Investigation summary 

What happened

On 15 July 2024, a QantasLink De Havilland Aircraft of Canada Limited DHC-8-402, (Dash 8), registered VH-QOD, was taxied to reposition to a different bay at Wagga Wagga Airport, New South Wales. This repositioning required the Dash 8 crew to taxi out onto the runway and then return to a different bay. As the Dash 8 entered runway 05, the crew were unaware that a Piper PA-28, registered VH-XDK, operating on a training flight by the Australian Airline Pilot Academy (AAPA), had already commenced its take-off roll from the opposite end of the runway. 

The instructor of the PA-28 was aware of the Dash 8 taxiing and assumed that the Dash 8 had received and understood their previous positional broadcasts. However, the crew of the Dash 8 were not aware of the PA-28 preparing for take-off on runway 23.

Upon entering runway 05, the Dash 8 was notified by a preceding aircraft that had landed, that there was another aircraft taking off from runway 23. The captain of the Dash 8 immediately stopped the aircraft and engaged reverse thrust to clear the runway. Simultaneously, the PA-28 pilot rejected their take-off from runway 05. 

What the ATSB found

The ATSB identified that during ground-based repositioning, the Dash 8 was unaware of the PA-28 that was taxiing for take-off on an opposing runway. This led to a situation where the crew of the Dash 8 had an incomplete comprehension of the ground-based traffic.  

Without any prior alert or expectation of the presence of the PA-28, the Dash 8 crew did not visually detect the PA-28 on take-off from the reciprocal end of the runway, prior to the Dash 8 entering the runway to taxi to the terminal. The PA-28 presented a difficult visual target due to its size and orientation over the 1,700 m distance, making unalerted detection unlikely.

Once made aware of the potential conflict of collision, the captain of the Dash 8 immediately reversed the aircraft away from the preferred runway without confirming that no other aircraft or obstacle was behind the Dash 8.

While the pilot of the PA-28 gave all the required radio calls, they did not directly communicate with the Dash 8 crew to identify the possible conflict, then positively arrange separation. Additionally, the Dash 8 ground-based reception, on VHF communications panel radio 2, had reduced strength and clarity. 

The QantasLink radio procedure required crew to use the number 2 VHF communications panel radio to broadcast and receive on local frequencies during operations at a non-controlled aerodrome. This likely reduced ground-based reception strength and the likelihood of the Dash 8 hearing other traffic in certain circumstances.

Reduced VHF ground-based communication was identified at the eastern end of runway 23 at Wagga Wagga Airport. Local operator reports and procedures indicate that reduced communication strength and clarity can be experienced on the eastern end of taxiway A to other areas of the airport.

During taxi for take-off, the PA-28 was not broadcasting transponder information, which also did not identify them to other traffic in the vicinity of the aerodrome. The AAPA procedure requiring the selection of the transponder to ‘ALT’ before entering the runway, rather than prior to taxi, became a missed opportunity to provide electronic enhancement of situational awareness to other airport operators. 

What has been done as a result

ATSB issued a safety advisory notice (AO‑2024‑041‑SAN‑01) to advise pilots and operators to review their procedures to ensure that mode S transponders are on from first movement of the aircraft. 

AAPA has advised the ATSB that changes have been made to its PA-28 Flight Crew Operating Manual, and Quick Reference Handbook, which reflects the transponder being selected to ALT mode after start in both the checklist as well as being incorporated as part of the scan action flow.

On 6 December 2024, De Havilland Aircraft of Canada Limited issued 2 flight operations service letters relating to radio communications, with one covering Dash 8 100-300 series aircraft and the other covering the Dash 8 400 series aircraft. The service letters remind operators that ground‑based VHF communications are affected by line of sight and can be impacted by buildings, terrain or aircraft structures and that use of VHF COM 1 is more effective for ground-based communications with other aircraft on the ground. 

The airport operator had issued a notice to airmen (NOTAM) notifying aircraft of a potential radio black spot at Wagga Wagga Airport under some conditions. 

QantasLink issued a safety alert notice encouraging crew to consider additional precautions when taxiing for departure or any other manoeuvring on-ground at Wagga Wagga Airport. This advice was issued due to on-ground VHF communications between aircraft possibly being affected by obstacles in the line of sight.

Safety message

Communication and self-separation in non-controlled airspace is one of the ATSB’s SafetyWatch priorities. Wherever you fly, into either non-controlled or controlled aerodromes, maintaining a vigilant lookout at all times is important. Situational awareness and alerted see-and-avoid is an effective defence against collision and good airmanship dictates that all pilots should be looking out and not be solely reliant on the radio for traffic separation. Being aware of other nearby aircraft and their operational intentions is important.

Effective use of all available sources of information is an effective risk control to achieve enhanced situational awareness and an accurate mental model of other traffic at a non‑controlled aerodrome.

Pilots are reminded that although accurate and timely radio calls play a critical role in ensuring collision avoidance in uncontrolled airspace, they cannot assume from an absence of other radio calls that there is no conflicting traffic. This is particularly important in an environment where there is high expectation of mixing with other aircraft of different sizes, flight rules and performance levels operating at the same time, in the same airspace. 

Pilots can enhance their situational awareness and mutual traffic separation by: 

• making the recommended broadcasts when in the vicinity of a non-controlled aerodrome
• actively monitoring the common traffic advisory frequency while maintaining a visual lookout and constructively organising separation through direct contact with other aircraft
• ensuring mode S transponders, where fitted, are selected to transmit altitude information before taxiing
• not hesitating to contact another aircraft if there is any uncertainty as to their position and/or intentions. 

The ATSB SafetyWatch highlights the broad safety concerns that come out of our investigation findings and from the occurrence data reported to us by industry. This investigation report highlights the safety concerns around reducing the collision risk around non-towered airports.

The occurrence

On 15 July 2024, a Piper Aircraft Corporation PA-28, registered VH-XDK, operated by the Australian Airline Pilot Academy (AAPA), with an instructor and student on board, taxied from the apron at Wagga Wagga Airport, New South Wales (Figure 1) for a training flight. At 1609 local time, the crew broadcast they were taxiing for runway 23 (Figure 2, position A). About 20 seconds later, the crew of an inbound Saab 340 responded, asking for the call sign to be repeated. The Saab 340 was 30 NM away at this stage and could not understand the call sign, even when repeated twice more. About 5 minutes later at 1614:12, the crew of the Saab 340 broadcasted that they were now on a 10 NM straight in approach for runway 23 (see Appendix A – CTAF Recordings).

At about the same time, the 2 crew members of a QantasLink De Havilland Aircraft of Canada Limited DHC-8-402 (Dash 8), registered VH-QOD, were on board the aircraft at the Wagga Wagga Airport apron. The crew were making preparations to reposition the aircraft from the apron to another bay (on the apron) by taxiing the aircraft via the runway (Figure 1). The Dash 8 crew reported they spent about 5 minutes[1] preparing the aircraft on the flightdeck prior to taxiing. However, they did not recall hearing the initial taxi call and callsign repeats from the PA-28, but did recall hearing the 10 NM inbound call from the Saab 340. 

Figure 1: Intended taxi routes for aircraft

Source: Google Earth, annotated by the ATSB

At 1616:27 the Saab 340 crew broadcast that they had joined a 5 NM final for runway 23. 

At 1617:31 (Figure 2, position B, Dash 8) the Dash 8 crew broadcast that they were taxiing for runway 05. At this time, the PA-28 was on taxiway A, approaching runway 23 (Figure 2, position B, PA-28). 

Figure 2: Position of respective aircraft at given times

Source: Google Earth, annotated by the ATSB. Note: aircraft not to scale

Being aware only of the inbound Saab 340, the Dash 8 crew communicated directly with them. This was to gain an understanding of the Saab 340 crew’s intentions on landing and exiting the runway. While still taxiing towards the runway at 1618:32, the Dash 8 crew told the Saab 340 crew of their intent to enter and backtrack from the threshold of runway 05, keeping out of the way of the Saab 340’s landing.

The instructor of the PA-28 recalled hearing the radio calls from both aircraft during this time. The PA-28 was aware of the Dash 8 taxiing for runway 05 and assumed that the Dash 8 had also heard their earlier calls, and would hold short of the runway for their departure. 

At 1619:20, the PA-28 was positioned on runway 23, with strobe lights on, behind the landed Saab 340 and the instructor made a ‘lining up and holding’ call (Figure 2, position B, PA-28). About 30 seconds later the Saab 340 crew made a broadcast that they were clear of the preferred runway (Figure 2, position C, Saab 340). The PA-28 instructor then broadcast their ‘rolling call’ (Figure 2, position C, PA-28) and commenced take-off at 1620:07. 

At 1620:25, the Dash 8 crew gave an entering call for runway 05 with intentions to taxi to taxiway C (Figure 2, position C, Dash 8) and then entered the runway.

On hearing the Dash 8 crew’s broadcast, the Saab 340 crew notified the Dash 8 that ‘there is an aircraft taking off on runway 23’ at 1620:33. Simultaneously, the PA-28 student pilot also rejected their take-off. 

The Dash 8 crew immediately stopped at their current position, with the aircraft still perpendicular with the runway and estimated by the captain to be about 5 m inside the runway gable markers.[2] The captain acknowledged the broadcast and then visually identified the PA-28 on the runway. 

The captain of the Dash 8 subsequently engaged Beta plus power mode[3] (reverse thrust) to reverse the Dash 8 backwards along taxiway B to ensure clearance of the preferred runway. The PA-28 crew broadcast they would exit runway 23 on taxiway D.

The Dash 8 captain then established direct radio contact with the PA-28 crew, stating ‘nil radio transmission heard and they are still broken’.

The PA-28 vacated the runway and taxied again for a subsequent departure. The Dash 8 then re-entered runway 05 and backtracked runway 23 and vacated to the apron.

Context

Pilot information

Flight crew VH-QOD (Dash 8)

The captain held an Air Transport Pilot Licence (Aeroplane) and a valid class 1 aviation medical certificate. They reported a total flying time of 2,912 hours with 2,389 of those being on the Dash 8 aircraft type. The captain reported being familiar with Wagga Wagga Airport and had operated there regularly. 

The first officer (FO) held a Commercial Pilot Licence (Aeroplane) (CPL-A) and a valid class 1 aviation medical certificate. They reported a total flying time of 1,312 hours with about 438 of those hours being on the Dash 8. The FO had been employed with the operator for approximately a year and had flown into Wagga Wagga Airport during their training and subsequent line operations.

Flight crew VH-XDK (Piper PA-28)

The instructor of the PA-28 held a CPL (Aeroplane) with a grade 2 instructor rating. They held a valid class 1 aviation medical certificate and reported a total flying time of 796 hours, with approximately 146 hours being on the PA-28. The instructor regularly operated out of Wagga Wagga Airport, as it was the company base of operation. 

The student pilot of the PA-28 held a class 1 aviation medical certificate and reported about 17 hours of flying time.

Aircraft information 

VH-QOD (Dash 8)

The De Havilland Aircraft of Canada Limited (DHC) DHC-8-402, was a high-wing, pressurised, commuter aircraft powered by 2 turboprop engines. VH-QOD was manufactured in Canada in 2006 and was first registered in Australia on 22 May 2006.

VHF radio antenna position and condition

The Dash 8 had 2 VHF radio systems, each using separate antennas installed on the upper and lower fuselage (Figure 3): 

• Number 1 VHF COM antenna (VHF COM 1) mounted on the roof of the aircraft forward of the wings.
• Number 2 VHF COM antenna (VHF COM 2) mounted on the belly of the aircraft.

Figure 3: Dash 8 with annotations of VHF antenna locations on airframe

Source: Gyo Kamata Jetphotos.com, annotated by the ATSB

Transponder and ADS-B system on the Dash 8

The Dash 8 was equipped with an automatic dependent surveillance-broadcast ADS-B mode S transponder. ADS-B uses digital positional data provided by a GPS sensor on board the aircraft. This information is transmitted via the same channels used in conventional aircraft transponder transmissions. The Dash 8, ADS-B transponder units are fully integrated through the Dash 8 audio and radio control and display unit. Positional data for each ADS-B transponder is provided through the aircraft flight management system.

TCAS system on the Dash 8

A traffic collision avoidance system (TCAS) was fitted to the Dash 8. A TCAS interrogates the transponders of nearby aircraft and uses this information to calculate the relative range and direction of this traffic. The system displays this information to the flight crew, providing situational awareness of the location of other aircraft, and is available on the ground and when airborne. 

TCAS is designed to be used as an airborne device, which performs surveillance of nearby aircraft and provides information on the relative direction and altitude of these aircraft so that the collision avoidance algorithms can perform their function (Federal Aviation Administration, 2011) 

Limitations exist for the use of TCAS in lateral separation. Current generations of TCAS are primarily focused on vertical separation (climb or descent); TCAS I and II generations of TCAS do not provide horizontal separation.

In certain circumstances functions and alerts within the TCAS system are inhibited. An example is documented in FAA AC 90-120 (Federal Aviation Administration, 2024) that noted:

2.11.5 ACAS[4] does not display aircraft on the ground and may not display an aircraft when own-ship ACAS estimates that the other aircraft is below 380 feet Above Ground Level (AGL), unless the other aircraft is operating with a Mode S transponder that reports airborne status.

While the use of TCAS for monitoring other aircraft on the ground was not an operator procedure, the Dash 8 crew identified that they regularly used it as an aid for situational awareness. The captain and FO both recalled reviewing their TCAS with only one aircraft (the Saab 340) being seen. 

Flightradar24 application 

The crew of the Dash 8 utilised an electronic flight bag (EFB).[5] Additionally during taxi, they reviewed the Flightradar24 application as a tool to enhance their situational awareness of potential traffic in the area. 

Flightradar24 is a global flight tracker that shows live air traffic from around the world and combines data from several sources including ADS-B. 

QantasLink did not have a formal policy referring to the use of the Flightradar24 application. However, the flight crew advised that this application is often used by pilots as a resource to aid situational awareness. As with TCAS, only the Saab 340 aircraft was visible on the Flightradar24 app during the Dash 8 taxi period.

VH-XDK (Piper PA-28)

VH-XDK was a Piper Aircraft Corporation, PA-28-161 Cherokee Warrior III, manufactured in the US in 2008. VH-XDK was a single‑engine, 4‑seat, low‑wing design aircraft, operated by Australian Airline Pilot Academy Pty Ltd (AAPA).

VHF radio systems on PA-28

The aircraft was equipped with 2 Garmin G430[6] VHF navigation/communication systems, transmitting at a minimum of 10 watts through 2 independent antennas. The VHF COM 2 antenna is located on the top of the fuselage and the VHF COM 1 antenna is located on the underside of the fuselage.

