after, the autopilot disengaged, which was likely the result of the FO depressing the ‘AP/TRIM DISC’ switch on their yoke or electrically trimming the elevator trim system. 

While the flight crew were attempting to locate the VS FD mode button, their attention was diverted from monitoring the aircraft state. When attention is directed to resolving an unexpected problem, monitoring of other information sources degrades, often without the flight crew being aware of it (Civil Aviation Authority, 2023). In this case, the flight crew’s attention shifted from monitoring the aircraft’s flight path to managing the automation system as the descent continued unchecked.

As the aircraft continued to descend, the flight crew were presented with 2 EGPWS alerts. The first occurred as the aircraft passed 2,100 ft (1,130 ft AGL) and was accompanied by a ‘caution, terrain.’ However, neither the captain nor the FO recalled being aware of the alert and they did not initiate the required terrain avoidance procedures. This was consistent with the known effects of narrowed attention, whereby other events and stimuli that would normally attract attention, including highly salient alerts, fail to do so (Civil Aviation Authority, 2023). At this point, the captain re-engaged the autopilot.

The second EGPWS alert occurred as the aircraft descended through 1,650 ft (650 ft AGL) and was accompanied by a ‘terrain terrain, pull up’ aural alert. The captain recalled being aware of the second alert and, after briefly directing the FO to fly towards the glideslope, they took control of the aircraft. During this time, the autopilot disengaged a second time, which was likely a result of the FO applying a sustained force through their control yoke in either the vertical or horizontal axis.

Recovery and unstable approach

At an altitude of about 1,480 ft (500 ft AGL) the captain took control of the aircraft and initiated a recovery. However, the recovery was not consistent with the required terrain avoidance procedures. During the recovery, the captain’s control inputs resulted in the aircraft reaching a bank angle of 50° with a pitch angle of 13° at around 2,100 ft (1,100 ft AGL). This exceeded the operator’s maximum bank angle deviation limit of 25°. However, neither of the flight crew performed the ‘bank angle’ deviation call as required by the Aircraft Operating Procedures or took timely action to correct the exceedance. 

Although the aircraft was subsequently re-established on the correct approach path, the manoeuvring during the recovery was consistent with the development of an undesired aircraft state (UAS). A UAS can occur due to flight crew action or inaction and results in the aircraft being placed in a condition of unnecessary risk (International Civil Aviation Organization, 2002). In this instance, the high angle of bank increased the risk of a loss of control at a low altitude.

Following the recovery, the flight crew elected to continue the ILS approach for runway 24. On passing 1,000 ft, the aircraft was within the operator’s required flight tolerances to continue the approach. However, as the aircraft continued below 500 ft, the airspeed exceeded the operator’s stabilised approach criteria for a period of about 31 seconds, reaching a maximum of 133 kt. The stable criteria required the aircraft to be operated at a calculated airspeed of no faster than 117 kt. Any exceedance of this speed required the flight crew to conduct a go-around. Neither crew member verbalised the speed exceedance, as required by the operator’s standard calls, and they continued with the approach and subsequently landed.

Air traffic control response to MSAW

The Perth Airport approach controller (approach controller) received an initial minimum safe altitude warning (MSAW) on their air situation display for VH‑OHE as the aircraft continued its descent below the assigned altitude. The approach controller cancelled the audible alert associated with the MSAW and assessed that the aircraft was likely above the minimum safe altitude of 2,700 ft. This is despite their air situational display indicating the aircraft’s altitude at that time was 2,100 ft.

The MSAW was observed in an area where the approach controller and Airservices Australia had previously identified an increase in the frequency of nuisance MSAW alerts due to a system change. A high number of nuisance alerts decreases controller trust in automation systems and desensitises controllers to genuinely hazardous situations (Federal Aviation Administration, 2008). Airservices Australia had acknowledged the hazards associated with the Perth Airport nuisance alerts in its safety case assessment and reporting determination documentation that stated that controllers were ‘becoming desensitised to the alarms’. While Airservices Australia had implemented a fix for the nuisance alerts and advised the controllers at Perth Terminal Control Unit, the approach controller reported they were unaware that the issue had been resolved.  

Expectation strongly influences where an individual will search for information, what information they will search for, and how they perceive information (Wickens and McCarley 2008; and Wickens and others 2013). People are more likely to perceive information that is expected, and less likely to perceive information that is unexpected. In this instance, expectation bias resulting from the approach controller’s previous exposure to the nuisance alerts likely influenced their perception of the initial MSAW information. This led them to conclude that the aircraft was likely above the minimum safe altitude despite the information presented to them.

The approach controller recalled observing a second MSAW and, in response, they issued a safety alert to the flight crew of VH‑OHE. However, this occurred 35 seconds after the initial MSAW was first triggered. This delay was a missed opportunity to alert the flight crew of the detected terrain proximity at a time when the aircraft was still descending. 

Flight director mode selector illumination fault

During the incident flight, the captain’s and FO’s flight director mode selector panels had a fault that resulted in the intermittent illumination of mode selector buttons. A similar defect was also observed on 3 mode selector buttons on the FO side during a flight in VH‑OHE that took place on 7 August 2025. Following that flight, the defect was not recorded in the aircraft flight log or maintenance release, as required by the operator’s procedures and the Civil Aviation Regulations 1988, and no related rectification maintenance had occurred. 

Cockpit camera  

The information sourced from the installed cockpit video camera, which provided an accurate record of the events as they unfolded from within the cockpit, was invaluable to the incident investigation. Analysis of the video recording provided an insight into flight crew communications and actions, environmental factors, aircraft systems status, and instrument indications.

Findings

From the evidence available, the following findings are made with respect to the descent below minimum safe altitude, ground proximity alert and unstable approach involving Cessna 560, VH‑OHE, 23 km north‑east of Perth Airport, Western Australia, on 13 August 2025.

Contributing factors

• The flight crew incorrectly set the approach decision altitude in the altitude selector before the autopilot levelled the aircraft at the assigned altitude. The incorrect altitude setting, in combination with the selected flight director modes, resulted in the autopilot not capturing the glideslope and continuing the descent towards terrain.
• The flight crew became preoccupied with the flight director mode selection, and their attention was diverted from monitoring the aircraft state. During this time the aircraft descended below the minimum sector altitude, the aircraft entered an uncommanded left bank, and the crew did not appropriately respond to the enhanced ground proximity warning system alerts. 
• The captain took control of the aircraft and commenced a recovery climb. During this time the aircraft exceeded the maximum angle of bank, which resulted in an undesired aircraft state and increased the risk of a loss of control at a low altitude.
• The aircraft’s airspeed exceeded the operator’s stabilised approach criteria during the late stage of the approach. However, rather than conducting the required go-around, the flight crew elected to continue the approach.
• The air traffic controller issued a safety alert to the flight crew 35 seconds after the controller received a minimum safe altitude warning about the aircraft. This delay was a missed opportunity to alert the flight crew of the detected terrain proximity at a time when the aircraft was still descending. The delay in communicating the information to the flight crew was likely influenced by the controller’s exposure to previous nuisance warnings within the same area.
• The flight director mode selector panels had a fault that resulted in the intermittent illumination of mode selector buttons. This likely degraded the crew's ability to identify and select the desired mode. This defect was present during a flight that took place several days prior, but it had not been entered into the maintenance release, and no maintenance action had been undertaken.

Other findings

• The information sourced from the installed cockpit video camera was invaluable to the incident investigation. Analysis of the video recording provided an insight into crew communications and actions, aircraft systems status, and instrument indications.

Safety action 

Safety action by the operator of VH-OHE

The operator of VH-OHE advised the ATSB that all company pilots were provided with additional ground-based training relating to the use of aircraft automation and crew resource management. The organisation also communicated to pilots the importance of reporting incidents and defects in accordance with the Transport Safety Investigation Act 2003 and established operator procedures.

Sources and submissions

Sources of information

The sources of information during the investigation included:

• flight crew of VH-OHE
• the operator
• aircraft manufacturer
• Civil Aviation Safety Authority
• Airservices Australia
• Bureau of Meteorology
• recorded data from the cockpit video camera fitted to the aircraft
• recorded data from the flight data recorder fitted to the aircraft.

References

Civil Aviation Authority. (2023). Flight-crew human factors handbook. Retrieved from https://www.caa.co.uk/publication/download/14984

Civil Aviation Safety Authority. (2025). Safety behaviours: human factors for pilots 4th edition. Retrieved from https://www.casa.gov.au/sites/default/files/2021-06/safety-behaviours-human-factor-for-pilots-8-threat-error-management.pdf

Federal Aviation Administration. (2008). Nuisance Alerts in Operational ATC Environments: Classification and Frequencies. Retrieved from https://hf.tc.faa.gov/publications/2008-nuisance-alerts-in-operational-atc-environments/full_text.pdf

Federal Aviation Administration. (2021). Airplane Flying Handbook. Retrieved from https://www.faa.gov/sites/faa.gov/files/regulations_policies/handbooks_manuals/aviation/airplane_handbook/00_afh_full.pdf

International Civil Aviation Organization. (2002). Line Operations Safety Audit (LOSA) Doc 9803. International Civil Aviation Organization.

