We recently attended the RAeS Flights Simulation Group conference in October 22-23 2024 in London. One of the topics that caught our attention was on the developing situation regarding Simulated Air Traffic Control Environment (SATCE) training technology. We have long been of the opinion that this is the missing element in the provision of a fully immersive operational environment in FSTDs. Not only with the surreal experience of pushing back, taxiing out and taking off at an airport like London Gatwick at 09:00 in the morning without any competing traffic but, more importantly from a training perspective, the lack of ATC induced workload and potential distractions.
After the conference we caught up with Dr Jeremy Goodman MRAeS, to find out what the current status of these systems were.
Q - Can you update us on the current products on the market, what’s the state of the art solutions at the moment?
A - Hopefully I can help with a long-term view of where we are today, having started working on SATCE 15 years ago (yes, it’s been that long!) in several roles with two organisations. I currently consult to ASTi, a US-based simulation specialist company, supporting their fielded SATCE solution called SERA in an advocacy role, interfacing with training providers, device manufacturers, regulators and supporting research.
Several years ago, there were a number of SATCE solutions in development globally, but only a few now remain. Technical challenges and the long time to market forced several development projects to close. Of note, has been ASTi, who were successful with early adoption of their SATCE solution, which began largely in military cadet pilot training. This continues to expand globally - more recently also incorporating advanced and multi-crew military training applications. There are fewer regulatory constraints in this sector, and demand is largely based on training value and efficiencies.
SATCE technology can now be considered mature, and suitable for both initial and advanced scenario-based military and commercial flight training. Many training experts I’ve spoken to don’t see SATCE useful for all training - especially manoeuvre-based training, where the focus is not re-creating line operations. For this training, SATCE can be turned off, or traffic removed, but for line-orientated training, where realistic operational threats and complexity approaching real-world conditions are necessary, then SATCE adds a new level of realism that’s long been missing.
The core function concerns simulating the operational environment outside the cockpit, and so it’s largely agnostic to the simulator aircraft type and air traffic management (ATM) procedures – this means it can be flexibly deployed for a wide range of training applications and locations worldwide. For example, ASTi has now fielded its SERA solution, with over 300 systems delivered, for military rotary wing, fixed wing, and commercial cadet pilot training, including the multi-crew pilot licence (MPL). There are eVTOL OEMs also adopting SATCE for future vehicle certification and pilot training.
Q - We recall that some of the earlier systems we have seen demonstrated did struggle with regional accents, have these issues been solved now?
In short, yes. Speech recognition technology has greatly improved. SATCE solutions using up-to-date technologies are no longer constrained by speech recognition or latency, which were challenges in the early years of development, let’s say, 10 years ago.
Strong regional accents are no longer problematic. In fact, recognition and computation speeds are now so fast that various artificial delays need to be built in to make the radio environment simulation more realistic! There have also been improvements in the robustness of the algorithms that interpret pilot radio transmissions and then decide appropriate ATC responses. These, and other enhancements, have added to the technology readiness.
Another early constraint was the capability for visual systems to render multiple other traffic, but modern FSTD image generators (IGs) are more powerful, and this is especially true of the next-generation IGs leveraging gaming engines, such as Unreal Engine. Synthetic ATC should automatically control aerodrome ground lighting, rather than the instructor having to do this at the IOS, and the latest generation of IGs are also more flexible to facilitate this new automation feature.
Q - As we said at the top of the interview, we believe that real life workload and distractions to aircrew from ATC interactions is significant, however has there been any research that supports our belief?
A - Yes, there are some early research results and more sizable global safety audit data that both support your belief, with more research underway and data analysis planned!
Let’s deal with workload first. Initial evidence from research using SATCE and observational evidence from line operations suggest both ATC and traffic present significant workload and threats to modern commercial flights.
It would be hard to measure workload in real world flight without ATC, because you cannot not have ATC! However, we do have evidence from initial trial results using an early version of ASTi’s SERA SATCE from the UK RAF Central Flying School. This was published last year and presented by Dr Jonathan Allsop and Flt Lt Richard Keeling at the RAeS Conference. Please bear in mind, this was a small trial, so we need to treat the results with a little caution, but they support exactly what you and most experts expect.