Figure 4: Antenna location on PA-28

Source: PA-28 operator 

Transponder and ADSB system PA-28 

The PA-28 was fitted with a Garmin GTX 330ES MODE-S, providing a mode S digital transponder signal. The transponder likely would have made the PA-28 electronically conspicuous to the Dash 8 crew through the flight crew’s Flightradar24 application on their EFBs, when selected to ON/ALT.

Meteorological information

Weather conditions at Wagga Wagga Airport around the time of the occurrence were identified as a moderate north-westerly wind at about 13 kt, with greater than 10 km visibility. The cloud was reported as broken (between 5–7 oktas[7]) between 2,000 ft and 2,600 ft above ground level.

Airport information

Wagga Wagga Airport

Wagga Wagga Airport is a regional, certified aerodrome, located about 6 NM to the south‑east of Wagga Wagga township in New South Wales, Australia. The aerodrome had an elevation of 724 ft (221 m) above mean sea level (AMSL), and had 2 runways running north‑east, south-west and north-west, south-east. The primary sealed runway at Wagga Wagga is runway 05/23 and is 1,770 m long with the secondary, unsealed runway 12/30 being 851 m long (Figure 5). 

Runway 05/23 exhibited a minor elevation differential between each runway threshold. The threshold of runway 05 was recorded at 216 m AMSL while the threshold of runway 23 was slightly lower at 214 m AMSL, resulting in a 2 m downward gradient from runway 05 to runway 23. Approximately 300 m from the runway 05 threshold, the elevation increased to 217.5 m AMSL, a rise of 1.5 m, before gradually declining towards the runway 23 threshold. This subtle elevation change created a minor undulation in the runway profile. 

Figure 5: Wagga Wagga Airport layout

Source: En Route Supplement Australia (ERSA). Airservices Australia, annotated by the ATSB

Airspace

Wagga Wagga Airport was located within non-controlled Class G[8] airspace which was available for use by aircraft operating under visual flight rules (VFR) and instrument flight rules (IFR). 

No air traffic control separation service was provided between aircraft operating in this airspace. Pilots are responsible for making themselves aware of nearby aircraft and maintaining mutual self‑separation. The primary method of traffic separation at Wagga Wagga Airport was by VHF communications coupled with visual reference, and relied on pilots using ‘alerted see-and-avoid’[9] practices (see See-and-avoid).

The Wagga Wagga CTAF operated on the VHF radio band and required pilots to monitor and make some positional broadcasts when operating within the vicinity of the aerodrome, or within 10 NM.

Reported radio dead spot

The Australian Airline Pilot Academy (AAPA) had issued an internal notice to all crew (NOTAC) on 23 August 2023, advising of a potential radio ‘deadspot’ on Wagga Wagga Airport. The NOTAC detailed:

NOTAC 032/17 CURRENT 23 08 1100 LST - Until Removed To: All Pilots 

POTENTIAL RADIO DEADSPOT ON WAGGA AIRPORT 

There is an identified radio dead spot on Taxiway Alpha between taxiway Charlie and the runway 23 holding point, prior to entering Runway 23, crew must visually check that no aircraft are entering the runway from taxiway Charlie. 

Authority: Chief Pilot.

Prior to the occurrence, the aerodrome operator had identified that ground-based vehicle radio communications were observed to have reduced effectiveness in some circumstances, however no further testing was carried out.

Radio signal transmission reception

The ATSB conducted an analysis to determine whether the radio line of sight (LOS) between the 2 aircraft, positioned at opposite ends of the runway, was obstructed by the terrain elevation at approximately 300 m down runway 05.

Firstly, a geometric LOS analysis was conducted. The analysis utilised a strict geometric LOS model, assuming a direct, unobstructed path between the transmitter and receiver, as described in standard International Telecommunications Union (ITU-R) (2019) radio propagation principles (Parsons, 2000). The results determined that the terrain elevation created a physical obstruction, blocking the direct radio LOS between the aircraft.

To assess the potential for radio communication despite the LOS obstruction, diffraction[10] effects were evaluated. The inclusion of diffraction effects was used as radio waves, particularly in the very high frequency (VHF) band used in aviation, can propagate beyond geometric LOS by bending around obstacles such as terrain features  (Rappaport, 2002). This phenomenon enables communication in scenarios where direct LOS is obstructed.

A knife-edge diffraction model was used to estimate the signal loss caused by the terrain. This model, a standard method in radio wave propagation analysis, approximates the terrain feature as a sharp, idealised edge obstructing the radio wave path (International Telecommunications Union (ITU-R), 2019). The knife-edge model was selected due to its applicability for analysing diffraction over a single, well-defined obstacle and its computational simplicity for VHF frequency analysis.

The calculations were completed using the Wagga Wagga CTAF frequency of 118.2 MHz and assumed values derived from generic radio and antenna specifications. 

Results indicated that despite the geometric LOS obstruction caused by the terrain, communication between the 2 aircraft was feasible[11] due to diffraction effects. 

A similar calculation was conducted between taxiway C and taxiway A5. It indicated that communication between 2 aircraft was also feasible. The calculated signal strength indicated that a radio transmission should have theoretically been received clearly. 

All calculations regarding VHF line of sight communication were theoretical and assumed idealised conditions. Noting the advice by AAPA of a potential radio dead spot on the airport, it should be noted that this observed dead spot may result from real-world environmental factors such as localised interference or other unmodeled conditions like vegetation or temporary obstacles, which cannot be fully replicated in theoretical calculations.

Visibility between opposite runway thresholds

A clear line of visual (eye level) sight existed between the taxiway B holding point at the threshold of runway 05 and the threshold of runway 23 (Figure 6). The pilot of the PA-28 reported they could clearly see the Dash 8 at the threshold of runway 05, as the Saab 340 vacated the runway, prior to them commencing take-off. The crew of the Dash 8 reported being able to see the Saab 340 land and vacate the runway, but did not visually detect the PA-28 prior to entering the preferred runway. After they were notified by the Saab 340, the captain and first officer were able to visually identify the PA-28 on runway 23 which was at that stage moving towards the Dash 8 on its take-off roll. 

Both the FO and the captain commented during interview of a possible obstruction to their line of sight from the PAPI[12] installation. The PAPI installation is positioned to the left of runway 05. The orientation was to the left of the Dash 8 when the aircraft was holding on taxiway B, prior to entering runway 05. PAPI units must be no more than 0.9 m above ground level. 

Figure 6: View from ground level, threshold runway 05

Source: airport operator, annotated by the ATSB 

It was noted that the visual range from holding point B to the threshold of runway 23 is greater than 1,700 m over undulating terrain, with the PA-28 presenting a small target for visual identification by the Dash 8 crew. 

ATSB calculations indicate there was a clear line of sight between the eye height of the Dash 8 crew and the eye height of the PA-28 crew. The visual line of sight between both aircraft was about 1.7 m above the geographical terrain undulations between the thresholds of runways 05 and 23. 

Human performance limitations

Object perception

There are limitations to the size of objects that are perceptible at a distance. The ATSB considered whether the crew of the Dash 8 could have been able to detect there was a small aircraft (PA-28) at the threshold of runway 23 when the Dash 8 was at taxiway B at the runway 05 end, based on the known limitations of distant object perception. It was reported by the PA-28 pilots that the Dash 8 was visible to them when they were lined up for take-off once the Saab 340 cleared the runway. However, the Dash 8 is a larger aircraft and was side-on to the PA-28 pilots at this point, and they were alerted to the Dash 8’s presence through broadcasts, however this does not mean that the Dash 8 crew could identify the PA‑28.

The 3 main factors that would affect the visual image size of the PA-28 aircraft in the Dash 8 pilot’s eye are: 

• the dimensions of the aircraft
• its relative orientation
• distance from the viewer. 

Past research (Hobbs, 2004) has shown distant objects can be seen when their visual angle is at least 24−36 minutes of arc (0.4−0.6°), and down to a minimum of 12 minutes of arc (0.2°) in ideal viewing conditions (ATSB, 2025).

Once the Saab 340 had cleared the runway and the visual path to the PA-28 was clear for the Dash 8 crew, the PA-28 was lined up for take-off. When lined up on the runway, the most prominent part of the PA-28 was the fuselage, which was 1.65 m high, with a wingspan of 10.79 m (noting that the wings would be edge-on). At the distance of 1,700 m, this would have presented a visual image that was about 3 minutes of arc (0.06°).[13] Additionally, as the PA-28 was moving down the runway, the gradual change in angular size would likely have been virtually imperceptible due to the eye’s limited sensitivity to small angular variations. Therefore, in the absence of confounding factors, it would have been very unlikely that the PA-28 would have been detectable by the Dash 8 pilots at that distance.    

Other factors can also exist that can affect whether a pilot will be able to see another aircraft, including the background that an object is seen against. As can be seen in Figure 6, the background behind the runway 23 threshold was dark and would have provided good contrast to the white PA-28 aircraft, making detection easier.

Finally, the use of lights by the PA-28 would have enhanced detection. Strobe lights would have enhanced detection though flash and movement. However, landing lights are angled forward and downwards, and may not have been as detectable by the Dash 8 crew.

See-and-avoid 

At and around non-controlled aerodromes, pilots are responsible for making themselves aware of nearby aircraft and maintaining separation. Safe operations at non-controlled aerodromes rely on all pilots maintaining an awareness of their surroundings and other aircraft, the principle of ‘see‑and‑avoid’. 

A visual traffic search in the absence of specific traffic information is less likely to be successful than a search where traffic information has been provided. Knowing where to look can greatly increase the chance of sighting the traffic.

An ‘unalerted’ search is one where reliance is entirely on the pilot searching for, and sighting, another aircraft without prior knowledge of its presence. An ‘alerted’ search is one where the pilot is alerted to another aircraft’s presence, typically through radio communications or aircraft-based alerting systems. An alerted search supports a pilot’s situational awareness and enhances their visual lookout for traffic by developing an expectation of visually acquiring the traffic in a particular area.

Issues associated with unalerted see-and-avoid have been detailed in the ATSB research report, Limitations of the See-and-Avoid Principles(Hobbs, 2004). The report highlights that unalerted see-and-avoid relies entirely on the pilot’s ability to sight other aircraft. 

Tools to enable ‘alerted see-and-avoid’ include:

• VHF radio
• transponders, used by traffic collision avoidance system (TCAS)
• automatic dependent surveillance broadcast (ADS-B)
• electronic flight bag (EFB) applications 

Hobbs (2004) identified that an alerted search is likely to be 8 times more effective than an unalerted search, this highlighted that knowing where to look greatly increases the chances of sighting other traffic.

Alerted see-and-avoid relies on pilot or crew awareness of all traffic in their vicinity, especially those that may be considered a hazard to their operations. Enhanced situational awareness requires the pilot or crew mental model of the location and intentions of nearby traffic to be updated regularly to form an evolving understanding of the nearby traffic. Without this information, the likelihood of effective situational awareness is degraded, and the mental model and shared understanding of hazards is compromised.

Operational procedures

QantasLink Dash 8 procedures

Non-controlled aerodrome VHF radio procedure

The QantasLink Dash 8 standard operating procedure for departure from a non‑controlled aerodrome (such as Wagga Wagga Airport), at the time of the occurrence, required VHF COM 1 to be set to the area frequency and VHF COM 2 to be set to the CTAF, as passenger boarding commenced. 

Prior to releasing the handbrake to taxi, a call was to be made to the relevant air traffic centre on VHF COM 1 and then followed by a taxi call to the CTAF on VHF COM 2. However, due to this being a repositioning flight, the crew only broadcast on VHF COM 2 on the CTAF frequency.

Controlled aerodrome VHF radio procedure

The QantasLink Dash 8 standard operating procedure for departure from a controlled aerodrome, at the time of the occurrence, required VHF COM 1 to be set to the primary air traffic control frequency for ground and air-based communications. The VHF COM 2 was set for other communications (such as ATIS,[14] AWIS,[15] PAL and CTAF). 

Reverse thrust on the ground

QantasLink policy for engaging reverse thrust during ground operations required prior approval from the Head of Flight Operations. This approval requirement was due to the risks associated with not being able to confirm clearance from other traffic or obstacles that cannot be visually identified from the flight deck. The captain advised during interview that they were aware of this policy.

Procedures for transponder use

The QantasLink procedure was to set the transponder to ‘ON/ALT’ with the applicable transponder code[16] as part of the pre-flight process. This is achieved either after obtaining airways clearance or (at airports where airways clearance is not obtained until after take-off), as part of the pre‑flight process before engine start.

AAPA PA-28 operator procedures

VHF radio procedure

The AAPA procedures, contained within the ‘Piper Warrior III Crew Operating Manual’, normal procedures, required COM 1 to be selected to the local area frequency, tower, ground, clearance delivery, approach or departure frequency as required. COM 2 was required to be set to the CTAF, ATIS, AWIS, MULTICOM or guard frequency (121.5 MHz) as required.

The instructor confirmed during interview that, at the time of the occurrence, COM 1 was set to Melbourne Centre (area frequency) and COM 2 was set to the Wagga Wagga CTAF.

Procedures for transponder use

The PA-28 operator’s’ procedure for transponder operation was to set the applicable transponder code and select the STBY[17] function during taxi. Prior to entering the runway, ON/ALT mode[18] was to be selected.

The PA-28 operator’s training manual notes:

All pilots must ensure ALT is selected on the transponder during the LINE UP CHECKLIST. Other aircraft equipped with TCAS rely on transponder information for pilot alerting and collision avoidance functions.

PA-28 external light use

The PA-28 operator’s procedure for the selection of external lights is documented in its Piper Warrior III flight crew operating manual, normal procedures. 

As required by the ‘pre line up scan action flow’, the pilot selects the landing lights ‘on’ (during daytime operations), immediately prior to commencing a take-off roll at a non‑controlled aerodrome. 

The instructor recalled that normal behaviour would involve the pilot selecting the landing lights ‘on’, when the ‘rolling call was issued’.  

Regulations regarding transponder use

The use of surveillance equipment such as ADS-B and transponder is outlined in the Civil Aviation Safety Regulations Part 91 Manual of Standards (MOS) and the Airservices Australia Aeronautical Information Publication (AIP).

Chapter 26 of the Part 91 MOS, operation of surveillance equipment – general requirements stated, among other things, that transponder equipment required to be fitted and carried on an aircraft must be continuously operated. It also identified that ‘continuous operation’ for a transponder means that the equipment must be operated in a mode that enables a secondary surveillance radar (SSR)[19] response to be transmitted and, where an altitude reporting capability is available, that this capability is also activated.

AIP Australia ENR 1.6 paragraph 7.1.9 stated:

A pilot operating a Mode S transponder must:

b. On receipt of ATC clearance, or requesting the earlier of Push Back or Taxi, select TA/RA/XPDR/ ON AUTO as applicable. 

CASA advised that item ‘b’ only applies at controlled aerodromes. At non-controlled aerodromes, transponders must be turned on prior to becoming airborne.