International Civil Aviation Organization. (2025). State of Global Aviation Safety. Retrieved from https://www.icao.int/sites/default/files/sp-files/safety/Documents/ICAO_SR_2025.pdf

Wickens CD, Hollands JG, Banbury S & Parasuraman R. (2013). Engineering psychology and human performance, 4th edition, Pearson Boston, MA. 

Wickens, CD & McCarley, JS. (2008). Applied attention theory, CRC Press, Boca Raton, FL.

Submissions

Under section 26 of the Transport Safety Investigation Act 2003, the ATSB may provide a draft report, on a confidential basis, to any person whom the ATSB considers appropriate. That section allows a person receiving a draft report to make submissions to the ATSB about the draft report. 

A draft of this report was provided to the following directly involved parties:

• flight crew of VH-OHE
• the operator
• Civil Aviation Safety Authority
• Airservices Australia
• Bureau of Meteorology.

Submissions were received from:

• the captain of VH-OHE
• the operator. 

Purpose of safety investigations The objective of an ATSB safety investigation is to improve transport safety through: identifying safety issues for action by organisations with the responsibility for managing that safety risk influencing safety action through engaging with stakeholders, communicating findings, and fostering awareness of safety issues and concerns.  In accordance with the TSI Act, the ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action, and cannot apportion blame, assist in determining liability, or, as a general rule, assist in court proceedings.  About ATSB reports ATSB safety investigation reports are developed in accordance with ATSB procedures and guidelines, and with regard to applicable international standards and instruments. Reports must include factual material of sufficient weight to support the investigation’s analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. An explanation of terminology used in ATSB investigation reports is available here.   Publishing information Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2026 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Commonwealth Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this report is licensed under a Creative Commons Attribution 4.0 International licence. The  CC BY 4.0 licence enables you to distribute, remix, adapt, and build upon our material in any medium or format, so long as attribution is given to the Australian Transport Safety Bureau.  Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

1. ^    Instrument flight rules (IFR) are a set of regulations that permit the pilot to operate an aircraft in instrument meteorological conditions (IMC), which have much lower weather minimums than visual flight rules (VFR).
2. ^    Pilot flying (PF) and pilot monitoring (PM): procedurally assigned roles with specifically assigned duties at specific stages of a flight. The PF does most of the flying, except in defined circumstances, such as planning for descent, approach and landing. The PM carries out support duties and monitors the PF’s actions and the aircraft’s flight path.
3. ^    Local time in Perth was Western Standard Time (WST), which is Coordinated Universal Time (UTC) +8 hours. Times in this report are WST unless otherwise noted.
4. ^    Flight level: at altitudes above 10,000 ft in Australia, an aircraft’s height above mean sea level is referred to as a flight level (FL). FL 410 equates to 41,000 ft.
5. ^    Standard instrument arrival (STAR): a designated IFR arrival route linking a significant point, normally on an ATS route, with a point from which a published instrument approach procedure can be commenced.
6. ^    Runway number: the number represents the magnetic direction of the runway centreline and is expressed in 10 degree increments of azimuth.
7. ^    Localiser: part of the instrument landing system and provides lateral position necessary to align with the runway centreline.
8. ^    Instrument landing system (ILS): a precision runway approach aid based on two radio beams which together provide pilots with both vertical and horizontal guidance during an approach to land.
9. ^    Altitude alert: an aural alert (tone) and/or a visual alert when an aircraft on climb/descent approaches the designated altitude and when deviating from that altitude during cruise. Aircraft conducting IFR operations in controlled airspace were required to have either an assigned altitude indicator or an altitude alerting system.
10. ^   Although the captain, who was the PM, announced ‘1,000 [ft] to go,’ Avcair Aircraft Operating Procedures required the PM to announce ‘Alert to 3,000 ft’ in response to the activation of the audible altitude alert. 
11. ^   Decision altitude (DA): a specified altitude (referenced to mean sea level) in a 3D instrument approach operation at which a missed approach must be initiated if the required visual reference to continue the approach has not been established. 
12. ^   Electronic attitude director indicator (EADI): an EADI is a part of the primary flight display (PFD) in glass cockpit aircraft. The EADI displays pitch and roll, and it also includes a slip/skid indicator for turn coordination. The EADI also provides the display for flight director symbols, which give the pilot a visual indication of the aircraft attitude required to follow flight director guidance.
13. ^   Initial approach fix (IAF): the commencement of an instrument approach.
14. ^   ‘Established’ means being within half full scale deflection for the ILS.
15. ^   ‘LAND Position’ on the 560 is 35° of flap extension. 
16. ^   Flight management system (FMS): an avionics system. It supports navigation, flight plans, positioning, guidance, and other flight-related tasks.
17. ^   The function of an MFD system is to provide the pilot access to various data, or combinations of data, used to fly the aircraft, to navigate, to communicate, and/or to manage aircraft systems.
18. ^   Maintenance release: a document that provides registered operators and pilots with evidence that an aircraft has had the required maintenance carried out, provides details of any defects that may require rectification before a flight can commence, and also provides essential information about any scheduled maintenance actions that may become due before completion of a planned flight.
19. ^   HAMC: head of aircraft airworthiness and maintenance control.
20. ^   MC: maintenance controller.
21. ^   VREF: The landing approach airspeed at the 50-foot point with flaps in landing position and landing gear extended (1.3 times the stalling speed in the landing configuration).
22. ^   HAT: Height above touchdown.
23. ^   The maximum operating airspeed for flap 35° (landing position) in the 560 was 173 kt.
24. ^   Few: a meteorological term used in aviation to describe cloud coverage that occupies 1 to 2 oktas (eighths) of the sky.
25. ^   Broken: a meteorological term used in aviation to describe cloud coverage that occupies 5 to 7 oktas (eighths) of the sky.
26. ^   An ILS approach procedure which provides for an approach to a decision height not lower than 60 m (200 ft) and a visibility not less than 800 m (2,400 ft) or a runway visual range not less than 550 m (1,800 ft).
27. ^   Very high frequency omnidirectional range station (VOR): transmitters that support non-precision (lateral guidance only) approach and en route navigation.
28. ^   The Manual of Air Traffic Services is a joint document of Defence and Airservices and is based on the rules published in Civil Aviation Safety Regulations Part 172 – Manual of Standards and International Civil Aviation Organization standards and recommended practices, combined with rules specified by Airservices and Defence.

Investigation summary

What happened

On 25 June 2025, the flight crew of a Jetstar Airways Airbus A321-251, VH-OYF, were conducting a scheduled passenger transport flight, JQ38, from Denpasar International Airport, Bali, Indonesia, to Sydney, New South Wales. The first officer was the pilot flying and the captain was the pilot monitoring.

During the landing at Sydney Airport, the aircraft floated for a prolonged period along the runway, was subject to a right crosswind and drifted left of the runway centreline. The captain responded by commanding a go-round which the first officer executed. 

The crew proceeded to continue with the published missed approach procedure and subsequently landed without further incident. 

What the ATSB found

The ATSB found that after the first officer initiated the flare manoeuvre, their control inputs resulted in a lateral deviation from the runway centreline when the aircraft floated for a prolonged period in crosswind conditions. 

After the captain commanded a go-around, they inadvertently manipulated their sidestick control, which resulted in a brief period where simultaneous control inputs occurred. The crew were alerted by a ‘dual input’ generated voice message and the captain took control. There was a moment of preoccupation which resulted in the first stage of flap being retracted out of sequence, however, there were no associated flight envelope exceedances or negative effects on aircraft performance. 

Safety message

Sound go-around decision-making is an effective defence against the hazards associated with low-level manoeuvring during the landing phase of flight, such as lateral runway excursions. If adequate safety margins cannot be maintained during an approach and landing, the correct and expected response is to go around.

Being go-around minded improves crew readiness and supports timely, coordinated actions during a period of high workload. This should involve crew members reviewing potential go‑around scenarios, procedures and responses prior to conducting an approach. 

When flight crews are faced with the unexpected need to execute a go-around even at the final stages of landing, effective crew resource management, with clear communication between flight crew, is essential. This promotes effective teamwork when responding to disruptions and increased workload under stress, ensuring that the aircraft remains on a safe flight path and is correctly configured for the relevant phase of flight.

