New study on Electronic Conspicuity published by aviation regulator

Hi, I’m Steve Jarvis, and welcome to this video from the CAA, GASCo, and Jarvis Bagshaw Ltd. We’ve been researching the human factors impact of electronic conspicuity (EC) devices in GA. We’ve carried out experimental flights with eye tracking glasses on GA pilots, in several aircraft types. We’ve been researching the impact on pilots of electronic conspicuity (EC) devices in GA.

We’ve carried out experimental flights with eye tracking glasses on GA pilots, in several aircraft types and over 2,000 GA pilots kindly responded to a comprehensive survey. Some findings will surprise you. I’ll be telling you about threats we’ve found and offering you short tips to address each. But rest assured that I am not simply going to tell you to stare harder out of the window. Mid-air collision risk is frightening, because it takes little account of pilot experience or skill and offers few opportunities for risk mitigation.

Electronic conspicuity (EC) devices DO offer some mitigation, and usage has grown strongly. But consider the following question. In current UK airspace, what is the likelihood of detecting an EC-equipped aircraft from another (assuming they are within range)? The answer is under 50% on average; even assuming 9 in 10 pilots are flying with a working EC device.

And this brings us to the first of six key points: Because EC detects less than 50% of other aircraft, There will be undetected aircraft around you, probably more so than detected aircraft. Visual lookout is still the top priority! In addition to this, it is important to note that each EC device has different likelihoods for detectability versus detection.

For example, one of the most popular devices has well over a 50% chance of detecting another aircraft in UK airspace, but well under a 50% chance of being detected itself. Not only did pilots over-estimate these likelihoods in the survey, but their estimates bore no statistical relationship to the actual devices they used themselves, suggesting a general lack of information about the strengths and weaknesses of specific EC devices. This was reflected in many survey comments, where pilots recalled feeling angry on seeing an aircraft out of the window that was not showing on their EC, or where a converging EC target took no avoiding action. But in most of these cases, it is likely that the offending aircraft did have EC,

And this brings us to point 2: Assume your aircraft is not showing on the others’ EC display, even when they are appearing on your EC display… and it follows that you should never expect an EC detected aircraft to avoid you. Now take a look at this eye tracker footage from a small helicopter. This is effectively a headcam video with a blue dot showing where the pilot’s eyes are pointing. There are currently no aircraft displayed on the pilot’s EC... that’s on the iPad there.

If we draw lines for eye movements and use circles (sized to represent the duration of each look) we get a ‘scan-path map’. Although flat, the picture represents about 180 degrees around the pilot. Notice how the lookout covers a wide breadth. But just one minute later, another aircraft makes a position radio call and appears in the 12 o’clock as an EC plot. Let’s see what happens now…

Now let’s compare the before and after scan path map. The new scanpath map (after the aircraft is known about) is underneath. Notice the pilot has locked into a narrow search pattern between the iPad and the expected target area. Crucially…. normal lookout has stopped. Before talking more about this, another quick question: “realistically… at what distance can you see another GA aircraft (in good conditions)?” Because searching beyond visual range removes attention from the 50% of non-EC traffic, for no gain. Research has found the answer to be between 1 and 2 miles, and it is realistic to consider three miles as the maximum in perfect circumstances.

So how much of this scanning was within realistic visual range?... The answer is none, it is all between 3 and 12 miles.

This brings us to point 3… Don’t prioritise searching for EC-targets beyond about three miles. But our helicopter event showed us a further risk, that was mirrored in another trial too. Only after landing did it become clear that the EC target and the radio call were from different aircraft, and we can call this the 2-in-1 illusion. It is where natural human cognition merges the evidence of two aircraft into one. Let’s return to the scanpath map.

All this time, the ‘radio-call’ aircraft was here … invisible to EC. The pilot had no suspicion of another aircraft because EC had offered his brain a way of falsely accounting for that aircraft’s radio call. Not only did this happen twice in four trials, but many survey anecdotes are suggestive of this happening often (regularly) (though without further data it is not usually possible for the pilots to have known).

So this brings us to point 4: Point 4 - Assume new signs of traffic do not belong to an existing EC detected target, until you know for sure.