Radio communication

VHF radio is the primary communication tool commonly used to provide ‘alerted see‑and‑avoid’ from sport and recreational private flying to air transport. Broadcasts on the CTAF to any other traffic in the vicinity of a non-controlled aerodrome are made to provide situational awareness, traffic separation and deconfliction to other traffic in the vicinity.

Positional broadcasts

Civil Aviation Safety Regulation 91.630 made certain radio calls mandatory for aircraft that are fitted with or carry a VHF radio. Chapter 21 of the Part 91 Manual of Standards (MOS) prescribed one type of mandatory broadcast that applies at all non-controlled aerodromes, namely:

When the pilot in command considers it reasonably necessary to broadcast to avoid the risk of a collision with another aircraft.

To aid in increasing situational awareness at non-controlled aerodromes, recommended broadcasts are published by the Civil Aviation Safety Authority (CASA). These broadcasts enable pilots to alert other traffic to their location and intentions before take‑off, inbound to land at, or if intending to overfly a non-controlled aerodrome. 

Standardised radio transmissions and phraseology assist with effective and efficient radio communication. To achieve this, the application of recommended positional broadcasts (Table 1) for VFR traffic are published in CASA Advisory Circular (AC) 91-10, Operations in the vicinity of non-controlled aerodromes. 

Table 1: Recommended positional broadcasts in the vicinity of a non-controlled aerodrome

Source: CASA AC 91-10 Operations in the vicinity of non-controlled aerodromes

Limitations of positional broadcasts

Positional broadcasts are a one-way communication and do not imply receipt of information by other parties unless direct radio contact is made between stations. 

Positional broadcasts rely on the accuracy of the information being broadcast and the ability of other traffic receiving, comprehending and reacting to this information. 

CASA AC 91-10 stated: 

8.2.1 Pilots are reminded that although correct and informative radio calls play a critical role in ensuring collision avoidance in uncontrolled airspace, to ensure the safety of their aircraft they cannot assume that an absence of other radio calls means there are no nearby or conflicting aircraft. 

8.2.2 Pilots must continually look out for other aircraft, even when their broadcasts have generated no response

8.2.3 Accidents and incidents have occurred where pilots incorrectly assessed the threat posed by another aircraft, either due to the pilot incorrectly assessing the relative aircraft flight paths, or inaccurate information being provided by other pilots.

Dash 8 radio reception and transmission

Two Dash 8 ground communication events has been previously identified by ATSB investigations (AO-2023-025 and AO-2023-050) at Mildura Airport (see Related occurrences).

These 2 events resulted in the second investigation testing the Dash 8’s VHF systems. The testing was conducted by the ATSB, the Australian Media and Communications Authority[20] and QantasLink. The testing measured the transmission power pattern of the Dash 8’s 2 communication systems (upper and lower antennae), when the aircraft was on the ground at Mildura Airport.

The testing identified that ground‑based Dash 8 signal strength reception could be adversely affected by the aircraft’s orientation relative to the other aircraft or antenna locations. Additionally, the average signal strength forward of the aircraft was 8.5 dBm[21] stronger than the average signal strength behind and to the side of the aircraft. A significant recorded signal strength and clarity reduction on both VHF COM 1 and VHF COM 2 radios was observed when the tail of the Dash 8 was pointed towards the receiver.

It was further identified that reception and transmission on VHF COM 2 on the ground (via the lower antenna, as used for QantasLink Dash 8 ground communications at non‑controlled aerodromes) had significantly reduced strength and clarity compared to VHF COM 1.

The Dash 8 manufacturer, De Havilland Aircraft of Canada Limited, advised that VHF COM 1 was expected to provide more reliable performance in ground-based communication with other ground stations. 

Two flight operations service letters (Appendix B – Flight Operations Service Letters) were released on 6 December 2024 for the Dash 8-100/200/300 (DH8-SL-23-008A) and Dash 8‑400 (DH8-SL-23-020A). Details included a description of the limitations of VHF line of sight communications and the recommendation that VHF COM 1 may provide a better signal (receiving and transmitting) to other stations on the ground, or nearby in the air. 

As a result of the first occurrence at Mildura on 6 June 2023 (see ATSB investigation report AO-2023-025), QantasLink issued a technical advisory bulletin, effective from 17 July 2024, which changed the VHF communications procedure for Mildura departures. The aim of the change was to improve ground-to-ground CTAF VHF communication during the taxi phase. However, this has only been adopted at Mildura Airport through a route manual amendment (now company port supplement) and does not provide any effective level of mitigation to this known risk for other non-controlled aerodromes. 

Related occurrences

ATSB investigation (AO-2023-025)

On 6 June 2023, a Piper PA-28-161, taxied for runway 36 at Mildura Airport, Victoria. At about the same time, a QantasLink De Havilland Aircraft of Canada Limited DHC-8-315 (Dash 8) began to taxi for runway 09. Both aircraft broadcast their intentions on the local common traffic advisory frequency. The pilot of the PA-28 was aware of the Dash 8, but the crew of the Dash 8 were not aware of the PA-28. Both aircraft commenced their take‑off at about the same time and the Dash 8 crossed ahead of the PA-28 at the runway intersection of 09/36 by about 600 m. 

The pilot of the PA-28 was unable to visually sight the location of the Dash 8 due to airport buildings and assumed that the Dash 8 was still backtracking on runway 09. They did not directly contact the Dash 8 to positively organise separation. They also incorrectly referred to the runway direction at Mildura Airport as ‘runway 35’ instead of ‘runway 36’. 

The Dash 8 crew was focused on obtaining their pre-departure information from air traffic control and had the volume for the radio tuned to the common traffic advisory frequency turned down. An over transmission from air traffic control meant that the Dash 8 crew only received certain elements of the PA-28 pilot’s radio calls. This likely led to an incomplete comprehension of traffic by the Dash 8 crew who believed that the PA-28 was not at Mildura (due to the incorrect reference to runway 35). However, they did not seek further information of the source of the radio calls to positively identify the traffic location.

The investigation found that, due to the topography and buildings at Mildura Airport, aircraft are not directly visible to each other on the threshold of runways 09, 27 and 36. The Dash 8 crew did not give a rolling call on runway 09, nor were they required to. The lack of a requirement for mandatory rolling calls increased the risk of aircraft not being aware of each other immediately prior to take-off.

ATSB investigation (AO-2023-050)

On 29 September 2023, De Havilland Aircraft of Canada Limited DHC-8-315 (Dash 8), taxied for runway 09 at Mildura, Victoria. A short time later, an amateur‑built Lancair Super ES aircraft taxied for runway 36 at Mildura, for a private flight to Ballarat. 

Both aircraft gave taxi, entering and backtracking calls on the local common traffic advisory frequency. Neither the pilot of the Lancair, nor the crew of the Dash 8, were aware of each other. The crew of the Dash 8 gave a rolling call and had commenced their take-off on runway 09 as the pilot of the Lancair gave a rolling call on runway 36, this was received by the Dash 8 crew with an immediate response given to the Lancair to hold on the runway. Another aircraft, taxiing behind the Lancair for runway 36, advised them to hold position while the Dash 8 departed.

The Dash 8 crossed the runway 09/36 intersection while the Lancair remained on the threshold of runway 36.

The investigation found that, due to the topography and buildings at Mildura Airport, aircraft are not directly visible to each other on the threshold of runways 09, 27 and 36. It was also identified that the crew of the Dash 8 were actively engaged in in organising separation with other airborne traffic and the Lancair’s entering and backtracking call was over transmitted. 

It was also identified that the Dash 8 had reduced ground-based radio reception and transmission strength and clarity on VHF COM 2 (which used an antenna on the aircraft underbelly) and was required to be used by company procedures. The investigation also found that the Dash 8 had reduced radio reception and transmission strength to and from other airfield users located behind the Dash 8 which affected radio call readability.

This led to a situation where the crew of the Dash 8 had an incomplete comprehension of the ground-based traffic at Mildura, and had no knowledge of the Lancair until during the take-off. In addition, due to the position and distance of the Dash 8, the pilot of the Lancair had no awareness of the Dash 8 until another aircraft advised that the Dash 8 was rolling on runway 09.

ATSB investigation (AO-2024-009)

On 19 March 2024, a Fairchild SA226-TC (Metroliner), taxied at Geraldton, Western Australia, for runway 03. About one minute later, a Beechcraft A36 (Bonanza), taxied for runway 14. After reaching their respective runway thresholds, both pilots attempted to contact the other, however, they did not hear each other, nor could they see each other. A third aircraft assisted by relaying information. Based on the information received, the Bonanza and Metroliner pilots commenced their take-off within 3 seconds of each other. The Metroliner crossed runway 14 about 400 m in front of the Bonanza, with a vertical separation of about 250–300 ft. 

The investigation found that, when aircraft were positioned at the thresholds of runway 03 and 14 (and 08), they will unlikely be visible to each other due to the position of the airport buildings. Further, they may not be contactable on VHF radio due to potential shielding effects. This resulted in the pilots being unable to verify each other’s position and intentions prior to commencing their take-off. 

While the pilot of the third aircraft was attempting to assist, the details provided were inaccurate and incomplete. This inadvertently resulted in misinterpretation by the Bonanza and Metroliner pilots and influenced their decision to take off.

Safety analysis

Introduction

On 15 July 2024, a QantasLink De Havilland Aircraft of Canada Limited DHC-8-402, registered VH-QOD, entered runway 05 at Wagga Wagga Airport, unaware that a Piper PA-28, registered VH-XDK, was commencing its take-off roll from runway 23. 

While the crew of the Dash 8 was entering the runway, the crew of a third aircraft (Saab 340), notified the Dash 8 crew of the PA-28’s presence. 

Once the Dash 8 crew was aware of the PA-28, the captain stopped the aircraft about 5 m past the holding point, then engaged ‘reverse thrust’ to back the Dash 8 clear of the runway. Concurrently, the pilot of the PA-28 rejected their take-off from runway 23. 

This analysis will explore operational considerations such as situational awareness, the breakdown of communication, shared mental model and alerted see‑and‑avoid, as well as the limitations identified within operational procedures and specific aircraft system limitations.

Situational awareness and communication 

The PA-28 crew was aware of the Dash 8 intentions and location. The PA-28 instructor subsequently held the assumption that the Dash 8 crew had, likewise, received their radio broadcasts and were aware of the PA-28’s location and intentions for take-off. 

The Dash 8 crew did not expect any other traffic, in addition to the Saab 340. The Dash 8 crew reported not identifying the PA-28 either visually or via electronic means, nor hearing the PA-28’s radio communications. During taxi to the opposite end of the runway, the Dash 8 was pointing in the opposite direction to the PA-28, and later the line of sight to the PA-28 would have been obscured by the landing Saab 340. As the PA-28 transponder was not turned on prior to entering the runway, it was not identifiable electronically either prior to this time. The remaining opportunity to increase situational awareness was the initial taxiing radio calls. The Dash 8 crew’s broadcast immediately after they were told about the conflicting traffic shows they were surprised to learn there was another aircraft operating from the airport, this suggests that the Dash 8 crew members did not hear these transmissions rather than forgot them.

Communication

Succinct and timely radio communication is important to ensure high levels of situational awareness and aids in providing alerted see-and-avoid safety outcomes. As such, the accuracy of the information broadcast by pilots is critical in ensuring minimum misunderstanding.

Common traffic advisory frequency (CTAF) recordings indicated that the PA-28 pilot made the recommended positional broadcasts in preparation for their departure – a taxi call when leaving the apron area, entering the runway and then a rolling call on the runway. However, positional broadcasts are ‘one-way’ communications by nature and do not explicitly mean that other traffic in the vicinity interpret and understand the intended information. 

The PA-28 instructor reported that they believed that the Dash 8 crew had received and understood their broadcasts. This misunderstanding was based on the presumption that because the PA-28 could hear the radio calls and see the Dash 8, that the Dash 8 crew also could hear the radio calls and see the PA-28.

The PA-28 crew could hear the communications between the Dash 8 and Saab 340, centring on the Dash 8 crew ensuring they did not impede the Saab 340’s landing. However, the PA-28 instructor expected that the Dash 8 would hold short of the runway to allow them to take-off without hearing any broadcasts to that effect. As a result, the PA-28 crew did not initiate any direct radio contact with the Dash 8 crew to clarify their intentions. 

Had the PA-28 instructor attempted to make direct contact with the Dash 8, and not been successful, they would have subsequently realised there had been a communication breakdown. 

Visual identification 

The topography at the aerodrome results in a runway height change between the thresholds of runway 05/23 at Wagga Wagga Airport. This change is up to 2 m in elevation between the thresholds, with a slight hump towards the runway 05 end. However, there was a clear visual line of sight from one threshold to the other.

When the Dash 8 was on taxiway B at the holding point of the runway, the crew’s vision of the PA-28 may have been blocked by the landing Saab 340. By the time that aircraft cleared the runway, the PA-28 was lined up on the runway and had started its take-off roll. 

Based on the limitations of vision, it is very unlikely that the Dash 8 crew would have been able to detect the presence of the PA-28 about 1,700 m away, simply by looking in that direction, as the image of that aircraft would have been too small for detection. The use of landing and strobe lights would have assisted, but over that distance would have been limited. Consequently, the Dash 8 crew, who reported looking before entering the runway, did not detect its presence.

If the Dash 8 crew had previously been alerted to the presence of another aircraft, this would have increased the chance of detection. Alerted see‑and‑avoid means that the presence and approximate location of another aircraft is known or expected, allowing the crew to narrow their visual search. However, in this case, the absence of the Dash 8 crew detecting radio calls from the PA-28 and no other aircraft identified during taxi either visually or electronically, meant that the Dash 8 crew held an expectation that there were no other ground-based aircraft operating at the airport at that time. Not detecting the PA‑28 was consistent with that expectation.

The ATSB identified that while the PA-28 crew conducted their taxi, engine run-ups and aircraft checks, their transponder was set to standby. Prior to entering the runway for take-off, the pilot of the PA-28 then switched the transponder from standby to ‘ON’. 

The Dash 8 crew reported that electronic surveillance equipment was used as an aid to supplement the identification of potential conflicting traffic in the vicinity of a non‑controlled aerodrome prior to take-off. Both the captain and first officer of the Dash 8 recalled conducting a check of their onboard surveillance systems (electronic flight bag, traffic collision avoidance system (TCAS) aircraft display and Flightradar24 application) during taxi and did not identify any traffic other than the third-party landing aircraft.

Non-active mode S transponders do not provide electronic surveillance information to other aircraft. If the transponder is not selected ‘on’, a missed opportunity exists to provide situational awareness to other aircraft. 

