The Smart Use of Data and Artificial Intelligence Are Transforming Commercial Aviation Training, Says Bhanu Choudhrie
Forget COVID for a moment. Life is well, you’re cruising at 38,000ft with a drink in hand and a bite to eat. You’re watching your favourite movie, perhaps. Your captain makes a familiar announcement, updating you on your flight and what you can expect on arrival [yes, it’s warm]. All is normal. Except, that pleasant and reassuring voice isn’t coming from the flight deck. Rather, it’s coming from behind a computer screen, many hundreds of miles away by a pilot monitoring your journey in much the same way that you or I play a flight simulator.
Granted, we are guilty of a little sensationalising and — for now — this is an extreme example, but it’s one that Bhanu Choudhrie believes isn’t entirely implausible. “If you fast forward 30 years or so into the future, I fully expect to see more autonomous flying. And remember that currently, close to 90/95% is already being done by computers, a typical pilot will fly an aircraft for no more than 15–20 minutes. You may have a pilot on board, or you may not.
“If Amazon can deliver you parcels like that, or Tesla can drive you anywhere you want to go, why should commercial aviation be any different?”
Bhanu Choudhrie
“Look at an equivalent example,” he continues. “Unmanned aerial vehicles (UAVs) have flown for the last 15 or 20 years without a pilot on board. Such is the pace of innovation that the lag time, which used to be a couple of seconds for this kind of remote control, is down to a couple of milliseconds. In many ways it would be no different to the simulators we operate on the ground. Your pilot would be in the simulator while your plane is in the air, monitoring and maintaining the systems, and they could intervene should they need to at any point. Other than that, your aircraft will do everything for you: taxi, takeoff, land and even disembark — it all happens seamlessly and digitally. If Amazon can deliver you parcels like that, or Tesla can drive you anywhere you want to go, why should commercial aviation be any different?”
Why indeed. After all, in recent years — as with every other industry sector — AI, machine learning and automation have seen increasing adoption across a host of aviation tasks from reducing operating costs for airlines, improving ticketing and customer service and ground handling, through to traffic management, delivering increased fuel savings and improved maintenance, and supporting pilots in their strategic decision making and mission management.
And, as Choudhrie goes on to explain, the same technologies are having a marked effect on the training and onboarding of new pilots, particularly in light of an evolving industry facing the reality of travel post COVID. Specifically, the AI-driven data that Choudhrie’s company, Alpha Aviation Group, provides to prospective airlines on its cadets’ training programmes can help analyse the routes they fly and specific information on flying skills, such as angle of descent, acceleration and fuel burn. For an aviation industry tentatively getting back on its feet, such information is invaluable in demonstrating which pilots are best suited for specific routes and for reducing costs.
“When it comes to simulator training, the basics are the basics,” he says, modestly. “Beyond that, technology evolution means that you can add significant value by implementing things like AI into that spectrum of training. It can take up to two years to train a student pilot, and that’s a mix of classroom training, simulator work and flying; the data that comes out of that journey is vast. Analysing that data using artificial intelligence rather than a training captain sitting next to the student just lets you be so much more accurate and get the information quicker.”
“We’re moving away from people towards technology”
Bhanu Choudhrie
Pilot training is one of the most significant costs to airlines, for good reason. But such is the granular detail of the data derived from training work that Choudhrie believes it will be a crucial factor in airline decision making in the future. “The beauty is that there’s so much data. It’s all there, and has been for some time, but nobody knew how to dissect and examine it, or to gain insight from it. That’s where AI has proved so beneficial. And, because it’s intelligent, it can pull out certain elements that even the most experienced training captains may not actually see.
“I’ll give you an example,” he continues. “Such is the focus on cost now that airlines want to know every detail of a flight. Say they’re flying seven or eight times a day between Manila and Hong Kong, the data from the simulator will show which pilot or crew consumes the most and least fuel on that journey. It will show the inputs they’re making during takeoff and landing and how they treat the aircraft when in charge of it, which is essential to forecast potential maintenance costs.”
You’d be forgiven for thinking this a little big brother-esque, but the work carried out by the likes of Alpha is anything but. Not only, for example, is such a high level of data providing insight to guide the aviation industry in one of its most perilous periods, it is improving airmanship and standards of safety across the board.