The investigation

The occurrence

On the evening of 25 June 2025, a Jetstar Airways Pty Limited Airbus A321-251 registered VH‑OYF was operating on a schedule passenger transport Jetstar flight, JQ38, from Denpasar International Airport, Bali, Indonesia, to Sydney, New South Wales. The flight was scheduled to arrive at Sydney Airport the following morning at 0630 AEST.[1] The operating crew included the captain, first officer, 6 cabin crew and 234 passengers. For the flight to Sydney, the first officer was the pilot flying (PF) and the captain was the pilot monitoring (PM).[2]   

After departing Denpasar, the aircraft climbed to flight level (FL) 330[3] and later descended to FL310 after reaching Australian airspace due to turbulence en route. Due to the turbulence en route, the captain elected not to take any controlled rest on the nearly 6‑hour flight, while the first officer stated they would not usually take controlled rest in flight. 

Prior to descent, the flight crew briefed for the arrival at Sydney, recalling that the turbulent conditions and the crosswind for the approach and landing were the main considerations. 

At 0554, the flight crew commenced their descent to the west-south-west of Sydney Airport and was cleared for the approach for runway 16R[4] which was conducted in day visual meteorological conditions[5] using the autopilot. The flight crew recalled there was a 30 kt crosswind down to about 500 ft above mean sea level (AMSL) and the approach up to that point was ‘pretty normal.’ Air traffic control (ATC) advised the crew to expect an 8 kt right crosswind for landing and the first officer chose to land in the flap 3 configuration,[6] which was consistent with guidance for landing in ‘rough’ conditions. (The first officer was procedurally restricted to a maximum crosswind landing component of 20 kt).

The aircraft reached 500 ft at 0621:14 and the captain called ‘stable’ (see Stabilised approach criteria). The first officer disengaged the autopilot 5 seconds later as the aircraft approached 400 ft and recalled encountering turbulence which placed the aircraft ‘a little higher’ on the approach. At 0621:45 at 90 ft, the first officer pitched forward, which they observed resulted in a 900 ft per minute rate of descent. 

At 0621:51, the first officer initiated the flare at 50 ft and reduced the thrust levers to idle at around the final approach speed (VAPP)[7] of 150 kt, which included a wind correction of 5 kt. At this point the first officer recalled they ‘over flared’. The captain also observed that the first officer applied the flare technique that was consistent with the technique for landing in the flap full configuration. The aircraft subsequently floated for a prolonged period along the runway after the first officer’s flare manoeuvre.

During the prolonged float, the aircraft was subjected to the crosswind conditions for a greater length of time. After observing the centreline deviation, the captain commanded a go-around approximately 600 m past the runway threshold, just prior to touchdown. The captain recalled they were ‘startled by the need to go around’ as the approach seemed ‘benign’ aside from the crosswind. They also reported a sudden stress response at this time as they had to rapidly transition from landing to commencing the go-around.

In response to the captain’s command, the first officer set take-off/go-around thrust at 0621:59 (Figure 1), which initiated the published missed approach procedure for the 16R GBAS landing system (GLS)[8] approach in the aircraft flight management system. The first officer also referenced their primary flight display (PFD) to command a target pitch attitude of 15° nose up.   

At this point, the captain recalled they instinctively applied control inputs via their sidestick while the aircraft was just above the runway, and the crew were alerted to this by the aircraft’s ‘dual input’ voice message (see Sidestick priority logic). 

The captain then engaged their sidestick pushbutton, and the first officer recalled hearing the ‘priority left’ voice message and the captain announce, ‘I have control.’ The captain subsequently took control of the thrust levers and the first officer relinquished control and became PM after the aircraft achieved a positive rate of climb. It was the role of the PM to retract the flap ‘one step’ at this point (see Go-around procedure). 

Figure 1: Overview of go-around 

Source: Google Earth, annotated by the ATSB

The captain announced the active flight modes on their PFD, which prompted the first officer to call ‘positive climb.’ The captain subsequently instructed the first officer to retract the landing gear, which was accomplished 42 ft above the runway at 0622:20. 

At this time, the captain looked up to the flight control unit located on the cockpit glareshield to engage the autopilot. After this was actioned, they looked back to their PFD and was ‘startled’ when they noticed that the aircraft suddenly banked right and responded by disengaging the autopilot at 0622:22. They subsequently realised that the aircraft flight director was providing commands for the published missed approach procedure and subsequently re-engaged the autopilot at 0622:29. 

The captain then requested flap 1, but the first officer noticed they were still configured with Flap 3 and retracted the flap by one step and announced, ‘flap 2.’ This occurred at 0622:32 when the airspeed reached 174 kt, which was below the maximum flap 3 speed of 195 kt.

They continued to follow the missed approach procedure, and the first officer advised ATC they were going around. The crew were given instructions to track for a right downwind for runway 16R at 4,000 ft. The captain recalled conducting a welfare check on the first officer, briefed the cabin manager via the interphone and made an announcement to the passengers through the public address system. 

The captain elected to remain as PF for the remainder of the flight, with the first officer acting as PM. The crew then conducted a second GLS approach for runway 16R, landing at 0638 without further incident.

Context

Flight crew information

The captain held an Air Transport Pilot Licence (Aeroplane), class 1 aviation medical certificate, and had accrued 5,921 hours total flying time, 1,480 of which were in the Airbus A320 and A321 aircraft types.

The first officer held a Commercial Pilot Licence (Aeroplane), class 1 aviation medical certificate, and had 2,212 hours total flying time, 551 of which were on the Airbus A320 and A321 aircraft types.

Fatigue

The captain reported that they felt 'moderately tired' during the go-around, likely due to the back-of-the clock[9] flight, which departed Denpasar at 0057 local time in Sydney. They also stated there was limited opportunity for controlled rest during the flight and their nap prior to the flight was disrupted due to noise at the hotel. The first officer reported feeling 'ok, somewhat fresh.’  

The flight crew also reported they had an adequate rest opportunity the evening prior to the flight and obtained around 6 hours sleep in the previous 24 hours and around 13‍–‍14 ‍hours in the previous 48 hours. Their sleep during the rest opportunity was reported to be good quality and the conditions at the hotel where they spent the night were suitable and therefore conducive to obtaining restful sleep. Biomathematical modelling[10] of the flight crew’s roster for the 2 weeks leading up to the flight indicated a low likelihood of fatigue.

The ATSB considered that fatigue was unlikely to have affected the flight crew’s performance at the time of the occurrence.

Aircraft information

General

The Airbus A321-251NX is a modern, fly-by-wire aircraft, powered by 2 CFM International LEAP-1A32 turbofan engines and had seating for 232 passengers in a single-class layout. 

All the flight controls are electronically actuated with the pilots using sidesticks to fly the aircraft in pitch and roll during manual flight. The 2 sidestick controllers are not coupled mechanically, and they send separate sets of signals to the flight control computers. 

Sidestick priority logic

Jetstar Airways A320-A321 Flight crew operating manual (FCOM) contains the following description of the aircraft sidestick priority logic: 

At all times, only one flight crewmember should fly the aircraft. However, if both flight crewmembers use their sidesticks simultaneously, their orders are algebraically added.

The flight control laws limit the combined order to the equivalent of the full deflection of one sidestick.

In this case the two green SIDE STICK PRIORITY lights on the glareshield come on and "DUAL INPUT" voice message is activated.

 A flight crewmember can deactivate the other sidestick and take full control, by pressing and keeping pressed the sidestick pb (Figure 2).

A “PRIORITY LEFT” or “PRIORITY RIGHT” audio voice message is given each time priority is taken.

Figure 2: Airbus A320/A321 captain's side sidestick and sidestick pushbutton

Source: Operator, annotated by the ATSB

Post-flight maintenance

The operator reported that there were no corrective maintenance actions that were required to be carried out in relation to the occurrence. The aircraft subsequently operated a scheduled passenger service the following day.

Meteorological information

The pre‑flight briefing package provided to the flight crew from the operator’s flight dispatcher included the aerodrome forecast[11] for Sydney Airport. The forecasted weather conditions for the scheduled time of arrival 0630 local time on 26 June indicated:

• wind direction of 240° at 15 kt with gusts up to 25 kt
• CAVOK[12]
• moderate turbulence[13] below 5,000 ft.

One-minute weather data for Sydney Airport from the Bureau of Meteorology indicated a wind direction of 255° at 17 kt with gusts up 20 kt at the time of the occurrence.

Airport information

Runway 16R at Sydney Airport is oriented on a magnetic heading of 155° and has a declared length of 3,962 metres with a width of 45 metres. A precision approach path indicator system is installed and set to 3° with a threshold crossing height of 64 ft. 

For daytime operations, the runway centreline, aiming point and touchdown zone markings provide visual references to assist pilots with approach and landing (Figure 3).