It is likely, had the PA-28 mode S transponder been activated at the first aircraft movement for taxi, the flight crew of the Dash 8 would have detected the presence of the PA-28 and developed an expectation of its location and potential confliction. This would have prompted the crew to coordinate both verbally and aided alerted see‑and‑avoid.

Australian Airline Pilot Academy procedures for transponder

The AAPA’s manual had a procedure that prescribed the use of the transponder on standby during startup and taxi, with a requirement to only switch it to ‘ALT’ prior to entering the runway. 

Widespread use of surveillance equipment, such as transponders with ADS-B installed, offer significant improvement to allow pilots to be more certain of the location of traffic, particularly outside controlled airspace environments. 

Had the PA-28 crew been required to select their mode S transponder to ‘ON/ALT’ prior to initial taxi, this would have made them electronically visible to the Dash 8 crew who reported actively seeking this information. 

Topographic shielding 

Local operators at Wagga Wagga Airport have observed and reported that a reduction in effectiveness of VHF communications had been experienced at certain locations on the airport. This observation of topographical radio shielding was identified by multiple parties. 

Topographic changes exist in and around Wagga Wagga Airport and were suggested by stakeholders to result in varying levels of topographical radio shielding at the eastern end of taxiway A and the end of runway 23. 

Prior to the occurrence, the PA-28 operator issued internal notifications of an ‘identified radio dead spot’, to its flight crews.  

Operator advice and reported experience regarding topographical shielding at Wagga Wagga Airport was documented procedurally for AAPA, but was not available to Dash 8 crew operating into Wagga Wagga Airport. The ATSB did not independently confirm the existence of radio shielding. Operator experience and local procedure indicates this is likely an ongoing issue at Wagga Wagga Airport.

Dash 8 radio reception and transmission

VHF radio transmissions facilitate exchanges between air traffic control, aircraft, and emergency services. These signals primarily rely on line of sight propagation, requiring an unobstructed path between the transmitting and receiving antennas. When line of sight is obstructed by terrain features, aircraft structure, or by man-made structures like buildings, VHF signals can be significantly attenuated, leading to reduced communication range, signal distortion, or complete loss of contact. However, under certain conditions, VHF signals may propagate beyond line of sight through mechanisms such as reflection and diffraction. While these phenomena can marginally extend communication range, their effectiveness is highly dependent on the environment.

Previous analysis of Dash 8 ground-based communications (AO-2023-050) identified a significant reduction in radio signal strength and readability when using the VHF COM 2 antenna for ground operations. The analysis determined that the aircraft structure and surrounding airport infrastructure at Mildura Airport impeded VHF COM 2 signal transmission and reception, limiting its effectiveness. In contrast, using the VHF COM 1 antenna provided a clearer line of sight to ground-based stations, resulting in more reliable communication.

ATSB analysis of the radio communications between the Dash 8, the Saab 340, and the PA-28 identified that the Dash 8 crew, while using VHF COM 2, could hear and respond to the Saab 340, likely due to its proximity to the aircraft. However, they could not hear the PA-28. 

Their inability to hear the PA-28 was likely due to previously identified issues with the VHF COM 2 antenna and the addition of runway undulations which further obstructed the line of sight between the Dash 8’s VHF COM 2 antenna and the PA-28. This likely exacerbated the antenna’s limitations and hindered effective signal propagation.

Qantas procedures for VHF

QantasLink procedures required the use of VHF COM 2 for ground-based communication at non-controlled aerodromes. However, the use of VHF COM 2 for ground-based communications was not required for controlled airspace, where ground‑based communications were conducted by VHF COM 1. 

In previous ground-based signal strength testing at Mildura Airport (see ATSB investigation report AO-2023-050), the ATSB identified a significant reduction in the Dash 8 VHF radio transmission strength and readability particularly when greater transmission and reception distances were involved. It was also found that the use of VHF COM 2 reduced ground-based transmission reception strength and clarity in comparison with VHF COM 1. 

Following a near collision occurrence at Mildura on 6 June 2023, QantasLink changed the VHF communications procedure for Mildura Airport departures. The changed procedure required use of the VHF COM 1 system, noting that this was found to have an improvement in both transmission clarity and reception. However, this has only been adopted at Mildura Airport and so does not provide any effective level of mitigation to this known risk for other non-controlled aerodromes.

Given the reported radio dead spot at Wagga Wagga Airport, the use of VHF COM 2 with the bottom‑mounted antenna by QantasLink Dash 8s likely increased the likelihood of Dash 8 crews not receiving strong and readable radio calls from other ground-based aerodrome users.

Part of the shared mental model for all operations at non-controlled aerodromes is the situational awareness of all crew, in this instance the crew of the Saab 340 identified a potential threat to the safety of the other 2 aircraft and clearly broadcasted their concern.

Crew interviews as well as CTAF recordings clearly showed that the proactive initiative of the Saab 340 crew to alert the Dash 8 crew of the presence of the PA-28 was instrumental in preventing a potential on runway collision.

After entering the preferred runway, the captain became aware of the hazard present due to the vacating third party aircraft report. The captain’s decision to apply brakes and then engage reverse thrust to clear the runway was almost certainly made to avoid the greatest perceived threat, that being, a possible collision on the runway. 

Reversing the aircraft under its own power is a non-normal manoeuvre with additional risks. As such QantasLink policy would normally require Head of Flight Operations approval to conduct the manoeuvre. 

The captain’s decision to reverse the aircraft could be attributed to the lower (perceived) risk involved with reversing the aircraft, versus the known risk of a possible collision.

The captain exercised their command authority, based on the information available, to make an assessment of the greatest threat, subsequently deciding to reverse the aircraft, without being able to positively confirm traffic and obstacle avoidance. 

Risk management focuses on reducing the potential risks associated with a decision. Thus, considering that decision‑making, when elements of uncertainty exist, focuses on making the best decision given the available information.

The pilot in command of an aircraft has the final authority over the safety of the aircraft and its occupants. When operating in dynamic environments, the pilot in command will need to consider possible outcomes and alternative courses of action to ensure clear risk-based decisions are made. In this instance the captain elected to reduce the likelihood of a catastrophic event by reversing clear of the preferred runway after having a reasonable expectation that the taxiway was clear behind the Dash 8.

Findings

From the evidence available, the following findings are made with respect to the runway incursion involving De Havilland Aircraft of Canada Limited DHC-8, VH-QOD, and Piper PA‑28, VH-XDK, at Wagga Wagga Airport, New South Wales, on 15 July 2024. 

Contributing factors

• During ground-based repositioning, the Dash 8 entered runway 05 while the PA-28 had commenced its take-off roll from runway 23. The Dash 8 crew were not aware of the PA‑28 until notified by the crew of a third aircraft of the potential conflict.
• The pilots of the PA-28 gave the recommended radio calls, however, did not directly communicate or engage with Dash 8 crew to arrange separation.
• Without any prior alert or expectation of the presence of the PA-28, the Dash 8 crew did not visually detect the PA-28 on take-off from the reciprocal end of the runway, prior to the Dash 8 entering the runway to taxi to the terminal.
• During taxi for take-off, the PA-28 was not broadcasting transponder information to identify them to other traffic in the vicinity of the aerodrome.
• The Australian Airline Pilot Academy flying school flight crew operating manual only required pilots to activate the transponder prior to entering the runway. The use of a transponder during taxi would normally provide an additional source of positional data to other pilots, aiding visual identification and alerted 'see‑and‑avoid' to other aircraft. (Safety issue)
• Reduced VHF ground-based communication was identified at the eastern end of runway 23 at Wagga Wagga Airport. Local operator reports and procedures indicate likely reduced communications on the eastern end of taxiway A to other areas on the aerodrome.
• Dash 8 ground-based transmissions on VHF COM 2 had reduced strength and clarity. This likely led to situations where other aircraft had difficulty in receiving and understanding radio transmissions, and the Dash 8 not receiving the PA-28 radio transmissions.
• QantasLink's radio procedure required crew to use communications panel radio 2 (COM 2) to broadcast and receive on local frequencies during operations at a non-controlled aerodrome. This reduced the likelihood of the Dash 8 receiving the calls from other aircraft at either end of runway 05/23 at Wagga Wagga Airport in certain circumstances. (Safety issue)

Other (key) findings

• Third party intervention by the Saab 340 crew prevented the Dash 8 from lining up on runway 05 whilst the PA-28 was engaged in the take-off roll.
• On assessing that a collision risk existed with the rolling PA-28, the captain held a reasonable expectation that it was clear and reversed the aircraft away from the preferred runway without being able to confirm that no other aircraft were behind the Dash 8.

Safety issues and actions

AAPA procedures for transponder usage

Safety issue number: AO-2024-041-SI-01

Safety issue description: The Australian Airline Pilot Academy flying school flight crew operation manual only required pilots to select ALT on the transponder, as part of the Pre Line Up Scan Action Flow and associated Checklist prior to entering the runway. The use of a transponder during taxi would normally provide an additional source of positional data to other pilots, aiding visual identification and alerted 'see‑and‑avoid' to other aircraft.

Safety advisory notice to pilots and operators of mode S transponder equipped aircraft

SAN number: AO-2024-041-SAN-01

The effective use of the mode S transponder from first movement of the aircraft can serve as an effective tool in adding another layer of collision avoidance between aircraft on the ground. Utilising the mode S transponder with ADS-B OUT enabled is the most effective way of making an aircraft electronically conspicuous and delivering maximum interoperability with other aircraft as well as the ground ATM environment. The ATSB advises pilots and operators to review their procedures to ensure that mode S transponders are on from first movement of the aircraft, particularly at non-controlled aerodromes.

QantasLink radio procedure

Safety issue number: AO-2024-041-SI-02

Safety issue description: QantasLink's radio procedure required crew to use communications panel radio 2 (COM 2) to broadcast and receive on local frequencies during operations at a non‑controlled aerodrome. This reduced the likelihood of the Dash 8 receiving the calls from other aircraft at either end of runway 05/23 at Wagga Wagga in certain circumstances.

Safety action not associated with an identified safety issue

Safety action by Wagga Wagga Airport 

After the occurrence, the aerodrome operator identified that it was possible that a radio black spot may be present based on observations of radio reception just beyond the threshold of runway 23. 

Wagga Wagga city council has issued a precautionary NOTAM after the occurrence which provides advice to all operators of possible visibility issues with aircraft on the threshold of runway 23 and the possibility of radio black spots existing under some conditions. The NOTAM note reads:

LIGHT ACFT AT THR OF RWY 23 NOT VISIBLE TO OTHER ACFT USING RWY 05. 

RADIO BLACK SPOTS MAY RESULT UNDER SOME CONDITIONS.  

Safety action by QantasLink

On 2 August 2024 QantasLink safety issued a company safety alert notice advising crew of a suspected VHF radio blackspot on taxiway A4 and A5 towards the threshold of runway 23 and other areas of the Wagga Wagga aerodrome. They issued the following advice to crew:

Whilst we investigate the issue with the aerodrome operator and other aerodrome users, we encourage crew to consider the following precautions when taxiing for departure or any other manoeuvring on the ground: 

- ensure any weak, garbled or carrier-wave only CTAF transmissions are clarified to ensure your separation plan remains valid; -

- stop at runway holding points and visually check for traffic prior to runway entry; -

- where possible all aircraft to use the into wind (duty) runway for departure so aircraft are not departing from opposite ends of the runway; and –

- ensure AFRU responses on the CTAF frequency are heard clearly. 

Safety action by Civil Aviation Safety Authority 

On 9 September 2025, CASA advised that new guidance and recommendations relating to the use of radios where the antenna is mounted on the underside of the aircraft fuselage to both Advisory Circular (AC) 91-10 and AC 91-14 had been updated. 

This guidance advised: 

Pilots operating aircraft with similar antenna placements are reminded that ground-based transmissions, when made using a radio with a fuselage underside antenna, are likely to have an increased risk of not being reliably received by other traffic. To enhance situational awareness and collision avoidance, especially at non-controlled aerodromes where radio-alerted see‑and‑avoid is critical, pilots are strongly recommended, wherever practicable, to use radios connected to antennas in unobstructed locations, such as an aircraft upper fuselage, for ground communications

Glossary

Sources and submissions

Sources of information

The sources of information during the investigation included: 

• the instructor of VH-XDK
• the crew of VH-QOD
• QantasLink
• Australian Airline Pilot Academy
• Civil Aviation Safety Authority
• Wagga Wagga Airport
• Aviation Bureau de la sécurité des transports du Canada
• De Havilland Aircraft of Canada.

References

ATSB. (2025). Cockpit Visibility Study. Supporting AO-2023-001 – Midair collision involving Eurocopter EC130 B4, . Canberra: Australian Transport Safety Bureau.

Bailey, L. L., & Thompson , R. C. (2000). The effects of performance feedback on air traffic control team coordination: A simulation study. United States: Department of Tranportation. Federal Aviation Administration. Office of Aviation. Civil Aerospace Medical Institute.

Civil Aviation Safety Authority. (2013, December). PIlot's responsibility for collision avoidance in the vicinity of non-controlled aerodromes using 'see-and-avoid'. Canberra, ACT, Australia.

Civil Aviation Safety Authority. (2014). SMS 3: Safety Risk Management: SMS for Aviation, A Practical Guide. Canberra, ACT, Australia.

Federal Aviation Administration. (2011). Introduction to TCAS II Version 7.1. U.S. Department of Transportation.

Federal Aviation Administration. (2024). AC 91-120. Operational Use of Airborne Collision Avoidance Systems. U.S. Department of Transportation.

Hobbs, A. (2004). Limitations of the see-and-avoid principle. Canberra, Australia: Australian Transport Safety Bureau.

International Telecommunications Union (ITU-R). (2019). Recommendation ITU-R P.526-15: Propagation by diffraction. Geneva: ITU.

Nguyen, T., Lim, C., Nguyen, N., Gorden-Brown, L., & Nahavandi, S. (2019). A review of situation awareness assessment approaches in aviation environments. IEEE Systems Journal, 3590-3603.

Parsons, J. D. (2000). The mobile radio proagation channel (2nd ed.). Wiley.

Rappaport, T. S. (2002). Wireless Communications: Principles and Practice (2nd ed). Cambridge University Press.

Reynolds , R., & Blickensderfer, E. (2009). Crew Resource managment and shared mental models: A proposal. Journal of Aviation/Aerospace Education & Research, 15‑23.

Stanton , N., Salmon, P., Walker, G., Salas, E., & Hancock, P. (2017). State of science: situation awareness in individuals, teams and systems. . Ergonomics, 449-466.

Wickens , C., & McCarley, J. (2008). Applied attention theory. Boca Raton FL : CRC Press.

Wickens, C., Hollands , J., Banbury, S., & Parasuraman , R. (2013). Engineering psychology and human performance. Person Bostan, MA

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 Civil Aviation Safety Authority
• the Australian Airline Pilot Academy
• QantasLink
• Wagga Wagga Airport
• Aviation Bureau de la sécurité des transports du Canada
• De Havilland Aircraft of Canada Limited
• the pilot of PA-28
• the crew of Dash 8.