The authors demonstrated that adding both simulated ATC and traffic in flight training adds significant workload and situational demands. SATCE increased overall pilot workload compared to sterile simulator conditions by 81%. They also noted that pilot performance in the simulator declined, which may become another outcome we can expect. [For more information, see the paper reference below.]
If, at any point in a flight, ATC becomes a distraction or challenge for flight crews, then this can be classed as a ‘threat’, according to the Threat and Error Management (TEM) framework. An example might be a busy frequency during a demanding flight phase, where a pilot needs to delay transmitting. This kind of threat and many others from ATC and traffic, cannot be easily simulated without using an automated SATCE system in FSTDs.
There’s an interesting paper from 2006, that analyses Line Operation Safety Audit (LOSA) observations from 4,500 flights that shows ATC is as common a threat as adverse weather – both accounting for approximately 25% of observed threats. These were higher than threats from the aircraft itself, which were at 13%, for example. [For more information, see the paper reference below.]
These data, and more recent LOSA observations from much larger datasets, all point to the fact that operational threats in real world flying, including those from ATC and other traffic, are not represented proportionally or replicated adequately in modern flight training using simulation. This is why SATCE is necessary – it addresses a significant flight training deficit using simulators and current approaches.
Q - So from the research you believe that this is a significant skill that needs to be trained for?
A – Yes. My own observations of the impact of SATCE align with indicative research results – that a more immersive synthetic operating environment raises workloads. I’m not a training expert, but I know that managing workload is considered a core pilot competency. As a result, I believe that SATCE has a significant role to play in enhancing pilot training, and that its deployment in training programmes at scale will make measurable impacts upon pilot proficiency and resilience.
It may be worth noting that SATCE doesn’t just deliver more realistic workloads, but it supports all eight pilot competencies in the ICAO Evidence Based Training (EBT) framework (and we can add knowledge to that list too). These are the core attributes of every pilot that the entire training ecosystem is aiming to develop and deliver.
Q - From ASTI’s presentations at the conference it seems that whilst military applications are being rolled out the civil market has been slower to adopt, why do you think this is?
A - The slower adoption of SATCE in civil flight training came down to a classic regulatory-market deadlock that we faced. Training device manufacturers were hesitant to invest in technology that wasn't required by regulation, while regulators wanted to see proven implementation before making any mandates. This created a significant barrier to early adoption in the commercial training sector that is now thankfully behind us.
Recently, the first major airline committed to implementing SATCE for their MPL training program which will go live shortly, and we're seeing other industry innovators beginning to adopt SATCE on a larger scale for 2025.
Also significant, the FAA signed a Cooperative Research and Development Agreement (CRADA) with ASTi in early 2024 to specifically study the effectiveness of SATCE in airline pilot training using FSTDs. This was largely driven by safety concerns, with their stated goal being “to improve the pilot training experience and reduce the rate of accidents and incidents related to pilot-controller miscommunication”. Initial results from the FAA study are expected mid-2025, which could further accelerate awareness and implementation.
Recent high-profile incidents, especially runway incursions, have also highlighted the opportunity to enhance pilot training using SATCE to generate much more realistic operational threats. Many experts across the flight training industry joined Cpt. Michael Varney and myself by contributing to white paper in July 2023 calling for the widespread implementation of SATCE. This was published by Halldale and is a good introduction if you’re relatively new to the subject. [For more information, see the paper reference below.]
Q - With the forthcoming change in qualification basis in EASA going toward a task-to-tool philosophy surely this paves the way for SATCE to fulfil in the training device definitions by way of the FSTD Capability Standard (FCS)?
A – For certain, the task-to-tool methodology, and the growing adoption of competency-based training and assessment (CBTA) will pave the way for the widespread use of SATCE. CBTA programmes, including the MPL and EBT require a “role relatable environment”, which in our case is an aircraft cockpit that includes a realistic simulation of the aircraft itself and the environment in which it is operating.