Figure 3: Sydney Airport runway 16R markings

Source: Google Earth, annotated by the ATSB

Recorded information

The aircraft’s quick access recorder data which captured the incident approach indicated that, as the aircraft descended below 1,000 ft, it maintained an appropriate speed and flightpath with no sustained exceedances of the stable approach criteria throughout the approach. 

At 0621:59, the recorded data captured the captain’s control inputs commencing concurrently with the initiation of the go-around, while the first officer was actively manipulating their sidestick control. Simultaneous control inputs lasted for a duration of 6 seconds (Figure 4), while the aircraft’s pitch attitude remained below the aircraft’s pitch limit of 11.5° until the aircraft had climbed through about 50 ft. 

The recorded data further indicated that the wind direction and speed varied following the flare manoeuvre, however the crosswind component remained well below the first officer’s operational limitation. The wind direction and speed was 315° at 13 kt with a crosswind component of 5 kt when the go-around was initiated.

Figure 4: Graphical representation of the recorded quick access data

Source: Quick access recorder from VH-OYF, annotated by the ATSB

Following the initiation of the go-around, the landing gear was retracted at 06:22:20 and 12 seconds later, the flap was retracted to the flap 2 configuration[14] at 174 kt.

Operational information

Stabilised approach criteria 

Jetstar Airways A320-A321 Flight crew operating manual (FCOM) defined a stabilised approach criteria as being established on the correct lateral and vertical flight path by 1,000 ft height above airport (HAA), configured for landing, and within the stated tolerances with the required checklists completed by 500 ft HAA. The FCOM also stated that if these criteria could not be met, or if the approach became unstable below 1,000 ft HAA, a missed approach was required. 

The crew reported the approach was stabilised against these criteria, which was consistent with the available recorded data.

Touchdown zone 

The FCOM provided the following operational information regarding the touchdown zone: 

The touchdown zone commences at 300 m (1000 ft) beyond the threshold and will not normally extend further than 600 m (2000 ft) beyond the threshold.

It is a requirement that the touchdown is planned to occur within the touchdown zone. Should it become apparent that the aircraft will touch down further than 600 m (2,000 ft) beyond the threshold, and the PIC believes that the landing is safe to continue, the PF must apply maximum reverse thrust and sufficient braking to ensure the aircraft stops within the landing distance available. If the PIC decides that a go-around is required, they will without delay, call “Go-Around”. In all cases this must be completed before the PF initiates reverse thrust.

The captain stated that runway 16R in Sydney was long enough to stop the aircraft on the runway if they had continued with the landing during the occurrence. This would have involved requesting maximum reverse and manual braking as necessary after the aircraft touched down. 

The FCOM did not specifically reference runway centreline tracking during a visual approach, however the captain stated that it was their personal expectation that a deviation from the runway centreline would lead them to calling for a go-around. 

Transfer of control  

The operator described procedures for transfer of control within the FCOM as follows:

The pilot relinquishing control of the aircraft shall say “You have control”. The pilot assuming control shall ensure that they have clear and unobstructed access to the flight controls and, when ready, say “I have control”. Only then is the pilot relinquishing control permitted to remove their hands and feet from the flight controls.

In critical phases of flight the PIC must be alert and positioned such that they can assume immediate control of the aircraft.

Following the occurrence, the captain stated the preferable method to conduct a go‑around at low level would have been to announce ‘I have control’ and initiate the go‑around themselves. They stated that their primary consideration when conducting a go‑around at low level was to avoid the risk of tail strike. 

Go-around procedure 

The FCOM defined the go-around procedure for the A320/A321, which specified the task sequence, memory-based crew actions and applicable guidance relating to techniques and navigation (Figure 5).

Figure 5: Jetstar Airway A320/A321 go‑around procedure below acceleration altitude

Source: Operator, annotated by the ATSB

Following the occurrence, the captain stated that although they could have taken over and landed, they believed that going around was considered the safest option. The first officer also stated, at about that time, that they were in the mindset of preparing to initiate a go-around themselves. 

Related occurrences

The following ATSB investigation highlights the importance of pilots maintaining their readiness for a go-around on every approach as it is typically a period of high workload requiring effective crew coordination. 

ATSB Investigation

On the morning of 18 May 2018, an Airbus A320 aircraft, registered VH-VQK, was being operated on a regular public transport flight by Jetstar Airways. The flight departed from Sydney for Ballina/Byron Gateway Airport, New South Wales.

The flight crew conducted a go-around on the first approach at Ballina because the aircraft’s flight path did not meet the operator’s stabilised approach criteria. On the second approach, at about 700 ft radio altitude, a master warning was triggered because the landing gear had not been selected DOWN. The flight crew conducted a second go‑around and landed without further incident on the third approach.

The flight crew did not follow the operator’s standard procedures during the first go‑around and subsequent visual circuit at 1,500 ft. In particular, the flaps remained at flaps 3 rather than flaps 1 during the visual circuit. This created a series of distractions leading to a non‑standard aircraft configuration for a visual circuit. Limited use of available aircraft automation added to the flight crew’s workload.

Safety analysis

During the approach to Sydney airport, with the first officer acting as the pilot flying (PF), the flight crew reported experiencing a crosswind of up to 30 kt until descending through about 500 ft above mean sea level. The crew were advised by air traffic control to expect a right crosswind component of 8 kt for landing, which was within the first officer’s operational crosswind limit of 20 kt. The captain confirmed the approach was ‘stable’ at 500 ft and the first officer continued the approach as PF.

At 50 ft, the first officer initiated the flare manoeuvre prior to landing. They recalled they ‘over flared,’ and the aircraft subsequently floated for an extended period along the runway. During this time, the first officer’s control inputs did not counteract the effect of the crosswind, and the aircraft drifted left of the centreline. After observing the lateral deviation from the centreline, the captain commanded the first officer to conduct a go‑around. 

This occurred just prior to the aircraft touching down when the flight crew would normally be focused on landing. The flight crew did not expect a go-around at the time and had to rapidly shift their focus to conducting the missed approach procedure. The captain recalled being ‘startled’ by the unexpected need to discontinue the landing, however they were more likely experiencing ‘surprise.’ Surprise is a cognitive-emotional response to something unexpected, which results from a mismatch between one’s mental expectations and perceptions (Rivera, Talone, Boesser, Jentsch, & Yeh, 2014). But their decision was consistent with the expectation that an approach be discontinued if the aircraft departed from the correct lateral flight path.

The unexpected change from landing to conducting a go-around close to the ground also resulted in the captain experiencing a sudden stress response at this time. When experiencing acute stress, people can respond quickly to a situation, but without conscious decision‑making (Wickens, Helton, Hollands, & Banbury, 2022). After the go‑around was commanded, there was a rapid increase in pitch attitude, engine thrust and airspeed, and in response the captain instinctively and inadvertently manipulated their sidestick while the first officer was flying, resulting in a dual-input alert. 

The captain reported they only realised they had manipulated their sidestick when they heard the dual input alert. Their primary consideration during the go-around was to avoid an excessive rotation rate to avoid a tail strike, which did not occur. Additionally, operator procedures directed captains to be alert and be positioned to ‘assume immediate control of the aircraft’ during critical phases of flight. 

Following the dual input alert, the captain took full control by engaging their sidestick push‑button and announced ‘I have control’, and the first officer assumed the role of pilot monitoring. A consequence of the control handover during the initial stages of the go‑around was the momentary interruption of sequential crew actions during the go‑around procedures. Interruptions typically disrupt the chain of procedure execution so abruptly that pilots turn immediately to the source of the interruption without noting the point where the procedure was suspended (Loukopoulos, Dismukes, & Barshi, 2009). 

Additionally, there was a further disruption (rapid task switching) associated with the first officer and captain exchanging pilot flying and pilot monitoring roles. As a result, some of the procedural items were completed out of sequence (flap 3 retraction occurred after gear retraction). 

Pilots are highly vulnerable to errors of omission when they must attend to multiple tasks. If one task becomes demanding, their attention is absorbed by these tasks demands and they can forget to switch their attention to other tasks (Loukopoulos, Dismukes, & Barshi, 2009). Although the flap retraction occurred out of sequence during the go-around, there were no associated flight envelope exceedances or negative effects on aircraft performance.  

Findings

From the evidence available, the following findings are made with respect to the control issues during landing and go-around involving Airbus A321, VH-OYF, at Sydney Airport, New South Wales, on 26 June 2025.

Contributing factors

• During the landing after crossing the threshold, the first officer’s control inputs resulted in a lateral deviation from the runway centreline during a prolonged float.
• After calling for a go-around, the captain inadvertently manipulated their sidestick while the first officer was the pilot flying, which resulted in a simultaneous control input and the go-around procedure being completed out of sequence.