Submissions were received from:

• the Civil Aviation Safety Authority
• the Australian Airline Pilot Academy
• QantasLink.

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

Appendices

Appendix A – CTAF recordings

Appendix B – Flight Operations Service Letters

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 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.

Investigation summary

What happened

On the morning of 25 July 2024, 4 Robinson R22 helicopters were involved in a mustering operation at Mount Anderson Station, located about 51 km south-south-east of Curtin Airport, Western Australia. 

The group planned to depart and travel together to the proposed muster location, about a 10-minute flight away. The lead pilot and another experienced mustering pilot departed first. Shortly after take-off, at about 150 ft, the helicopters departed controlled flight, collided with terrain and were destroyed. Both pilots were fatally injured.

What the ATSB found

The ATSB found that during the initial climb, while flying in close proximity, the lead helicopter manoeuvred to the right, and neither pilot detected their converging flight paths, which resulted in the midair collision. The ATSB found that the tools used by the operator to consider and manage operational risk were not tailored to their business. In the context of their primary flying activity of aerial mustering involving multiple helicopters, the risk of collision had not been identified in operational risk assessments, and the operator’s manuals did not provide documented procedures to ensure pilots establish and maintain adequate separation between helicopters. 

The investigation also identified that company pilots were permitted to arrange their own separation based on personal preference. Pilots routinely flew with reduced vertical and lateral separation, and over time this became an accepted operating practice. Consistent with this practice, the pilots flew in proximity on the day of the accident. 

While it did not contribute to the accident, the topics discussed in the morning brief were operational in nature and did not include elements of aviation safety risk such as separation. This represented a missed opportunity to establish a shared understanding of how the helicopters would be coordinated, which may have provided the pilots with a better understanding of what to expect, and how other pilots would be operating their helicopters.

What has been done as a result

Following the accident, the operator conducted a review of the operations manual and defined procedures for operations involving multiple aircraft and the conduct of operations in the vicinity of other aircraft. These changes were contained in a company notice to air crew (NOTAC) and all pilots were briefed on the changes prior to resuming flying operations. Additionally, a specific risk assessment for mustering operations was generated that included multi‑helicopter operations and considered the risk of collision. 

Safety message

The regulations that govern CASR Part 138 aerial work activities are intended to provide operators with flexibility through scalable risk management practices. 

For risk management to be effective, it must identify, assess and mitigate actual threats to a flying activity. Without consideration of the actual work being conducted, including commonplace activities, it is more likely that operational risks will not be uncovered, and risk management opportunities will be missed. 

The ATSB encourages operators to review available guidance to assist with identification and management of hazards. For mustering operators, CASA provides a sample manual that includes sample acceptable procedures. Other resources such as the Flight Safety Foundation’s Basic Aviation Risk Standard for Aerial Mustering are a valuable resource to assist operator’s in coordinating aviation activities to manage and understand the aviation risk to their operation.

This occurrence also reinforces the fallibility of the see‑and‑avoid principle as the primary tool for identifying and managing the threat of collision. Defined separation minimums and pre-planned safe exits designed to provide opportunity to identify and respond to emerging collision threats would enhance the pilot’s ability to detect a conflict and recover from an unsafe condition. Additionally, airframe obstructions that limit the application of the see‑and‑avoid principle, in even the most open aircraft cabins, should be a consideration when establishing how aircraft should be positioned when flying in close proximity.

The occurrence

On the morning of 25 July 2024, 4 Robinson Helicopter Company R22 Beta II helicopters were being operated at Mount Anderson Station, located about 51 km south-south-east of Curtin Airport, Western Australia (Figure 1). The helicopters were operated by Pearl Coast Helicopters and were being used as part of a cattle mustering operation, which also included multiple land vehicles.

Figure 1: Accident location

Source: Google Earth, annotated by the ATSB

At about 0515 local time, a briefing was held by the lead pilot to cover the planned muster along a section of the Fitzroy River, a 10-minute flight to the west-north-west of the station homestead (Figure 2). This briefing was attended by the helicopter pilots and the head stockman, who provided input on where the ground crew would be located. The stated plan was for the helicopters to take-off one after the other and travel together to the muster location. Once there, the lead pilot would show the other pilots various locations before each pilot would peel off and work their section. 

Figure 2: Planned muster location

Source: Google Earth, annotated by the ATSB

At about 0540, the pilots commenced their pre-flight preparations. VH-HQH (Heli 1) and VH-HYQ (Heli 2) were parked in a grassed clearing behind the homestead (Figure 3). The other 2 helicopters, VH‑HMR (Heli 3) and VH-YKC (Heli 4), were parked about 400 m to the west. Heli 3 was parked on a small mound about 3 m high, and Heli 4 was parked about 20 m to the north-east at the base of the mound. Both aircraft were parked facing east towards the homestead. 

The pilots started the helicopters just before 0600, and established communication with each other on a discreet company VHF radio frequency. The lead pilot in Heli 1 radioed they were ready to depart and would lead the group out. This radio broadcast was acknowledged by the pilots of Heli 2 and Heli 3. The pilot of Heli 2 replied that they were also ready and would follow the lead pilot. The pilot of Heli 4 did not have their flight helmet on at this time and did not hear these radio calls.

At 0603, Heli 1 (lead helicopter) and Heli 2 (following helicopter) departed the overnight parking area and flew west. The track to the planned muster took them past the location where the other helicopters were parked (Figure 3). The intention was for the remaining helicopters to join them as they flew past and the group would fly together to the muster. 

A witness (witness 3) located near the stockyards about 200 m to the south of the homestead watched the 2 helicopters depart, climbing above the buildings with the following helicopter slightly lower and off to one side. They recalled that the helicopters were already established on the departure track, travelling about the same speed, and appeared to be close to one another when they first saw them.

Figure 3: Witness locations

Source: Google Earth, annotated by the ATSB

The pilot of Heli 3 observed Heli 1 and Heli 2 flying together off the right of their helicopter’s nose. They recalled that the helicopters were above the tree line to the east, slightly north of the camp kitchen at about 100 ft, moving from left to right in their windscreen and would have passed about 80 m south of where Heli 3 was parked. One helicopter appeared higher than the other, with the following helicopter slightly behind and off to one side. The pilot of Heli 3 recalled that they were probably closer than 100 m apart at this point, but it was difficult to be sure because the helicopters were silhouetted against the bright sky. 

The pilot of Heli 4 was a few minutes behind in their own pre-flight preparations. The pilot recalled seeing the 2 helicopters approach from the east, but did not see a risk of collision at that point. The pilot of Heli 4 put on their helmet and established radio contact with the group. The pilot in nearby Heli 3 acknowledged the radio check and broadcast that they would lift into a low hover.

Both pilots were still on the ground and heard the pilot of Heli 1, the lead helicopter, acknowledge this radio transmission with the reply “Yep, I see you there”. The pilot in Heli 3 lifted into a low hover, focusing on the manoeuvre, and the pilot of Heli 4 turned their attention to their own helicopter before they heard Heli 1 and Heli 2 collide.

Witness 3 recalled that when the helicopters collided, they were partially obscured behind a tree but recalled that neither helicopter appeared to manoeuvre prior to colliding. After hearing the collision, the pilot of Heli 3 looked up and saw that Heli 1 was descending toward their location. Already established in the hover, the pilot manoeuvred the helicopter away to avoid being struck. The pilots of Heli 3 and Heli 4 advised that after the collision, Heli 1 maintained a generally straight track over the ground as it descended. Witnesses were unable to recall any details regarding the movement of Heli 2 following the accident. Several witnesses including the other helicopter pilots reported seeing smoke or flames as the helicopters descended.

As a result of the midair collision, both helicopters were rendered uncontrollable, collided with terrain and were destroyed. Both pilots were fatally injured.

Context

Pilot information

Both pilots involved in the collision were qualified and authorised to conduct aerial mustering operations.

Pilot of VH-HQH

The pilot of VH-HQH (Heli 1) held a commercial pilot licence (helicopter) with a low‑level endorsement and an aerial mustering endorsement. They had 6 years of aerial mustering operation experience and had been with the operator for 4 years. Their total flight time was 4,750 hours, with about 3,600 flight hours mustering, and about 4,250 flight hours in R22 helicopters. Their most recent flight review was a low-level single engine helicopter flight review on 1 April 2024.

The pilot held a class 1 aviation medical certificate that was valid until 20 March 2025. There were no medical conditions listed on their licence. A post-mortem examination identified no significant pre-existing medical conditions. 

Pilot of VH-HYQ

The pilot of VH-HYQ (Heli 2) held a commercial pilot licence (helicopter), with a low-level endorsement and an aerial mustering endorsement. They had 4 years of aerial mustering operation experience and had been with the operator for 2 years. Their total flight time was 2,100 hours, with about 1,350 flight hours mustering, and about 2,000 flight hours in R22 helicopters. They had conducted a low-level single engine helicopter flight review on 30 March 2024.

The pilot held a class 1 aviation medical certificate that was valid until 21 February 2025. There were no medical conditions listed on their licence and a post-mortem examination identified no significant pre-existing medical conditions. 

Aircraft information

The Robinson Helicopter Company R22 Beta II helicopter is powered by a 4‑cylinder carburetted piston engine. The R22 has 2 seats, with the pilot flying from the right seat, and each seat was fitted with a 3-point seatbelt and inertia reel shoulder strap, similar to those used in motor vehicles. The 2-blade main rotor assembly rotates counterclockwise when viewed from above. 

Fitment of doors is optional (Figure 4) and it is normal for mustering operations to operate without doors. On the day of the accident both helicopters were operated without doors.

Figure 4: Exemplar R22 without doors

Source: ATSB

Both helicopters had a high visibility paint scheme on the upper surface of the main rotor blades (Figure 5).

Figure 5: High visibility main rotor blades

Source: ATSB

VH-HQH (Heli 1)

VH-HQH, serial number 3936 (Figure 6), was manufactured in the United States in 2005 and first registered in Australia in September 2005. The helicopter had undergone a periodic inspection along with a 2,200 hr / 12-year inspection and overhaul that was completed on 8 July 2024 at 6,087 hours total time in service. The maintenance release current at the time of the accident was not located, however, the operator reported the helicopter had accrued about 60 hours since the rebuild.

Figure 6: VH-HQH (left) and VH-HYQ (right)

Source: Helico

During this recent overhaul, HQH was fitted with several optional items including a carbon fibre console cover, carbon fibre battery box, and a smoker unit to aid in aerial mustering operations (Figure 7).[1] The smoker consisted of a small tank (about 5 L capacity), that was filled with either diesel or paraffin oil and mounted on the right-side lower frame.

Figure 7: VH-HQH smoker unit installation

Source: Helico, annotated by the ATSB

VH-HYQ (Heli 2)

VH-HYQ (HYQ), serial number 3835 (Figure 6), was manufactured in the US in 2005 and first registered in Australia in May 2005. A 2,200 hr / 12-year inspection/overhaul was completed in April 2023. The most recent periodic inspection, at 4,946 hours, was conducted on 19 July 2024. The current maintenance release was not located, however, the operator reported the helicopter had accrued about 40 hours since the last periodic inspection.

HYQ was not fitted with a smoker unit.

Accident site information

Overview

The accident site was located about 400 m west-north-west of the Mount Anderson homestead at an elevation of about 42 m above mean sea level (AMSL). The small hill beyond the site had an elevation of about 80 m. Power to the property was produced via generator. There were some low powerlines immediately above the buildings, but the powerlines did not extend along or across the flight path (Figure 8).

Figure 8: View from departure point towards the accident site

Source: ATSB 

The local vegetation was dry, low, open woodland and a flock of Cacatua sanguinea (Little Corella) were observed by the ATSB around the homestead during the onsite phase of the investigation. There was no evidence of any tree or wildlife strike in the vicinity of the start of the debris trail or among the wreckage.

Wreckage information

The first items in the debris field, about 300 m west-north-west of the take-off location, were consistent with a rotor blade strike to the right side of VH-HQH/Heli 1 (Figure 9), and consisted of:

• smoker unit pump cover panel
• smoker tank sump (separated horizontally from most of the upper portion of the tank)
• smoker unit lower mount
• fuselage skin from just forward of right position light
• right seat ‘pocket’ material section (located on forward panel of the pilot seat structure)
• trim from around the forward and right edges of pilot seat.

Figure 9: VH-HQH (pre-accident) showing approximate location of rotor strike and first items in debris field

Source: Helico and ATSB, annotated by the ATSB

The main debris trail was arranged in 2 key areas (Figure 10), in a north-westerly direction ending with the wreckage of Heli 1, and in a west-north-westerly direction, ending with the wreckage of Heli 2. The heaviest items of wreckage were found to have followed these general directions.

Kepert (1976) looked at wreckage trajectories to assist with fixed wing aircraft accident investigations. The research found that heavy items possess a high weight/surface area ratio and are little effected by wind drift with their trajectories depending primarily on the velocity of the aircraft at the time of break-up. With regard to heavy items including engines and batteries, the research identified the following:

Such items are generally recovered from points lying close to the line of the extended flight path of the aircraft. A mean line drawn through the ground impact points of heavy items will define the extended flight path of the aircraft with sufficient accuracy.

Kepert’s research must be considered in the context of a helicopter midair collision. While helicopters and their components can be subject to changes of trajectory when damage occurs while airborne, when considering heavier items of wreckage, the research provides a likely scenario for establishing flight path information. The flight path estimates for the accident are depicted as white dotted lines in Figure 10.

Figure 10: Debris field showing main areas of wreckage for each helicopter

Note: The image does not show all items of wreckage identified in the debris field. Source: Google Earth, annotated by the ATSB

Wreckage examination of both helicopters identified all the major components were present at the accident site with no evidence of an in‑flight break-up prior to the midair collision. Heli 1 impacted the ground about 50 m to the west of the start of the debris trail, with Heli 2 a further 5 m west, and about 30 m to the south of Heli 1. Various components separated from the helicopters during the accident sequence, with fracture surfaces consistent with overstress. 

Heli 1 was consumed by fire, contained within an approximate 6 m diameter. The fire damage to Heli 1 limited examination of the flight controls. However, though flight control tubes had been destroyed, security was established through observation that control tube rod ends were secured to identified controls and bellcranks. 

Heli 2’s main fuel tank bladder was destroyed by impact forces, however there was no fire. The auxiliary fuel tank assembly was ejected, likely at the second impact with the ground. Its outer structure was impact damaged, however, the bladder remained intact. Flight control continuity of Heli 2 was established. 

Samples of fuel taken from Heli 2’s auxiliary tank and gascolator were visually examined and noted to be clean and clear of visible debris. The samples were also tested for the presence of water, with nil indication. Further, all 4 helicopters had been using the same bulk fuel with no quality issues reported.