The technical requirements for SATCE have been incorporated in EASA’s most recent draft CS-FSTD (NPA 2024-102) – and these are aligned with the industry consensus standard for SATCE, which is ARINC Specification 439.
However, I believe EASA has decided not yet to include SATCE as an FSTD feature, which would mean, as things currently stand, that the CS-FSTD will not align with ICAO Doc 9625, where SATCE is considered an FSTD feature called “Environment – ATC”.
The draft CS is still in final revision, so not yet finalised. Currently, if SATCE is qualified as part of an FSTD, then I understand that the SATCE functionality and fidelity level would be noted on the FSTD Qualification Certificate (QC) under the heading “Additional Capabilities”.
I’d personally advocate for EASA to align fully with Doc 9625 and view SATCE as an FSTD feature – it’s an essential component of the simulated environment. In submitted comments we’ve also suggested “Environment – ATC” is renamed “ATC & Traffic”, because that is a succinct description of the scope. We’re not asking for it to be mandated, but if the framework is in place, it’ll help with the coming roll out in commercial flight training in FSTDs qualified under EASA.
Q - What’s the rationale from EASA for this, the original NPA 2020-15 seemed clear?
A - I think it’s important to recognise that EASA will be the first regulatory authority to publish SATCE technical requirements for FSTDs. Furthermore, EASA is essentially the first authority to implement the FCS framework. Both of which are significant steps forward.
I was surprised by the change from NPA 2020-15 (that is, SATCE not being considered an FSTD Feature), as SATCE is required in EASA Part-FCL for devices that support the latter phases of the MPL. I asked EASA and the Rule-making Task (RMT) the same question you’ve asked at the focussed consultation for NPA 2024-102, and we received several answers.
More recently at the RAeS Conference, an EASA representative answered a question about SATCE from the Conference Chair, where they reported they had to draw a line somewhere to changes in the draft, and only had so much time and resourcing. So, it appears to have been a pragmatic decision more than anything. Based on the information currently available, it’s anticipated that new and future technologies (including SATCE) will be addressed in the next RMT work package. Work on these developments is expected to continue immediately after the comments on NPA 2024-102 are reviewed and resolved.
My hope is that EASA may reconsider how SATCE is categorised. But if there’s no change from the current draft, at worst this is a temporary setback - it’s unlikely to stop growing global adoption. We are already past the tipping point. We’re at the breaking wave of this technology in our industry.
We’ve waited a lot of years for SATCE and it’s finally fielded and in training. It’ll have a very significant impact on our industry. It affects so much – here’s a quick shortlist: CBTA, MPL, EBT, human factors, crew resource management, TEM, and changes to the instructor’s role. It enables fully remote IOS. It supports XR technologies. It’ll be disruptive and bigger than most expect.
At long last we will be able truly to say we “train as we fly” and “fly as we train”.
References:
Jonathan Allsop, Richard Keeling. (2023). RAF Central Flying School Sleaford, Lincolnshire, UK. On Approach to Reality: The Impact of a Simulated Air Traffic Control Environment (SATCE) on Workload and Situational Awareness in Military Aviators. 2023 Interservice/Industry Training, Simulation, and Education Conference (I/ITSEC). Paper No. 23166. https://www.researchgate.net/publication/379957057_On_Approach_to_Reality_The_Impact_of_a_Simulated_Air_Traffic_Control_Environment_SATCE_on_Workload_and_Situational_Awareness_in_Military_Aviators
Ashleigh Merritt, Ph.D. & James Klinect, Ph.D., December 12, 2006. Defensive Flying for Pilots: An Introduction to Threat and Error Management. The University of Texas Human Factors Research Project. The LOSA Collaborative. https://skybrary.aero/sites/default/files/bookshelf/1982.pdf
Varney M, Goodman J. (2023) Bolstering safety and resilience with an AI-based pilot training technology. Industry white paper. Halldale Group. Halldale article: https://www.halldale.com/civil-aviation/21451-cat-industry-calls-for-mandatory-adoption-of-satce-in-pilot-training
Comentários