Sources and submissions

Sources of information

The sources of information during the investigation included:

• Jetstar Airways Pty Limited
• Bureau of Meteorology
• the flight crew
• recorded data from the quick access recorder from VH-OYF.

References

Loukopoulos, L., Dismukes, R., & Barshi, I. (2009). The perils of multitasking. AeroSafety World, 4(8), 18-23.

Rivera, J., Talone, A., Boesser, C., Jentsch, F., & Yeh, M. (2014). Startle and surprise on the flight deck: Similarities, differences, and prevalence. In Proceedings of the human factors and ergonomics society annual meeting (Vol. 58, No. 1, pp. 1047-1051). Sage CA: Los Angeles, CA: SAGE Publications.

Wickens, C. D., Helton, W. S., Hollands, J. G., & Banbury, S. (2022). Engineering psychology and human performance, 5th edn. Routledge, doi: 10.4324/9781003177616.

Submissions

Under section 26 of the Transport Safety Investigation Act 2003, the ATSB may provide a draft report, on a confidential basis, to any person whom the ATSB considers appropriate. That section allows a person receiving a draft report to make submissions to the ATSB about the draft report. 

A draft of this report was provided to the following directly involved parties:

• Civil Aviation Safety Authority
• the flight crew
• Jetstar Airways Pty Limited
• Bureau of Meteorology.

Submissions were received from:

• the flight crew
• Jetstar Airways Pty Limited.

The submissions were reviewed and, where considered appropriate, the text of the report was amended accordingly.

Purpose of safety investigations The objective of a safety investigation is to enhance transport safety. This is done through:  identifying safety issues and facilitating safety action to address those issues providing information about occurrences and their associated safety factors to facilitate learning within the transport industry. It is not a function of the ATSB to apportion blame or provide a means for determining liability. At the same time, an investigation report must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. The ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action. About ATSB reports ATSB investigation reports are organised with regard to international standards or instruments, as applicable, and with ATSB procedures and guidelines. Reports must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. An explanation of terminology used in ATSB investigation reports is available here. This includes terms such as occurrence, contributing factor, other factor that increased risk, and safety issue. Publishing information Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2026 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Commonwealth Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this report is licensed under a Creative Commons Attribution 4.0 International licence. The CC BY 4.0 licence enables you to distribute, remix, adapt, and build upon our material in any medium or format, so long as attribution is given to the Australian Transport Safety Bureau.  Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

Investigation summary

What happened

On 15 April 2025, an Embraer ERJ 190-100, registered VH-UZD, was conducting a passenger transport flight from Sydney, New South Wales, to Launceston, Tasmania. After commencing approach to Launceston, the flight crew received multiple caution messages including a SLAT FAIL caution. The flight crew discontinued their approach and after completing the relevant checklists elected to divert to Melbourne, Victoria, as it was the longest available runway in the region. The remainder of the flight was uneventful, and the aircraft landed safely.

Post-flight troubleshooting determined that a torque tube in the left wing slat drive system had disconnected as it had been incorrectly assembled when it was last refitted.

What the ATSB found

The ATSB identified a similar occurrence with another of the operator’s Embraer ERJ 190‑100 aircraft, VH-UYB, where a torque tube in the left wing flap drive system had disconnected as it had been incorrectly assembled when it was last refitted.

The occurrences were similar in that the locking bolts that secured the torque tubes to their actuators had not been fitted correctly into the holes of the splined shafts, since the torque tubes had been incorrectly positioned during installation.

In both occurrences, those carrying out and certifying for the torque tube installations did not identify that they had been incorrectly assembled.

These errors occurred at different maintenance providers, and reportedly from January 2005–August 2011 in the worldwide fleet of Embraer 170, 175, and 190 aircraft (all sharing similar componentry), there have been 5 similar occurrences related to incorrect torque tube installation.

What has been done as a result

The operator, Alliance Airlines, issued a maintenance notice that detailed the flap torque tube disconnect affecting VH-UYB and the slat torque tube disconnect affecting VH-UZD. This notice reiterated the aircraft maintenance manual information for the correct installation of flap and slat torque tubes.

The maintenance organisation added an additional task card that is automatically issued when work is scheduled on the E190 slat system torque tubes that provides guidance in addition to the aircraft maintenance manual to mitigate the incorrect assembly of torque tubes on their splines. A similar additional task card was being developed for the E190 flap system torque tubes.

Safety message

Historical occurrence and technical information provide an opportunity to review known errors prior to commencing particular maintenance activities, thereby reducing the possibility of further errors occurring. When an error does occur, this information also provides a means to bolster the actions taken to prevent re-occurrences.

This information can be available from multiple sources including the manufacturer, national aviation authorities (such as CASA or the FAA), accident investigation authorities, and the safety management systems of operators and maintenance organisations.

The investigation

The occurrence

Previous maintenance

In November 2024, an Embraer ERJ 190-100 aircraft, registered VH-UZD and operated by Alliance Airlines, commenced a heavy maintenance[1] check by Rockhampton Aviation Maintenance in Rockhampton, Queensland. A team comprising 2 aircraft maintenance engineers (AMEs) was tasked with inspecting and lubricating the leading-edge slat drive system (see Embraer E190 slats and flaps). This involved removing, cleaning, lubricating, and refitting each slat torque tube in turn. A licensed aircraft maintenance engineer (LAME) briefed the AMEs on what was required.[2] The LAME was familiar with the task but was unaware of any historical issues with the task (see Maintenance requirements). The work was carried out in a new facility with good lighting. Access to the components was good, and a purpose-built platform allowed the work to be carried out with the relevant components at eye level.

Prior to commencing work, brakes internal to the power drive units (PDUs) (which drive the flap and slat torque tubes) were electrically released as required by the aircraft maintenance manual (AMM) procedure. The AMEs printed a copy of the relevant AMM procedure, and worked together on the torque tube driving the left-wing outboard actuator for slat number 4. The PDU brakes were also required to be released prior to installing the torque tubes, however, it could not be established whether this took place (the PDU brakes reapply when power is removed). After refitting the outboard actuator torque tube, a push-pull check was carried out to ensure it was locked in place, as required by the AMM. Unknown to the AMEs, when this torque tube was refitted, it had not been positioned far enough onto the actuator’s splined shaft for the locking bolt to secure it (Figure 1, lower right). The locking bolt was inadvertently installed beyond the end of the spline (shown in grey) rather than through the hole as required.

One AME then continued work on the left wing and the other moved to the right wing slat drive system to work alone. The remaining slat torque tubes were correctly fitted.

After this work was completed, the LAME inspected the installation of the torque tubes and their locking bolts, and a second LAME carried out an independent inspection[3] of the work. The heavy maintenance check was completed in March 2025, and the aircraft was returned to service. 

Figure 1: Aircraft maintenance manual torque tube installation illustration

Source: Embraer

Flight control event

On 15 April 2025, 50 flights after returning to service from heavy maintenance, the aircraft was being operated on a passenger transport flight from Sydney, New South Wales, to Launceston, Tasmania, by Alliance Airlines for QantasLink. After commencing approach to Launceston, the flight crew received multiple caution messages[4] on the aircraft’s engine indicating and crew alerting system (EICAS) including a SLAT FAIL caution. The flight crew discontinued the approach and requested clearance from air traffic control for vectors[5] so they could action the relevant quick reference handbook (QRH) checklists for the caution messages.

The flight crew completed the QRH checklist. As the slat failure would require landing with the slats and flaps up, the flight crew elected to divert to Melbourne Airport, Victoria, as it had the longest available runway in the region. The flight crew declared a PAN PAN[6] and commenced the diversion to Melbourne. After climbing to 19,000 ft the aircraft was flown to Melbourne at 220 kt as required by the QRH because of the slat failure. The aircraft landed at Melbourne without further incident.

Post-flight inspection

Post-flight inspection determined that the torque tube for the left wing slat number 4 outboard actuator had disconnected as the locking bolt fitted to the torque tube had not passed through the corresponding hole in the actuator’s splined shaft when it was last refitted (Figure 2).

Figure 2: VH-UZD left wing outboard actuator for slat number 4 and torque tube, shown disconnected after the occurrence flight

Source: Alliance Airlines, annotated by the ATSB

Context

Aircraft information

The Embraer ERJ 190-100 IGW (E190) is a narrow-body aircraft used for air transport operations and powered by 2 General Electric CF34-10E5 turbofan engines. VH-UZD was manufactured in Brazil in 2008 and registered in Australia on 31 January 2022.

Embraer E190 slats and flaps

The E190 is fitted with devices to increase the lift produced by its wings during take-off and landing. On the leading edges of the wings there are 8 slat panels and on the trailing edges of the wings there are 4 flap panels (Figure 3), where each set (slats/flaps) extends and retracts together.