Loss of control following the midair collision

VH-HQH (Heli 1)

Distribution of tailcone and tail rotor driveshaft and control tube segments, along with yellow paint transfer, was consistent with multiple strikes from the helicopter’s own main rotor blades early in the accident sequence. Both teeter stops exhibited impact damage, from main rotor spindle contact with the mast,[2] coincident with extreme teetering of the main rotor assembly that resulted in the tailcone strike. Such excessive movement of the main rotor and loss of tail cone and tail rotor components would render the helicopter uncontrollable.

Heli 1 collided with terrain left side low and almost inverted about 50 m to the west of the start of the debris trail and came to rest a further 9 m west. The main rotor assembly and mast separated after colliding with terrain and were located adjacent to the ground scar.

VH-HYQ (Heli 2)

Examination of the wreckage identified signatures indicative of the engine providing power to a rapidly slowing main rotor assembly, including torsional distortion at mast fracture location, as would occur with Heli 2’s rotor contacting Heli 1. One main rotor pitch link was secured. The other main rotor pitch link had failed in overstress at the lower rod end. The pitch horn had also fractured due to overstress. Loss of continuity in either pitch link will lead to loss of control of the helicopter. The mast fractured, about mid length, with the main rotor assembly and mast fairing separating and coming to rest about 10 m southwest of the other helicopter’s main rotor. This was indicative of the mast separation occurring during or immediately following the contact with the other helicopter.

Heli 2 impacted terrain on its left side and slightly nose down, about 50 m to the west of the start of the debris trail, and about 30 m to the south of Heli 1. Heli 2 then came to rest about 10 m further west. 

Survival aspects

Onsite examination identified the storage compartment under both seats in VH-HYQ were filled to volumetric capacity.[3] Once items were removed from under the right seat, distortion to the seat structure was evident. While this reduction in energy absorption may have contributed to some of the pilot’s injuries, the extent to which it influenced those injuries could not be determined. 

Accident site inspection showed that both pilots were correctly restrained and wearing flight helmets at the time of the collision. However, the accident was determined to be unsurvivable. 

Recorded information

Neither helicopter was equipped with a flight data recorder, tracking device or onboard camera, nor were they required to be.

Onboard recording devices

Two damaged Garmin aera 660 portable aviation GPS units recovered from the accident helicopters were transported to the ATSB facilities for further examination.

The GPS unit from Heli 1 was largely intact but showed signs of moderate fire damage. The memory board was installed in a donor chassis and a successful board-level data recovery was conducted.

The ‘current.gpx’[4] file had data from the day of the accident with 3 track points recorded, however, these were found to have been recorded while Heli 1 was on the ground. Timing data indicated that this was likely within one minute of the accident occurring. 

Damage to the unit recovered from Heli 2 was consistent with something penetrating the device. As a result of that damage the memory chip was knocked from the board. The memory chip was not found, and no data could be recovered from the device.

Witness video

Video of livestock in a nearby yard was captured on a handheld device at the time of the accident. The video was 14.37 seconds in length and while it did not contain any footage of the helicopters, it did capture audio of the departure and the collision.

Spectral analysis of the audio determined the following:

• The ratios of the main rotor, tail rotor and piston firing frequencies throughout the recording are consistent with the helicopters operating at normal engine and rotor speeds prior to the collision.
• There are notable frequency changes from key helicopter components between 7 and 11.5 seconds in the audio recording, likely due to one of the helicopters starting to move away from the recording source.
• On the balance of this evidence, it is considered likely that prior to the collision, Heli 1 conducted a right turn while both helicopters were travelling at about 40 knots in a similar direction. After 3 seconds, Heli 1 likely stopped turning and continued on this path for another 1.5 seconds.
• Following the change in flightpath, the helicopters maintained this general orientation for about 1.5 seconds before high amplitude sounds and noises consistent with midair collision between the 2 aircraft were recorded.
• Following the collision, a higher pitched whirring sound was recorded, indicative of engine and/or drive components running in an overspeed condition.

Weather and environmental information

Weather

The site was remote and there were no official weather observations available. The nearest official data was obtained from Curtin Aerodrome located 51 km to the north‑west of Mount Anderson Station.

The aerodrome forecast (TAF)[5] issued at 0321 local for Curtin Aerodrome indicated CAVOK[6] conditions with a south‑east breeze (150°) around 8 kt (Figure 11). The TAF included a low chance of fog, however, this was removed in an updated forecast issued 15 minutes prior to the occurrence.

Figure 11: Terminal Area Forecast for nearby Curtin

Source: Airservices, annotated by the ATSB

The meteorological aerodrome report (METAR)[7] for Curtin Aerodrome reported wind from the south‑east (130°) at 7 kt, visibility greater than 10 km and no cloud detected. There was no rainfall recorded in the previous 24 hours (Figure 12).

Figure 12: Meteorological aerodrome report for nearby Curtin

Source: Airservices

While the graphical area forecast (GAF)[8] that covered Mount Anderson Station and was valid at the time of the accident included isolated fog, witnesses and video recorded on the morning indicated there was no fog at the station.

Ambient lighting conditions

The Civil Aviation Safety Authority[9] defines daytime as the period between the beginning of morning civil twilight to the end of evening civil twilight. Geoscience Australia[10] states that in morning civil twilight, large objects may be seen but no detail is discernible. 

Geoscience Australia recorded the first light at Mount Anderson Homestead at 0551 and sunrise at 0614. At the time of the accident (0603), it was civil twilight with the sun approximately 3° below the horizon. Witnesses to the accident recalled being able to see, however when looking east, objects were silhouetted against the light background and depth perception was diminished (Figure 13).

Figure 13: Witness video showing ambient lighting conditions

The image on the left is the view looking to the east in the direction of the rising sun. The image on the right is facing towards the south and shows the increased detail visible when not looking directly towards the east. Source: Witness video provided by Western Australia Police Force

It is likely that during twilight conditions, the greatest degradations will be in the ability to perceive detail and colour. This reflects physiological changes within the eye, where cone photoreceptors become more effective as light levels increase (Purves and others, 2001). Performance on visual tasks such as motion perception and the ability to detect larger environmental features such as the horizon are less affected. Liebowitz and Owens (1991) suggest:

During civil twilight visual guidance and spatial orientation functions may be carried out normally, but tasks for which contrast sensitivity, resolution, or reaction time are critical may be degraded or no longer possible.

Organisational information

General information

Pearl Coast Helicopters Pty Ltd (PCH) commenced operations in 2004 to provide aerial mustering services across the Kimberley region of Western Australia. At the time of the occurrence, it employed 6 pilots and operated a fleet of 8 Robinson R22 helicopters and 1 Robinson R44 helicopter.

PCH operated under a Civil Aviation Safety Regulations (CASR) Part 138 (Aerial work operations) aerial work certificate that authorised it to conduct dispensing, external load and task specialist activities, including aerial mustering. The PCH chief executive officer (CEO), who was also the head of flight operations (HOFO), reported that aerial mustering was the company’s primary business and accounted for about 95% of the flying conducted by the company. They added that approximately 70% of musters were flown with multiple helicopters and advised that the utilisation of multiple aircraft on a muster was consistent with industry practice.

Mustering operations

Aerial Mustering incorporates the use of aircraft, primarily helicopters,[11] to locate, direct and concentrate animals. It is conducted at low level (below 500 feet above ground level) and used to complement traditional animal management techniques using vehicles, horses and motorcycles. The operator reported that where large areas need to be covered, it is common practice to utilise multiple helicopters on a single muster. 

Data published by the Bureau of Infrastructure and Transport Research Economics available in the CASA agricultural flying sector safety risk profile[12] showed that aerial mustering operators conducted 145,500 hours of flying in 2023. The data showed that for the period between 2014 and 2023, the average number of aerial mustering hours flown per year was 149,700 hours. 

Transit to a muster 

The accident occurred shortly after take-off during the transit to the planned muster. With respect to the transit to the area that the mustering activity will take place, there are 2 CASR rule sets that could apply. Certain Part 138 provisions disapply general operating and flight rules when aerial work is being conducted. The relevant CASA definition of flight is outlined in section 3 Interpretation of the Civil Aviation Act 1988 (CA Act) as follows:

flight means:

(a) in the case of a heavier than air aircraft, the operation of the aircraft from the moment at which the aircraft first moves under its own power for the purpose of taking off until the moment at which it comes to rest after being airborne

If the entirety of the flight was just to position the aircraft and no aerial work took place, that would be covered by Part 91 - General operating and flight rules. If a helicopter travels to site and begins aerial work activity such as mustering without coming to rest, the Part 138 rule set will apply. 

Part 138 scalable risk management

The interactions of the general Part 91 rules and Part 138 are covered in CASA AC 138-01 v3.1 Part 138 core concepts. It contained the following explanation of how risk is considered when Part 138 alleviations are applied:

Recognising that the absence of certain Part 91 rules could result in a lower level of aviation safety assurance, Part 138 compensates by setting out a series of safety risk management processes that require an aerial work operator to manage the risks of their aerial work operations. The requirement to manage risk is fundamental to Part 138 and is achieved both directly within the legislation, for example, by specifying particular performance requirements for carriage of aerial work passengers at night, and by requiring an operator to conduct a risk assessment specific to their operations together with implementing procedures to mitigate the identified risks. These steps are designed to ensure that the operation is within the operator's capabilities and that it can be carried out safely.

This guidance outlines how the provisions of the regulations are designed to recognise the differences in organisational elements (structure, processes, flight activity) without reducing the safety margin of the operational risk assessments required. Operators that did not conduct ‘complex operations’ (defined in regulation 138.140) were not required to have a safety management system (SMS). CASA advised that mustering by day in VFR single piston engine helicopters would not be considered a complex operation and confirmed that the operator was not required to have a safety management system (SMS) as part of its manual suite. However, the operator was still required to consider its operational risk. 

In lieu of this SMS requirement, Part 138 Manual of Standards Chapter 13 required operators wanting to conduct aerial mustering to consider hazards associated with the operation. PCH did this using safe work method statements (SWMS) to consider and manage their operational hazards (see Safe Work Method Statements).

In the context of an operator conducting mustering operations using multiple helicopters, the operator had to provide instructions in its operations manual for how additional risk associated with the use of more than one helicopter would be mitigated. This was not discussed in the operator’s SWMS.

Operations manual

Following CASA’s regulatory reform in 2021, there was a significant change in the regulations that applied to aerial work operators not conducting charter or air transport operations. While still required to comply with the requirements of Part 91 – General Operating and Flight Rules, under the new Part 138, aerial work certificate holders were able to scale risk management practices to meet the complexity of their aerial work activities. These changes necessitated significant changes to many operators’ existing manuals.

An operations manual for a Part 138 operator does not need to include the Part 91 general operating flight rules that are intrinsic to the operation of an aircraft. If the operator wishes to place additional obligations on its flight crew that exceeds the Part 91 requirements, the operations manual must contain these instructions. For example, regulation 91.265 prescribes the pilot in command responsibilities for minimum height rules over populous areas.[13] If the aerial work activity requires the pilots to fly lower or closer than the Part 91 regulations specify, then additional instructions must be provided to flight crew.

Management of the manual

The CEO advised that a third party was contracted by PCH to produce an operations manual (OM) that was compliant with the new regulations. The author of the manual advised the ATSB that there was no ongoing contractual obligation for management, update or oversight. The expectation was that the ongoing oversight and updating of the documents would be the responsibility of the operator which would tailor the basic elements to its operation. 

Following the initial issue of the new OM in July 2022, the manual update history showed updates in March 2023 to capture a change of CEO; and June 2024 in response to a CASA surveillance audit. 

Management of risk

Safe Work Method Statements

To comply with the risk-based approach of the Part 138 regulations, a collection of Safe Work Method Statements (SWMS) was generated as part of the new manual. The use of SWMS to manage risk was adapted from other industries based on information available from CASA at the time. These were in essence, standalone risk assessments for the operator’s flying activities that were not part of an integrated SMS. The author advised that SWMS were also used by other operators conducting similar flying activities in Western Australia.

The HOFO and company pilots recalled during interview that the SWMS were the company’s primary risk assessment tool and they referred to these documents to identify hazards and manage threats for all flying activities.

Update history

A review of the operator’s master forms register showed that version 1 of the SWMS published in July 2022 were current at the time of the accident. The register contained the following SWMS:

• General operations
• Stock and feral animal control
• Powerline inspection
• Incendiary
• Underslung load.

Notably, there was not a specific SWMS available for mustering. The HOFO advised that they had been developing a mustering SWMS, but it had not yet been incorporated into the operations manual at the time of the accident. 

Operational risk assessments

The HOFO advised that they referred to the general flying SWMS for their mustering risk assessments but did not produce risk assessments for each muster location. The HOFO would then provide a brief to the lead pilot including a general overview of the property and the areas to be mustered. Following the initial brief, the lead pilot would then assume control of the muster and brief the other pilots and ground crew on the plan for the muster. 

The HOFO maintained oversight of the muster by remaining in contact with the lead pilot throughout the day. If the lead pilot had any questions for the HOFO, or the nature of the muster changed, the HOFO was contactable via satellite phone and daily check-in calls at the end of the day.

Mount Anderson HLS

The operator held a helicopter landing site (HLS) register,[14] but because Mount Anderson Station was not visited often, there was not a dedicated plate in the register. However, the lead pilot had previously operated out of Mount Anderson. The HOFO considered the lead pilot to be very experienced and familiar with the area.

Operational information

Operator’s procedures

General flying separation and minimum heights

Section 2 of the OM outlined each pilot's responsibilities around separation and avoiding a collision. The manual required the pilots to 'maintain a constant look-out at all times' and 'refer to SWMS' for further risk mitigation practices. Elsewhere, the manual required pilots to make use of the radio and helicopter lighting to increase visibility.

The company procedures for aerial work activities were contained in part 6 of the OM. The operator’s procedures noted an exception to the standard 500 ft vertical and 300 m lateral separation standard for helicopters operating over unpopulous areas in CASR 91.267. The relevant set of conditions that were required to disapply the normal minima were as follows:

All of the following apply:

a. The pilot in command of the aircraft is authorised under Part 61,[15] or holds an approval under regulation 91.045, to fly the aircraft below the height mentioned in paragraph (2)(a);

b. The pilot in command of the aircraft conducts a risk assessment of the area to be flown over;

c. The point on the ground or water vertically below the aircraft is not within 150 m of a person, vessel, vehicle or structure or of livestock.

Unless all of these conditions were met, the OM required pilots to adhere to the standard Part 91 rules.

Mustering brief

The task specialist activity of aerial stock mustering was covered in part 6 of the OM. The procedures identified threats to be managed including obstacles on the ground, weather and fatigue. Based on the content of the OM, the manual did not contain procedures relating to the use of multiple helicopters and there was no requirement to conduct a separation briefing. 