Now consider the following: Can EC make pilots accept more risk? Consider the following anecdotes, loosely based on findings from the survey: Two pilots plan to fly from Airfield A to Airfield B in separate aircraft at the same time. They recognise the risk of being in close proximity to each other, but visibility is excellent. However, before the flight, visibility has reduced considerably. Fortunately, they have compatible EC and so they can conduct the flight safely. The next anecdote is as follows An instructor needs to do an intense IMC training sortie with IMC goggles on the student. It is midday Saturday, and the normally busy airspace is even busier than normal with two glider competitions in the nearby vicinity. The instructor recognises the threat and decides that it can still go ahead because once airborne, they will use FLARM-equipped EC to locate the best local area to conduct the training. Free of other aircraft. In both these situations, the pilots have rightly recognised the potential threat and have considered the mitigations.

This is of course good! On the other hand, both have factored the existence of EC into their risk-based decision making, in terms of whether it is safe to carry out the flight. This is often subtle, and can happen in flight too. Care is required because current EC detection rates are relatively low and devices can be unreliable (for instance aircraft structure can blank the signals).

This brings us to point 5… 5. When making big decisions, ask yourself “would I do this if I did not have EC? …If the answer is no, then re-consider whether it is OK. Now I’d like to raise another important issue: Over three-quarters of powered aircraft midair collisions happen in the circuit (including approach and climb out). Many alarming survey anecdotes related near misses in the circuit with an unknown aircraft appearing as if from nowhere, having not shown on EC.

In the trial, a 2-in-1 illusion in a circuit led to a late avoidance manoeuvre. Let’s take a look at the situation… Our aircraft is HERE, and we are trying to spot the EC displayed aircraft here. Evidence would suggest that with the extra workload and traffic volume of the circuit, we become even more vulnerable to that false sense of security from EC. To make things worse, screens can get very cluttered in busy areas around airfields, and spurious audio alerts increase. This can distract heavily from other priorities, including lookout. It can be helpful to declutter the display using the filters; for example, by removing EC targets over 4000ft above. Something everyone can do to help, is to avoid leaving EC devices switched on when parked, taxying, or waiting to line up, because these cause large numbers of spurious alerts, resulting in less EC usage in the circuit itself. In busy circuits, many pilots say that they switch off audio alerts, or even their EC device, because it is too distracting, and some pilots avoid taking EC at all if only flying circuits. Paradoxically therefore, there may be a lower proportion of EC usage in a busy circuit than a quiet one, and there is often less usage in circuits than elsewhere, despite circuits representing a higher collision risk.

All this brings me to the final point. You will increase your EC reliance in the circuit, particularly under high workload. Try to add thorough visual searches into your circuit and remember that EC usage in busy circuits can be even lower than elsewhere. Whereas the research showed threats of EC usage, it also raised some fascinating evidence showing why EC is, nevertheless, very important. It can be tempting to attribute poor lookout to mid-air collisions. But take a look at this scan path map from a fixed wing aircraft pilot. During this period of lookout there was an aircraft approaching within 3 miles. Let’s see where it was… here. Despite youth, experience, excellent eyesight, and looking directly in the right area, our pilot did not see this. This demonstrates that even perfect lookout in very good conditions might not be enough. This shows us how important EC can (and will) be, and therefore whereas this video highlights human factors threats of EC, we are in no way suggesting that EC is bad. The message is that safety is improved by EC complementing good lookout.

Let’s summarise the six points

1. On average, EC detects less than 50% of other GA aircraft at the moment. There will be undetected aircraft around you, probably more than detected aircraft. Visual lookout is still the top priority! The second point was…
   
   
2. Assume your aircraft is not showing on the others’ EC display, and never expect an EC detected aircraft to avoid you.
   
   
3. Don’t prioritise searching for EC-targets beyond 3 miles.
   
   
4. Assume new signs of traffic do NOT belong to an existing EC detected target, until you know for sure. I.e. avoid the two-in-one illusion.
   
   
5. The Fifth point was When making big decisions, ask yourself “would I do this if I did not have EC? …If the answer is no, then it’s a good idea to re-consider whether it is going to be ok.
   
   
6. That last point was you will increase your EC reliance in the circuit, particularly under high workload. Try to add thorough visual searches into your circuit and remember that EC usage in busy circuits can be even lower than elsewhere. Overall point however – EC can help mitigate for lookout vulnerabilities, and good lookout is as important as ever! I hope that’s been of interest and of use.

This research was by GASCO and Jarvis Bagshaw Ltd, for the CAA I’m Steve Jarvis Many thanks for watching.