Figure 3: Embraer E190 slats and flaps

Source: Embraer, annotated by the ATSB

Slat and flap extension and retraction is controlled from the cockpit by using the slat/flap control lever (SFCL). When the SFCL is moved from its 0 (up) position,[7] the flap and slat power drive units (PDUs) drive torque tubes which in turn drive actuators, transferring the rotary motion of the torque tubes to linear motion that extends the slats and flaps (Figure 4 and Figure 5).

Each PDU has 2 internal brakes that are engaged under spring force and released electrically, such that the brakes would re-engage when power is removed. There are 26 torque tubes in the slat drive system and 22 torque tubes in the flap drive system.

In the event of a slat or flap failure, redundant detection and protection systems prevent them operating in such a way that may compromise safety of flight.

Figure 4: Embraer E190 slat drive system

Source: Embraer, annotated by the ATSB

Figure 5: Embraer E190 flap drive system

Source: Embraer, annotated by the ATSB

Maintenance requirements

The slat and flap torque tubes are removed periodically for the actuator splines to be lubricated with grease. They may also need to be removed to replace associated components. A detailed visual inspection of the slat and flap drive system is also carried out periodically and includes a requirement to check that the torque tubes are correctly secured in place by their locking bolts. No detailed visual inspections of the slat system had been required between the heavy maintenance in November 2024 and the occurrence flight.

The procedure to remove and install the slat and flap torque tubes is detailed in the aircraft maintenance manual (AMM). As part of this procedure, the slat or flap PDU brakes are disengaged electrically to eliminate any residual torque in the system that may impede (through friction) the removal of the torque tubes. For the same reason, the brakes must also be disengaged for their installation.[8] Embraer advised the ATSB of the importance of removing residual torque for the installation.

Rockhampton Aviation Maintenance noted during its investigation into the occurrence that excessive amounts of grease on the actuator splines can produce hydraulic resistance to re-assembly of the torque tube and therefore no more than what is required to lubricate the splines should be applied. It could not be determined whether this occurred during the maintenance of VH-UZD. The installation procedures for torque tubes in the AMM requires the old grease to be removed, new grease to be applied, and any unwanted grease to be removed prior to assembly.

The torque tubes interface with other components via splined shafts and are secured by locking bolts in conjunction with castellated nuts and split pins to prevent their inadvertent disconnection. There are 24 locking bolts in the slat drive system and 18 locking bolts in the flap drive system, all with this configuration.

The AMM describes and illustrates a ‘push-pull’ check to determine the locking bolt has been correctly installed and had showed representative examples of correct and incorrect installation (Figure 1).

The torque tube locking bolts pass through holes close to the end of each actuator’s splined shaft. A correctly installed torque tube is visually apparent by less exposed splines (Figure 6). If a slat torque tube is incorrectly positioned[9] on a slat actuator the locking bolt will not capture the splined shaft and can lead to the torque tube disconnecting and slat failures.

Figure 6: Exemplar slat torque tube correctly fitted (upper image) and incorrectly fitted (lower image) to a slat actuator 

A slat actuator and torque tube were correctly and incorrectly assembled on a workbench to create these images. Source: The maintenance organisation, annotated by the ATSB

Actions taken to prevent installation errors

In 2010 the AMM was amended to include the previously mentioned illustration (Figure1) showing the correct and incorrect installation of slat and flap torque tubes along with the push-pull test. This revision also added the requirement to release the PDU brakes.

Embraer communicated these changes by publishing a service newsletter SNL 190‑27‑0050 noting reports of incorrect slat or flap torque tube installation, advising that the AMM had been revised to mitigate future occurrences, and provided an overview of the revisions. This information was also published in Embraer’s safety magazine[10] (available to operators of E190s) and was contained in a document[11] published by the National Civil Aviation Agency of Brazil.

In October 2017 Embraer published an update on the issue in a document[12] that reiterated the previous actions taken to mitigate these occurrences. This document noted that from January 2005–August 2011 in the worldwide fleet of Embraer ERJ170, 175, 190, and 195 aircraft[13] there were 483 reports of slat or flap system failures. Of these, 5 were occurrences related to incorrect torque tube installation. Additionally, the document stated that the subject of incorrect torque tube installation was presented to civil aviation authorities in Europe and the Americas. It was concluded that no additional actions were required, as there were a small number of exposed aircraft, and there had been no reported events since the AMM was revised in 2010, and the manufacturer considered the issue closed.

Related occurrences

Incorrect flap torque tube installation

In late 2024, an Embraer ERJ 190-100 aircraft, registered VH-UYB and operated by Alliance Airlines for QantasLink, commenced a heavy maintenance check at a facility in Singapore. The torque tube driving the left wing flap actuator number 2 (see Embraer E190 slats and flaps) was removed to carry out flap actuator torque limiter checks. When fitted, the torque tube had not been positioned far enough onto the actuator’s splined shaft for the locking bolt to secure it.

On 10 November 2024, 35 flights after returning to service from heavy maintenance, the aircraft departed for a passenger transport flight. After take-off, the flight crew received a FLAP FAIL caution on the EICAS as the flaps were retracting. The flight crew initiated a turnback and the aircraft landed safely.

Engineering personnel later found that the locking bolt for the left wing flap actuator number 2 torque tube had not passed through the corresponding hole in the actuator splined shaft when it was last refitted (Figure 7).

Figure 7: VH-UYB left wing flap actuator 2 and torque tube

Source: Alliance Airlines, annotated by the ATSB

Other flight control event involving VH-UZD

On 18 April 2025, VH-UZD was operating from Adelaide, South Australia, to Canberra, Australian Capital Territory. When flaps were selected down, the slats began to extend but the flaps did not deploy, and the crew received multiple failure warnings. The flight crew diverted to Melbourne. Post-flight troubleshooting determined that the flap power drive unit (PDU) torque limiter had tripped, which is a problem unrelated to the investigation occurrence or the recent heavy maintenance check.

Safety analysis

Incorrect fitment of actuator torque tubes

When the torque tube for the left wing slat number 4 outboard actuator was refitted to VH-UZD in November 2024, it had not been positioned far enough onto the actuator’s splined shaft for the locking bolt to secure it in place. After re-entering service and conducting 50 flights, the torque tube disengaged from the actuator, and the slat system failed. Protection systems ensured the safety of flight was minimally affected.

Similarly, when another E190, VH-UYB, was under heavy maintenance at a different facility at around the same time, the torque tube driving the left wing flap actuator number 2 was incorrectly assembled in that the locking bolt had not passed through the hole in the actuator’s splined shaft. The torque tube disengaged 35 flights after the aircraft re-entered service and the flap system failed. 

Non-detection of the error

The 2 AMEs who fitted the torque tube in VH-UZD did not identify that the torque tube had been incorrectly fitted. Further, the LAME checking this work and the second LAME carrying out the independent inspection of this work did not identify that it had been incorrectly assembled. The similar error affecting VH‑UYB also apparently remained undetected by those carrying out and certifying for the work.

As far as could be established, there were no physical or environmental factors that may have influenced the incorrect assembly of the torque tube. The work on VH-UZD was carried out in a new facility with good lighting, and access to the work area was good and could be carried out with the relevant components at eye level.

Ultimately, it is likely that not knowing the subtle difference in appearance of an incorrectly assembled slat torque tube (that is, as little as about 6.35 mm more of the actuator spline visible) contributed to the error not being detected by the 2 AMEs and the 2 LAMEs involved. Further, the remaining torque tubes in the slat drive system were correctly assembled, however their subtly different appearance did not trigger recognition that the original torque tube had been incorrectly assembled.

Available relevant information

Installation of the slat and flap drive system torque tubes is a simple task, but errors have occurred. Embraer noted that from January 2005–August 2011 in the worldwide fleet of Embraer 170, 175, 190 aircraft (all sharing similar componentry) there were 5 occurrences related to incorrect torque tube installation. The Embraer 190 has 24 locking bolts in the slat drive system and 18 in the flap drive system representing a total of 42 opportunities to incorrectly secure the torque tubes.

In 2010, Embraer made amendments to the aircraft maintenance manual to reduce the possibility of assembly errors. These were intended to remove any residual torque loads during removal and installation (by releasing the PDU brake), highlight the possibility of error with an illustration, and through the addition of the push-pull check, provide a means to detect an installation error.

These changes were communicated in multiple documents, such as a service newsletter, that were available to operators and maintainers of E190s. Review of such documents can assist in highlighting known issues and thereby prevent reoccurrence.

Findings

From the evidence available, the following findings are made with respect to the flight control event involving Embraer E190, VH-UZD, 29 km south-east of Launceston Airport, Tasmania, on 15 April 2025. 