A Flight Safety Foundation article on effective briefings[16] outlined that an effective briefing should seek to establish a common mental model between participants with an agreed understanding of key aspects of the proposed flight. These should include:

• defining how the flight will be conducted
  • objectives
  • route
  • spacing between aircraft and responsibilities for its continued application
  • how unplanned changes will be managed
• what communication will be provided and when
• unexpected events and criteria indicating the flight is not proceeding as planned
• a corrective action plan to either restore or abandon the proposed flight
• other aspects identified during the activity risk assessment.

Standard practice

Pilots advised that morning briefings were led by the lead pilot and covered operational aspects with the lead pilot often referring to topographical maps or OzRunways to point out significant locations and landmarks.

Elements typically covered in the morning brief included:

• where the stock were located
• how many head of stock there were likely to be
• in which direction the muster should go
• the location of fences, yards, windmills and any specific reference points such as hills, and powerlines etc
• any hazards or obstacles in the area to be operated in
• location of fuel and that the refuelling point is suitable for landing and taking off.

Company pilots and the HOFO stated that separation was not usually discussed in the briefing. 

While the OM did include a dedicated section on collision avoidance, the OM did not include instructions on what the operator considered acceptable separation. Besides requiring pilots to ‘maintain a constant look-out at all times’, there were no instructions on how separation should be maintained. The expectation was that pilots would arrange their own separation once airborne. 

Pilots reported that the improvised application of reduced vertical and lateral separation became a common and widely accepted practice, and limits varied between pilots. In the absence of formal separation minimums, the accepted practices employed by the pilots became standard operating practice. The HOFO and company pilots stated that this was typical of how mustering operations were conducted. 

Flight conduct on the day of the accident

The HOFO advised that it was normal for company helicopters to arrange a departure sequence prior to take-off and then depart together and travel to the muster as a group. Pilots advised that the sequencing was arranged over the radio on the morning of the accident, and there was nothing unusual about how the group were preparing to depart. 

The use of airframe strobe lights and anti-collision lighting were intended to support see‑and‑avoid by attracting the attention of a pilot looking for other aircraft. While the pilots did not recall seeing the strobe lights on the accident helicopters, they recalled having their own lights on, and believed that the other pilots would also have their lights on, having routinely observed the other pilots fly with the strobes on.

Undocumented risk controls

The operator had identified certain risks associated with mustering activities and had developed practices to mitigate the risks. However, some of these were not published in the OM or SWMS. 

The operator had implemented a practice that utilised GPS lines between locations as demarcation of operational boundaries. This was used when more than one helicopter was operating in the same location to alert pilots to the potential for conflict with a helicopter in the adjoining area. 

In support of the use of these virtual boundaries, radio calls were required when operating near the perimeter of an area to increase awareness among pilots. Company pilots reported use of this practice and the HOFO stated that this method of separation had been effective for many years on musters involving up to 5 helicopters. 

Visibility

The limitations of the see‑and‑avoid principle as the primary means of conflict detection are well understood and covered in numerous ATSB investigations and the ATSB research report 

. While the rate of midair collision during mustering is very low, with the ATSB Aviation Occurrence Database showing about one mustering midair collision every 10 years for the past 20 years, the risk of a collision increases with the introduction of additional aircraft and collision pairs.[17]

R22 cockpit visibility

Based on the wreckage examination and the location of the Heli 2 main rotor strike to Heli 1, the approximate relative lateral position of each helicopter was calculated (Figure 14).

Figure 14: Approximate relative positions of Heli 1 and Heli 2 at the point of collision

Source: ATSB

Similarly, based on the main rotor strike location on Heli 1, the vertical orientation of the aircraft was estimated in Figure 15.

Figure 15: Approximate vertical orientation of Heli 1 and Heli 2 at the point of collision

Source: ATSB

Due to the lack of available flight path and aircraft body angle information it was not possible to complete a full visibility analysis from both aircraft. However, it was possible to demonstrate the limitations to visibility presented by the R22’s cockpit by using a cockpit mask from the pilot’s perspective (Figure 16). Refer to Appendix A – Visibility and R22 exemplar cockpit mask for more detail.

Figure 16: Equirectangular cockpit image of exemplar R22

Source: ATSB

The image was overlaid with a grid to indicate the viewer’s eye position as azimuth and elevation angles from the centre of the view. As the eyes were moved upwards and to the left, so too would these boxes (Figure 17). The cockpit mask shows that although visibility is generally very good out the front of the cockpit, obstructions to line of sight exist above and to the left of the helicopter. 

Figure 17: R22 generic cockpit mask

The blue and orange circles represent the approximate location of the other helicopter at the time of collision. Source: ATSB

For visual detection to be effective, a target must be visible. That is to say it must be, large enough, unshielded, conspicuous, in a location where it is likely to be seen and, detectable against the background.

The operator and other pilots referred to the presence of a ‘blind spot’ above and to the left of the R22 from the pilot’s seat (Figure 17). This is relevant when considering the opportunity available to the pilot of Heli 2 to detect a conflict in this direction.

For the pilot of Heli 1, although the relative position to Heli 2 is unshielded, that position, at about 135° from centre, is behind the pilot.

Regulatory oversight

CASA employs a surveillance schedule to ensure a consistent level of oversight of industry to address compliance issues. Operators, or authorisation holders,[18] are assigned to a surveillance group based on their risk profile which considers the risk to safety posed by their flying activity. 

PCH was assigned to a group with aerial work certificate holders that did not have a training and checking manual or SMS. The surveillance schedule required one surveillance event focusing on core scope items every 5 years. 

Previous ATSB investigations have noted that regulatory oversight processes will always have constraints in their ability to detect problems. There is restricted time and limited resources available for these activities. Regulatory surveillance by CASA is therefore a sampling exercise and cannot examine every aspect of an operator’s activities. The CASA audit schedule seeks to address this limitation by scheduling different types or levels of surveillance over the approval period of an aerial work certificate.

Surveillance

Recent activity

In April 2023, a change of PCH’s CEO triggered a formal review of the organisation’s operations manual finalising the transition to Part 138. A systems level, or level 1 audit was conducted in January 2024. Systems audits are based on a defined scope that references mandatory planning and scoping development documentation. They consider the specific activities conducted by an operator, ensuring their compliance with regulations and the suitability of operators’ systems to manage their flying activities. Additional items can be added by the lead auditor. None were added to the scope of the audit conducted in January. 

Findings

Incomplete procedures and plans to safely conduct operations were identified by CASA during the January surveillance event. The operator made the required changes to its manual, and all findings were acquitted in May 2024. 

While the surveillance activity identified areas where improvements could be made, it did not identify that separation procedures for operations involving multiple helicopters had not been specified. Considering the low likelihood of this type of accident in the mustering sector, the attention of the surveillance team was probably focused on other areas of the operator's activities. 

Post‑accident review

Following the accident, CASA conducted a review of the operator’s procedures manual and issued a safety alert to PCH after it was identified that the OM and SWMS did not provide adequate assessment of the risk, nor sufficient guidance, regarding lateral or vertical separation to pilots when conducting multi‑helicopter operations. The operator reviewed and updated its operations manual (see Safety issues and actions section). This finding was acquitted August 2025.

Related occurrences

The ATSB reviewed 7 investigation reports which involved 2 or more aircraft colliding midair while working together (Table 1). Most of the accidents were fatal. All took place in visual meteorological conditions, and almost all involved a breakdown of see‑and‑avoid as a control for preventing midair collision. Across the investigation reports there are common findings around briefings, planning and procedures.

Table 1: Investigation reports concerning midair collision between aircraft working together

Risk management and planning was highlighted in JIAAC report 088/15. It reported that there was no formal assessment of the risks of operating in close proximity, and no plans for loss of visual contact were in place, such as performance of an evasive manoeuvre. It stated that 2 of the contributing factors were:

Lack of a formal assessment of the safety risks for an unusual operation (filming and flight in proximity), which prevented the identification and analysis of the dangers inherent to that operation.

and

Deficiencies in the operation planning that led to the accident, including the failure of observing the “see and be seen” concept or an evasive maneuver if visual contact is lost between both aircrafts.

Four of the reports (JIAAC 088/15, ATSB AO-2008-010, AO-2023-001, NTSB AAR‑09/02) found that a lack of procedures to support operations with multiple aircraft contributed to the midair collisions. They note that providing procedures specific to the operation would support maintenance of separation and assist pilots to maintain awareness of each other. ATSB investigation report AO-2008-010 found that:

There were no formalised operating procedures detailing the conduct of multiple aircraft culling operations, including the assurance of aircraft separation, that would have assisted the pilots to maintain separation from each other.

ATSB report AO-2023-001 stated:

…documented procedures for communication between inbound and outbound helicopters were not specific to their usual operation and location, and permitted a reactive model of separation, increasing the likelihood that an outbound pilot would not form awareness of relevant traffic.

Pre-flight briefings between pilots were also raised in 2 of the reports. ATSB AO‑2023‑057 found that a manoeuvre conducted had not been briefed on the ground prior to departure and an in-flight discussion did not allow for full consideration of the risks associated with it. AAIB EW/C2017/09/05 stated that while a brief was conducted it did not focus on the whole operation and contributed to an incorrect recall and subsequent incorrect joining manoeuvre when operating in formation.

Together, the investigation reports highlight the risk of midair collision between aircraft, often from the same company, operating in close proximity. The findings highlight the importance of managing the risks of operating multiple aircraft in close proximity to each other and show that pre-prepared proactive methods of separation and deconfliction should be briefed and available to pilots prior to commencing flight. They show that sound operating procedures can support pilots and help ensure that multi-aircraft operations are managed to safe outcomes.

Guidance material

In addition to the acceptable means of compliance and guidance documents produced by CASA, there are a number of other resources available to assist mustering operators to develop the procedures contained in their manuals. 

CASA Sample Part 138 OM

CASA published a Part 138 sample operations manual (Mustering). This sample manual contains guidance on content and examples of acceptable procedures for operations using multiple helicopters and separation. 

Basic Aviation Risk Standard 

The Flight Safety Foundation’s Basic Aviation Risk Standard (BARS) is a suite of risk‑based aviation industry standards with supporting implementation guidelines. The standards are developed by the industry and contracting companies and based around the specific risk these operations face in their day-to-day aviation activities.

The Aerial Mustering Standard is presented in a risk-based format to emphasise the relationship between threats to aerial mustering operations, associated controls and recovery measures. It has been designed to be used as a primary reference tool for aerial mustering operators and contracting companies to manage and understand the aviation safety risk in their operation.

Safety analysis

Introduction

On 25 July 2024, 4 helicopters were involved in a multi‑helicopter mustering operation along the Fitzroy River, about 13 km north-west of the Mount Anderson Station near Curtin, Western Australia. The pilots of 2 Robinson R22 helicopters, VH-HQH (Heli 1) and VH‑HYQ (Heli 2), departed from a clearing behind the homestead to track initially to the west where they would be joined by the 2 other helicopters before following the river to the muster. Shortly after take-off, at a height of about 150 ft, the 2 helicopters collided resulting in an immediate loss of control. Both aircraft collided with terrain and the pilots were fatally injured.

No evidence of pre-existing mechanical defects were found with the helicopters. Additionally, there was no evidence that either helicopter had struck flora, fauna or a wire prior to the collision. The smoke that was observed from the lead helicopter immediately following the collision was likely due to the damaged smoker tank spilling diesel onto the hot engine and exhaust after it was struck by the main rotor of the following helicopter. 

This analysis will discuss the:

• pilots’ awareness of the other helicopter prior to collision
• briefing that was conducted prior to the flight
• typical separation employed by company pilots
• operator’s standard operating procedures and management of operational risk.

Collision detection

Estimated flight path prior to collision

A witness reported that the helicopters were travelling in a similar direction and at a similar speed to one another prior to the collision. Wreckage examination revealed that when the helicopters collided, they were almost alongside each other in an attitude consistent with level flight, with Heli 2 being about 2 m behind and about 2.5 m lower. This was not consistent with the agreed sequencing that Heli 2 would follow Heli 1 out. 

The pilot of Heli 3 reported the last sighted track of Heli 1 and Heli 2 was of them passing about 80 m south of where Heli 3 was parked, closer than 100 m apart. This meant that prior to the collision, they were probably established on a parallel west-north-west track (about 285° magnetic) from the overnight parking area.

Environmental considerations

The twilight conditions should not have impacted the pilot of Heli 2, given Heli 1 was higher and set against the sky. The conditions would have an uncertain effect on the pilot of Heli 1 who would need to observe Heli 2 against the background consisting of terrain and vegetation. There was no reason to suspect the helicopters were not utilising standard position indication (conspicuity) lighting, which would also mitigate that effect. Additionally, while both helicopters were fitted with high-visibility paint schemes on the main rotor blades to increase conspicuity, the contrast against the background in the morning light could not be determined. 

Audio analysis

Audio analysis of video recorded at the time of the collision (that did not include footage of the helicopters) showed that prior to the collision, a notable reduction in frequency was observed from one source. As the helicopters were established to be operating normally, this frequency change was consistent with one of the helicopters manoeuvring away from the microphone of the recorder at a greater speed. As the helicopters were moving from right to left relative to the microphone, likely at similar speeds, this was most likely due to one of the helicopters turning right. 

Based on the relative position of Heli 1 being in front of Heli 2, the turning helicopter was almost certainly Heli 1. On the balance of evidence, this change established the helicopters on converging flightpaths. About 2 seconds after the completion of the turn, audio analysis showed sounds consistent with the collision between Heli 1 and Heli 2.

Wreckage trajectory

Following the midair collision, witnesses observed the helicopters descend on different trajectories. The main debris trail was arranged in 2 key areas. Each consisted of the heaviest components for each helicopter (except the main rotor of Heli 2 which detached at the point of collision or immediately thereafter). Based on the wreckage trajectories from the estimated point of collision, the helicopters were likely maintaining different tracks at the point of collision.

Conflicting traffic undetected

Shortly prior to the midair collision, the pilot of Heli 1 was communicating with Heli 3 and Heli 4 on the ground. It is plausible that the pilot of Heli 1 was looking towards the helicopters at this time, focused ahead and to the right of the helicopter with the expectation that Heli 2 would have been behind. 

Additionally, the pilot of Heli 2, operating slightly lower than Heli 1, may also have looked towards Heli 3 and Heli 4 on the ground. Heli 3 and Heli 4 were preparing to lift, and the pilot of Heli 2 may have needed to assess a potential risk of collision. For the pilot of Heli 2, the helicopters on the ground would have been located to the right of the nose and any attempt to sight them would have drawn their attention away from Heli 1 located above and to their left. This relative position is known to include a potential ‘blind spot’ caused by cabin structure. While limited data prevented a full visibility study, the wreckage examination indicated that the orientation of the helicopters at the point of collision may have placed Heli 1 in this area which could have resulted in an obscured line of sight for the pilot of Heli 2, limiting their opportunity to detect the conflict.

