Developed by software-first aerospace company Airspace Intelligence, Flyways AI is currently being used by Alaska Airlines. Experts say it’s one of the first instances of AI being used in the operation of an airline, and it’s proving to be a game-changer for how airlines plan flights and for passengers arriving at their destination on time. “It really was an eye-opening thing to see that there was a way to do flight planning that was so different from the way the industry approached it before,” Pasha Saleh, Alaska Airlines’ director of flight operations, strategy, and innovation, told Lifewire in a phone interview.
Behind the AI
Alaska Airlines began exploring and developing how to help the dispatching process two and a half years ago. The idea was to look at how AI and machine learning could potentially work on the routing side of flying and if the technology could help make dispatchers’ jobs a little easier. “The dispatcher’s role begins about two hours before every flight pushes back from the gate,” Saleh said. “They develop a mental picture of what the path will look like by manually looking at a variety of factors.” Dispatchers are tasked with deciding that a particular path is safe to fly, that the amount of fuel the plane is carrying is enough, that the way that bags and people are boarded and loaded is safe, and more. By applying machine learning, Flyways creates prediction models of the US air space that can accurately predict the weather, traffic, and other constraints impacting flights. “With this kind of predictive modeling, Flyways can provide its airline partners with the ability to look into the future as far as eight-plus hours ahead,” the Airspace Intelligence team wrote to Lifewire in an email. The tech behind the AI works like this: Flyways uses dynamic mapping using a real-time 4D map from the Federal Aviation Administration’s data feeds, turbulence reports, and weather reports. The system then looks at all the scheduled and active flights across the US from all different airlines, scanning air traffic systemically, rather than focusing on a single flight. “It continuously and autonomously evaluates the operational safety, air-traffic-control compliance, and efficiency of an airline’s planned and active flights,” the Airspace Intelligence team said. “The predictive modeling is what allows Flyways to ’look into the future.'”
Making the Sky Easier to Navigate
The results of using Flyways AI were significant for both Alaska Airlines and its passengers. “During the six-month trial, our flights were reduced in time by 5.3 minutes, on average,” Saleh said. “And so obviously, flying less means less fuel burning which means less CO2 emitted.” Specifically, Flyways enabled Alaska Airlines to save 480,000 gallons of fuel and avoid 4,600 tons of carbon emissions, which will help the airline’s goal of net-zero carbon emissions by 2040. The reliability of planes taking off and landing was also much more accurate as a result, which frequent flyers should be excited for. “We foresee more predictability for travelers [with AI],” the Airspace Intelligence team added. “Airline passengers will have more predictable travel experiences, less bumpy flights, and a higher likelihood of arriving on time at their destination.” Saleh notes that while these results are impressive, the AI will never fully take over a dispatchers’ important job and professional judgment, but rather, become a very powerful tool they now can use at their disposal. “Dispatchers can choose to accept [the AI recommendations] or not accept them,” he said. “During the trial, we accepted about a third of the recommendations, and for two-thirds, they decided to do their own thing. And so over time, the AI will just get smarter because of this.” The future of AI in aviation is a promising one since there are so many more ways AI can help the skies. “What we’re learning is that there are so many more applications [for AI],” Saleh said. “We’re not ready to share the details of them just yet, but we have some really exciting stuff coming up in the next few months to build on this.”