Contributing factors

• During scheduled maintenance, the locking bolt for the left outboard slat torque tube was not passed through the hole in the actuator’s splined shaft as the torque tube had been incorrectly positioned. The aircraft was released from maintenance, and 50 flights later, the torque tube disconnected, causing the slat system to fail.
• Both licensed aircraft maintenance engineers inspecting the left outboard slat torque tube did not identify that it had been incorrectly assembled.

Safety actions

Safety action taken by Alliance Airlines

On 17 April 2025, Alliance Airlines issued a maintenance notice that detailed the flap torque tube disconnect affecting VH-UYB on 11 November 2024 and the slat torque tube disconnect affecting VH-UZD on 15 April 2025. This notice reiterated the aircraft maintenance manual information for the correct installation of flap and slat torque tubes.

Safety action taken by Rockhampton Aviation Maintenance

The maintenance organisation added an additional task card that is automatically issued when work is scheduled on the E190 slat system torque tubes. This task card provides guidance in addition to the aircraft maintenance manual to highlight the possibility of hydraulic lock caused by lubricant and the importance of releasing the PDU brake. Additionally, this task details a dimensional check to confirm the correct installation of torque tubes on their splined shafts. A similar additional task card was being developed for the E190 flap system torque tubes.

Sources and submissions

Sources of information

The sources of information during the investigation included:

• Alliance Airlines
• Centro de Investigação e Prevenção de Acidentes Aeronáuticos (Brazil)
• Civil Aviation Safety Authority
• Embraer
• Rockhampton Aviation Maintenance
• licenced aircraft maintenance engineer that made the final certification of the work
• both aircraft maintenance engineers.

Submissions

Under section 26 of the Transport Safety Investigation Act 2003, the ATSB may provide a draft report, on a confidential basis, to any person whom the ATSB considers appropriate. That section allows a person receiving a draft report to make submissions to the ATSB about the draft report. 

A draft of this report was provided to the following directly involved parties:

• Alliance Airlines
• Centro de Investigação e Prevenção de Acidentes Aeronáuticos (Brazil)
• Civil Aviation Safety Authority
• Embraer
• Rockhampton Aviation Maintenance
• licenced aircraft maintenance engineer that made the final certification of the work
• both aircraft maintenance engineers.

Submissions were received from:

• Embraer
• Rockhampton Aviation Maintenance.

The submissions were reviewed and, where considered appropriate, the text of the report was amended accordingly.

Purpose of safety investigations The objective of a safety investigation is to enhance transport safety. This is done through:  identifying safety issues and facilitating safety action to address those issues providing information about occurrences and their associated safety factors to facilitate learning within the transport industry. It is not a function of the ATSB to apportion blame or provide a means for determining liability. At the same time, an investigation report must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. The ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action. About ATSB reports ATSB investigation reports are organised with regard to international standards or instruments, as applicable, and with ATSB procedures and guidelines. Reports must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. An explanation of terminology used in ATSB investigation reports is available here. This includes terms such as occurrence, contributing factor, other factor that increased risk, and safety issue. Publishing information Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2026  Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Commonwealth Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this report is licensed under a Creative Commons Attribution 4.0 International licence. The CC BY 4.0 licence enables you to distribute, remix, adapt, and build upon our material in any medium or format, so long as attribution is given to the Australian Transport Safety Bureau.  Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

Investigation summary

What happened

At 1021 on 7 April 2025, a Bankstown Helicopters Robinson R44 helicopter, registered VH‑EWM (EWM), with a pilot and 2 passengers on board, departed from Bankstown Airport, New South Wales, for a local scenic flight around Sydney Harbour. Shortly after 1028, as EWM was entering the Parramatta River helicopter lane behind an EC120 helicopter, the occupants of EWM experienced a sudden onset of turbulence followed by an uncontrolled descent. 

In response, the pilot applied full collective, which resulted in a low rotor speed condition as the helicopter descended towards the water. The pilot was able to manoeuvre the helicopter and complete a forced landing on the river shoreline.

What the ATSB found

The ATSB found that it is likely that EWM entered the rotor wake from a preceding heavier EC120 helicopter, which resulted in the control difficulties, an uncontrolled descent, low rotor speed warning and the forced landing. 

What has been done as a result

Following review of the draft report, the Civil Aviation Safety Authority undertook proactive safety action to improve existing guidance about helicopter wake vortices in Advisory Circular 91-16. The updated advisory circular was released on 17 July 2025 and can be found at the link: AC 91-16 v1.2 - Wake turbulence.

Safety message

Flight tests have demonstrated that helicopter wake turbulence is comparatively larger and less predictable in its behaviour than for aeroplanes of the same weight. Helicopter rotor vortices can descend, remain level or climb, and the duration of their persistence can increase significantly in conducive weather conditions. The United States Helicopter Safety Team website recommends remaining 3 rotor disks clear of a hovering or taxiing helicopter and allowing 3 NM and/or 2 minutes for the rotor wake from a preceding helicopter to dissipate.

The investigation

The occurrence

At 1021 local time on 7 April 2025, a Bankstown Helicopters Robinson R44 Raven 1 helicopter, registered VH‑EWM (EWM), with a pilot and 2 passengers on board, departed from Bankstown Airport, New South Wales for a local scenic flight around Sydney Harbour. Bankstown Tower air traffic control (TWR) cleared EWM to depart via ‘Choppers West’, which was a standard procedure for helicopters departing to the north when runway 29 was active at Bankstown. 

The pilot reported that they climbed to about 1,000 ft above mean sea level.[1] The pilot’s plan was to join the Parramatta River on the west side of the Ryde Bridge and descend to 500 ft to follow the helicopter lane[2] along the south side of the river to Sydney Harbour (Figure 1).

Figure 1: Key locations

Source: Google Earth, annotated by the ATSB

About 1 minute after EWM departed, an Airbus EC120B helicopter departed Bankstown, also following the Choppers West departure. Shortly after the EC120 departed, TWR advised the EC120 pilot that there was ‘R44 traffic 1 NM ahead’, to which the EC120 pilot reported that they had the traffic sighted. Bankstown TWR then advised the pilot of EWM that they were not receiving their transponder data, which the pilot acknowledged. The pilot of EWM then turned their transponder off and on in an attempt to transmit transponder information, but no data was received from it throughout the incident flight.

As the 2 helicopters tracked north towards the Parramatta River, the EC120 flew to the west of EWM and passed it before reaching the river. The EC120 then turned right to join the Parramatta River helicopter lane, tracking towards the Ryde Bridge and Sydney Harbour, and passed over the Ryde Bridge at a recorded radar altitude of 600 ft. 

The pilot of EWM reported that they descended the helicopter to 500 ft as they approached the river. Just before the pilot turned EWM right to join the helicopter lane, another larger helicopter (the EC120) suddenly appeared in front of them (Figure 2). The pilot of EWM estimated the EC120 was about 500‍–‍600 ft (150‍–‍180 m) in front of them and about 100 ft above them. While there was no recorded altitude for EWM, primary radar data indicated that EWM entered the lane about 9 seconds behind the EC120.[3] Primary radar data for EWM was lost about 10 seconds later, just after 1028, indicating it had descended below radar coverage.

Figure 2: Primary radar return (left) and loss of primary radar return (right) for VH‑EWM

Source: Airservices Australia, annotated by the ATSB

The pilot of EWM made a radio broadcast that they were entering the helicopter lane as they crossed the Ryde Bridge behind the EC120. They then experienced what they described as very strong turbulence from a vertical motion in the atmosphere. A passenger later described it as ‘like heavy turbulence … rolling left and right’ followed by ‘diving towards the water’.

The pilot noted that the helicopter was descending through 400 ft and responded by raising the collective lever.[4] However, the helicopter continued descending towards the water as it tracked behind and below the EC120. A passenger recalled the pilot announced ‘brace for impact’ as the helicopter approached the water. The pilot applied full collective to avoid the water, which caused the rotor speed to decay sufficiently for the low rotor speed warning horn to activate. They also reported feeling that they could not escape what they believed to be the rotor wake from the EC120. The pilot then sighted a suitable forced landing area at Cabarita Park and, using the helicopter’s remaining airspeed and rotor speed, manoeuvred the helicopter to the shoreline for a landing.

Following the landing, the pilot rolled the engine throttle back to idle and proceeded through their after‑start checks and confirmed normal operations on the ground. The pilot then conducted a hover check and again confirmed normal operations. The pilot attempted radio contact with their operations base but received no reply. They then conducted a return flight to Bankstown without further incident. 

Context

Pilot information

The pilot held a commercial helicopter pilot licence, issued on 26 November 2024, with a single‑engine helicopter class rating and low‑level rating. The pilot held a class 1 aviation medical certificate with no restrictions and expiration date of 30 May 2025. The pilot had accumulated about 112 hours flying experience and the incident flight was the pilot’s first commercial flight.