The evidence shows that the pilots involved in the accident were aware of the other helicopter’s existence but did not detect the converging flight paths either at all, or with sufficient time to take evasive action. The strongest indication of this was the lack of any radio transmission or evasive manoeuvring from either pilot. 

Established separation practice

Normalisation of deviance

Repeated exposure to risk without consequence is known to affect pilots' decision‑making (Hollenbeck and others, 1994). The concept of ‘normalisation of deviance’ describes the desensitisation to risk experienced by individuals or groups who repeatedly deviate from safe operating practices, without encountering negative consequences. A prominent feature of the normalisation of deviance is the desensitisation process, where frequent deviations result in the practice’s normalisation and perceived standardisation within everyday operations. This sets a precedent for what is considered tolerable and establishes a new normal from which further deviations may occur. In the absence of intervention, this cycle of deviance is disrupted only when the behaviour results in an undesirable outcome such as an accident (Sedlar and others, 2022).

Standard practice

The other pilots involved in the muster confirmed that although sequencing was arranged between the accident pilots on the day, the collision occurred before the pilots had arranged their own separation. While sequencing was one means of establishing initial separation, it would not ensure separation was maintained.

An established plan or standard procedure to manage the risk of collision after take-off would have provided the pilots with a tool to support effective decision‑making around in‑flight separation. By not defining procedures for operations that carried inherent risk, the operator did not have control of how the risk was managed. With pilots permitted to set their own separation, this produced varying results based on personal preference and over time, the practice of flying in proximity became the accepted norm.

In this system, there was a reliance on the pilots maintaining a lookout for other helicopters and taking avoiding action as necessary. See-and-avoid, though a well‑established defence against midair collision in uncontrolled airspace, is known to be fallible (CASA, 2021) and its limitations as a primary means of collision avoidance are well documented.  

Repeated exposure to the established separation practice probably affected the mustering pilots’ perception of the associated risk. Flying in proximity to other aircraft reduced the opportunity to detect a conflict and manage the situation by decreasing the time available to respond to a threat. Further, this practice increased the risk of collision in the event of an unexpected or unanticipated manoeuvre.

Briefing

Based on the information obtained during interview, the morning brief was attended by all pilots and included the content specified in the operations manual. It focused solely on the operational aspects of the planned muster and did not include safety elements such as how the aircraft would be flown with regard to sequencing and separation.

The concept of a shared mental model has been found to be an essential component of how teams process information and perform (Salas & Fiore, 2004). Shared mental models affect team coordination by providing mutual expectations from which team members can know what their teammates are doing and predict future actions (Cannon‑Bowers and others, 1993). Through situation awareness, a team has a better understanding of the current and future state of the system and is more likely to be successful (Endsley, 2015). Mechanisms to support team members in monitoring each other and adapting to each other’s performance to modify their own task performance accordingly, result in better outcomes (Marks & Panzer, 2004). 

The morning briefing was an opportunity for the pilots to establish a mutual understanding of how the day’s flying would be conducted and their responsibilities when changing the plan or recovering from a deviation from the plan. Guidance on effective briefings by the Flight Safety Foundation[19] suggests that including these elements in the pre-muster brief with some prescriptive minimums would have allowed the pilots to form a better understanding of how the flight would be conducted and, importantly, what the other pilots were trying to achieve and how the other helicopter would be flown. It could have prompted the pilots to consider where other aircraft would be during departure and may have helped the pilots detect a potential conflict. While not including these safety elements in the morning brief did not contribute to the accident, it represented a missed opportunity to manage the risk of collision shortly after departure.

Operator’s procedures

Awareness of operational risk

The majority of the work conducted by the operator was aerial mustering utilising multiple helicopters. A review of the operations manual and evidence obtained during interview confirmed that the operator documented its expectation that pilots would maintain a good lookout and use visual aids such as aircraft lights to avoid a collision. In addition, the operator had implemented undocumented procedures for managing separation during mustering. This indicated an awareness that the risk of collision existed when operating with more than one helicopter. 

The operator permitted pilots to establish their own separation relying on the pilot’s ability to identify and manage the risks of collision in multi-helicopter operations. Operating with unofficial guidelines can be one way to allow for autonomy and flexibility, however, an inherent problem with this approach is that even though guidelines have no official status, in practice, they can take on the character of formal policy in the absence of an alternative, however well-reasoned, without being sanctioned (Brandsen and others, 2006).

Importance of separation procedures

The importance of procedures to manage separation were identified in a previous ATSB report into a midair collision involving a Robinson Helicopters R44 and Piper Super Cub.[20] The report identified the following:

A procedural framework comprising, for example, lateral separation minima, radio communication requirements, the designation of a ‘coordinating pilot’, and specified responses to abnormal situations, could have assisted the pilots to maintain adequate separation, and to respond appropriately once they knew separation had broken down.

It was found that the operator did not provide a procedural framework for establishing and maintaining separation during aerial work activities. 

Like the benefits of an effective briefing, defined separation procedures would provide participants with a shared understanding of how elements of the flight would be conducted in proximity to other machines. Further, practices enabling sighting of the helicopter ahead would assist with detection, and appropriate separation practices would assist the pilots by providing additional time to identify and respond to emerging threats. 

The accepted separation practice increased the risk that 2 helicopters would be flown close enough to cause a collision hazard and the operator’s existing procedures did not manage this operational risk.

Operator’s management of risk

Historically, there has been a preference for using multiple helicopters on a single muster. The risk of collision was known to the operator, demonstrated by the undocumented strategies and practices implemented to avoid collisions during mustering operations. However, the risks associated with operating more than one aircraft in proximity had not been captured in the operator’s Safe Work Method Statements (SWMS) which were the primary tool for identifying and considering operational risks. 

Records showed that the contents of the SWMS had not been updated since their introduction 2 years prior and did not include a mustering‑specific SWMS even though SWMS were available for other task specialist activities. Critically, the existing SWMS did not identify the hazard of midair collision or provide controls to manage aircraft separation. In the context of its primary business, this represented an under-utilisation of the operator’s available risk management tools. Not considering the nature of the flying activities increased the likelihood that operational risk would not be identified or managed appropriately. 

Findings

From the evidence available, the following findings are made with respect to the midair collision involving Robinson R22 Beta II, VH-HQH, and Robinson R22 Beta II, VH‑HYQ, 51 km south-south-east of Curtin Airport, Western Australia, on 25 July 2024. 

Contributing factors

• During initial climb while flying in close proximity towards the muster location, the pilots of VH-HQH and VH-HYQ did not detect their converging flight paths, which resulted in the midair collision.
• Consistent with the operator's accepted and routine practice, the pilots flew in close proximity without an established plan for separation, increasing the risk of collision.
• Pearl Coast Helicopters did not establish appropriate separation standards for its helicopters or provide documented procedures to ensure pilots established and maintained appropriate separation. (Safety Issue)
• Pearl Coast Helicopters did not formally manage risk in the context of its primary business which was multiple helicopter mustering operations. (Safety Issue)

Other factors that increased risk

• The morning briefing was solely operational and did not include elements of safety. Critically, separation between helicopters was not discussed.

Safety issues and actions

Absent procedures 

Safety issue number: AO-2024-044-SI-01

Safety issue description: Pearl Coast Helicopters did not establish appropriate separation standards for its helicopters or provide documented procedures to ensure pilots established and maintained appropriate separation.

Identification of risk not tailored to flying operations

Safety issue number: AO-2024-044-SI-02

Safety issue description: Pearl Coast Helicopters did not formally manage risk in the context of its primary business which was multiple helicopter mustering operations.

Glossary

Sources and submissions

Sources of information

The sources of information during the investigation included the:

• Pearl Coast Helicopters Pty Ltd
• other pilots employed by the operator involved in the muster
• Civil Aviation Safety Authority
• Western Australia Police Force
• Robinson Helicopter Company
• Helico Australia
• Bureau of Meteorology
• accident witnesses
• video footage of the accident flight and other photographs and videos taken on the day of the accident
• recorded data from the GPS unit on the aircraft.

References

Brandsen, T., Boogers, M., & Tops, P. (2006). Soft Governance, Hard Consequences: The Ambiguous Status of Unofficial Guidelines. Public Administration Review, 66(4), 546–553. http://www.jstor.org/stable/3843940)

Cannon-Bowers, J. A., Salas, E., & Converse, S. (1993). Shared mental models in expert team decision making. In N. J. Castellan, Jr. (Ed.), Individual and group decision making: Current issues (pp. 221–246). Lawrence Erlbaum Associates, Inc.

CASA. (2014). Sector safety risk profile – Aerial mustering.

CASA. (2021). AC 91-14 v1.0 Pilots’ responsibility for collision avoidance. Canberra: Civil Aviation Safety Authority

CASA. (2024). AC 138-01 v3.1 Part 138 core concepts. Canberra: Civil Aviation Safety Authority

CASA. (2025). Agricultural flying sector safety risk profile. https://www.casa.gov.au/operations-safety-and-travel/safety-management-systems/sector-safety-risk-profiles/agricultural-flying-sector-safety-risk-profile/

CASA. (2025). Part 138 guide to sample operations manual (Mustering). Canberra: Civil Aviation Safety Authority

CASA. (2025). Visual flight rules guide v8.2 Canberra: Civil Aviation Safety Authority

CASR Part 91 (General flying and operating flight rules), (1998) Cth.

CASR Part 138 (Aerial work operations) Manual of Standards, (2020) Cth.

Civil Aviation Act (1988) Cth.

Civil Aviation Safety Regulation (1988) Cth.

Endsley, M. R. (2015). Situation Awareness Misconceptions and Misunderstandings. Journal of Cognitive Engineering and Decision Making, 9(1), 4-32. https://doi.org/10.1177/1555343415572631

Flight Safety Foundation. (2025). Basic Aviation Risk Standard for Aerial Mustering v4.0. Flight Safety Foundation. https://flightsafety.org/

Hobbs, A. (1991). Limitations of the see-and-avoid principle. Canberra: Australian Transport Safety Bureau.

Hollenbeck, J. Ilgen, D. Phillips, J. Hedlund J. (1994) Decision risk in dynamic two-stage contexts: beyond the status quo. Journal of Applied Psychology, Vol. 79, Issue 4, pp. 592–598.

Kepert, J.L. (1976). The use of wreckage trajectories in aircraft accident investigation. Department of Defence. Defence science and technology organization aeronautical research laboratories. 

Leibowitz, H. W., & Owens, D. A. (1991). Can normal outdoor activities be carried out during civil twilight? Applied Optics 30 (24), 3501-3503.

Marks, M. A., & Panzer, F. J. (2004). The influence of team monitoring on team processes and performance. Human Performance, 17(1), 25-41.

Purves D, Augustine GJ, Fitzpatrick D, et al., 2001.Functional Specialization of the Rod and Cone Systems. Neuroscience. 2nd edition. Editors. Sunderland (MA): Sinauer Associates.

Salas, E. E., & Fiore, S. M. (2004). Team cognition: Understanding the factors that drive process and performance. American Psychological Association.

Sedlar, N. Irwin, A. Martin, D. & Roberts, R. (2022). A qualitative systematic review on the application of the normalization of deviance phenomenon within high-risk industries. Journal of Safety Research. School of Psychology, William Guild Building, University of Aberdeen, Aberdeen, UK. & Aberdeen Business School, Robert Gordon University (RGU), Aberdeen, UK.

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:

• Pearl Coast Helicopters Pty Ltd
• other pilots employed by the operator involved in the muster
• Civil Aviation Safety Authority
• Western Australia Police Service
• Robinson Helicopter Company
• Helico Australia.

Submissions were received from:

• Pearl Coast Helicopters Pty Ltd
• Civil Aviation Safety Authority
• Robinson Helicopter Company.

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

Appendices

Appendix A – Visibility and R22 exemplar cockpit mask

Capture of the base image

The cockpit mask in Figure 16 was developed from a 360° equirectangular image taken from a location approximating the eye position of a person seated in the pilot’s seat. This position was determined from medical records and an exemplar pilot of similar height.

While there are limitations to this method of image capture, these limitations do not invalidate the use of the mask.

 Limitations include:

• mask developed from a single position
• image based on an aircraft on level ground and not in flight attitude
• use of wide‑angle lenses can introduce image distortion at the edges of an image
• flight controls captured in a static position.

Visual detection of a target aircraft is impacted by 3 different but related factors: visibility, detectability and conspicuity. Visibility considers whether an aircraft is visible to the person looking for it. Detectability considers how easily a target can be detected from the eye position of the person looking for it. Conspicuity is whether a target stands out from its background sufficiently to be identified.

Creating the mask

The relative position of a target from the viewer’s eye position is represented as azimuth and elevation angles from the centre of the view. As the equirectangular image and subsequently the mask represents a full 360° of azimuth and 180° of elevation, a grid can be overlayed on the image to assist the viewer to understand how far from the centre of the view objects will appear or the relative angle between 2 things in their view.

Figure 18 shows the mask overlain with a 15° marked grid in both azimuth and elevation angles.

Figure 18: R22 generic mask with grid overlay

Source: ATSB

Field of view 

The cockpit mask is a spherical representation of the aircraft structure around the estimated pilot’s eye position. While the 360° camera can capture this whole area, the human eye has a limitation to how much of this area that it can detect. This is referred to as the field of view (FOV), it consists of approximately 200° of azimuth and 135° of elevation. The opportunity for target detection is not constant across this area, it varies depending on where the target is located. For the purposes of ATSB visibility analysis, the FOV is broken down into 3 areas consisting of the foveal, inner and full visual fields, progressing from the centre of the visual field outwards. Further detail of the size and differences between these areas can be found in 

.

Figure 19 shows approximate areas of the visual field overlain on the R22 generic cockpit mask as developed earlier. In this case the field of view has been positioned as if the eyes are directed straight ahead to demonstrate the size of the field. Rotation of the eyes will change where in the FOV parts of the structure appear, which may impact visual acquisition of the target. It is important to note that movement of the head may also impact FOV but it will impact the relative location of the structure so modelling rotation of the head would require a different cockpit mask. 

Figure 19: R22 generic cockpit mask showing fields of view

Source: ATSB

Conclusions

The mask shows that when looking ahead there is minimal obstruction to the pilot’s view and with the doors removed there are large areas on both sides that are unobstructed. However, there are areas where a target will be shielded such as behind the instrument cluster and the cabin roof. For a pilot seated in the right seat, this shielding is greater in the left side of the pilot’s field of view. This means that a potential conflict is more likely to be obstructed when the threat originates from the left of the helicopter. 

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 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.