Helicopter information

The incident helicopter, EWM, was a piston‑engine 2‑bladed Robinson Helicopter Company R44 Raven 1 with a maximum take‑off weight of 1,089 kg. The weight and balance data provided by the operator indicated it was within limits for the flight. 

The Airbus EC120B was a turbine-engine 3‑bladed helicopter with a maximum take‑off weight of 1,715 kg. Therefore, the EC120 was about 57% heavier than EWM at their respective maximum take‑off weights.

The maintenance release for EWM indicated the helicopter was operated by Bankstown Helicopters in the operational category of Part 133 Air Transport. The maintenance release current at the time of the incident was issued on 3 April 2025 at 4,349 hours total time in service with an expiry date of 3 April 2026 or 4,400 hours. A maintenance test flight was certified on the maintenance release as conducted on 3 April with ‘nil defects evident.’ 

After the incident, the operator’s maintenance organisation inspected the helicopter and found no defects. As the flight hours remaining on the helicopter were close to the next overhaul, the operator elected to remove the helicopter from service and have the maintenance organisation complete the overhaul. 

Meteorological information

The METAR[5] recordings for Bankstown Airport at 1000 and 1030 indicated that the wind was westerly at a speed of 9 kt at 1000 and 7 kt at 1030. No cloud was detected. These conditions were consistent with the Bankstown Airport forecast for 8 kt westerly winds. The pilot reported their assessment of the weather was 5 kt of variable wind and CAVOK[6] conditions, but when they encountered the turbulence over the Ryde Bridge it felt like 40 kt of wind.

Rotor wake turbulence

In 1996, the United States Federal Aviation Administration (FAA) produced a report on the subject of Flight test investigation of rotorcraft wake vortices in forward flight. They used a laser doppler velocimeter to measure the vortices and small probe aircraft to test the actual flying conditions. Smoke generation was used to visualise the wake vortices for the probe aircraft. Their investigation concluded that:

• The measured vortex circulation diminished with decreasing airspeed for helicopter airspeeds below 40 knots. At these lower speeds, the wake vortex structure begins to break down and changes to a distinct downwash.
• Vortex duration depends strongly on ambient weather conditions and a variance of 300% was observed on those days most conducive[7] to vortex persistence and duration compared with those observed on typical days.
• Typically, helicopters with higher gross weight, larger rotor diameters, and larger numbers of rotor blades generated vortices of larger core diameters.
• Probe tests revealed that helicopter vortices did not descend in the same predictable manner as for fixed‑wing aircraft. Some vortices descended; some remained level; and some initially descended, levelled off, and then ascended above the altitude of the generating helicopter.

Figure 3: Rotor wake vortices visualised with smoke generators

Visualisation of the wake vortices behind an S‑76A helicopter in forward flight with smoke generators from the FAA (1996) flight tests. Source: Reddit

Meiris (n.d.) provided an article for the United States Helicopter Safety Team website, on the subject of Avoiding helicopter wake turbulence. The article referenced the FAA 1996 flight test report and provided the following recommendations:

As a result of these findings and the studies conducted regarding helicopter downwash in a hover, a few guidelines have been developed to increase awareness around helicopter wake turbulence:

• For hovering flight or a hover taxi, stay three rotor diameters away.

• For forward flight, a minimum of 3 nm [NM] separation is recommended, especially from larger helicopters. The investigation we discussed previously discovered that even at 3nm [NM], the planes encountered uncommanded pitch and roll oscillations.

• Leave 2 minutes for the rotor vortices to dissipate behind a helicopter in forward flight.

Related occurrences

The French Bureau of Enquiry and Analysis for Civil Aviation Safety investigation BEA2019-0234, Accident to a paraglider involving the Airbus - EC135 - T2 PLUS registered F-HTIN, examined a fatal paraglider accident in 2019. The paraglider’s wing collapsed after encountering the rotor wake from an Airbus EC135 helicopter, which drifted with the wind from the helicopter’s flightpath onto the paraglider (Figure 4).

Figure 4: Simulation of rotor wake drifting onto the paraglider

Source: YouTube – Bureau of Enquiry and Analysis for Civil Aviation Safety, annotated by ATSB

The 2022 United States National Transportation Safety Board investigation WPR22LA072 found that the pilot of a Cessna 120 attempted a go‑around about 20 seconds behind the passage of a Bell UH‑1H helicopter. During the go‑around the Cessna encountered wake turbulence, resulting in a loss of control and collision with terrain (Figure 5). The report indicated light wind conditions of 4 kt at the airport.

Figure 5: Loss of control accident from rotor wake

Source: YouTube – Aviation Safety Network, annotated by ATSB

Safety analysis

Primary radar data and the pilot’s report indicated that EWM entered the Parramatta River helicopter lane and passed over the Ryde Bridge about 9 seconds behind and slightly below the EC120 helicopter. At this point, EWM encountered heavy turbulence, an uncontrolled descent and a low rotor speed when the pilot applied full collective to avoid a collision with the water.

The uncontrolled descent and low rotor speed condition resulted in the pilot conducting a forced landing on the shoreline of the Parramatta River. 

The incident occurred under relatively calm wind conditions and EWM operated in a serviceable condition for the return flight. Subsequent maintenance inspections of the helicopter found no fault. Furthermore, EWM passed overhead the Ryde Bridge in sufficient proximity to a preceding heavier 3‑bladed helicopter to be subject to a rotor wake induced upset. Therefore, the ATSB concluded that the sudden onset of turbulence and uncontrolled descent were likely the result of EWM encountering rotor wake turbulence from a preceding EC120 helicopter. 

Findings

From the evidence available, the following findings are made with respect to the wake turbulence encounter and forced landing involving Robinson R44, VH-EWM, about 15 km north‑east of Bankstown Airport, New South Wales, on 7 April 2025. 

Contributing factors

• It is likely that the incident helicopter entered the rotor wake from a preceding heavier helicopter, which resulted in control difficulties, an uncontrolled descent, low rotor speed warning and a forced landing.

Safety actions

Safety action by the Civil Aviation Safety Authority

Following review of the draft report, the Civil Aviation Safety Authority undertook proactive safety action to improve existing guidance about helicopter wake vortices in Advisory Circular 91-16. The updated version of the advisory circular was released on 17 July 2025 and can be found at the link: AC 91-16 v1.2 - Wake turbulence.

Sources and submissions

Sources of information

The sources of information during the investigation included:

• Airservices Australia
• Civil Aviation Safety Authority
• the operator and maintenance organisation for VH-EWM
• the pilot and passengers of the incident flight

References

Bureau of Enquiry and Analysis for Civil Aviation Safety. (2021). Accident to a paraglider involving the Airbus - EC135 - T2 PLUS registered F-HTIN on 11 May 2019 at Le Conquet (Finistère). https://bea.aero/fileadmin/user_upload/BEA2019-0234.en.pdf

Federal Aviation Administration. (2023). Aeronautical information manual. http://www.faa.gov/air_traffic/publications

Federal Aviation Administration. (1996). Flight test investigation of wake vortices generated by rotorcraft in forward flight (DOT/FAA/CT-94/117). https://apps.dtic.mil/sti/tr/pdf/ADA318103.pdf

Meiris, J. (n.d.). Avoiding helicopter wake turbulence. https://ushst.org/avoiding-helicopter-wake-turbulence/

National Transportation Safety Board. (2022). Aviation investigation final report (WPR22LA072). Investigation docket https://data.ntsb.gov/Docket?ProjectID=104480

Submissions

Under section 26 of the Transport Safety Investigation Act 2003, the ATSB may provide a draft report, on a confidential basis, to any person whom the ATSB considers appropriate. That section allows a person receiving a draft report to make submissions to the ATSB about the draft report. 

A draft of this report was provided to the following directly involved parties:

• Civil Aviation Safety Authority
• the maintenance organisation for VH-EWM
• the operator and pilot of the incident flight.

No submissions were received.

Purpose of safety investigations The objective of a safety investigation is to enhance transport safety. This is done through:  identifying safety issues and facilitating safety action to address those issues providing information about occurrences and their associated safety factors to facilitate learning within the transport industry. It is not a function of the ATSB to apportion blame or provide a means for determining liability. At the same time, an investigation report must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. The ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action. Terminology An explanation of terminology used in ATSB investigation reports is available here. This includes terms such as occurrence, contributing factor, other factor that increased risk, and safety issue. Publishing information Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2025   Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Commonwealth Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this report is licensed under a Creative Commons Attribution 4.0 International licence. The CC BY 4.0 licence enables you to distribute, remix, adapt, and build upon our material in any medium or format, so long as attribution is given to the Australian Transport Safety Bureau.  Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.