Airlines are reportedly relying on machine-learning algorithms to assist with ultra-long flights.
Air New Zealand Ltd. and Qantas Airways Ltd. are relying on software to determine fuel-efficient flight paths, avoiding stops, according to Bloomberg.
Swiss International Air Lines and Lufthansa have also begun to use A.I., including forecasting models developed by Google Cloud, according to a January Fortune report. American airlines are also among carriers investing in the sector.
Such programs are able to help pilots avoid extreme weather and have the ability to instruct them to fly slower, with mapping software designed to improve as it is used more.
The outlet said artificial intelligence is impacting much of the airline industry, from ticket sales to cockpit procedures, and Austria-based Flightkeys helps calculate about 380,000 flight plans each day.
Its mission is to meet emerging requirements of cost-optimized airline operations, trajectory-based operation and reduction of emissions.
Flightkeys co-founder and former pilot Raimund Zopp, 67, told Bloomberg that finding a minimum-cost trajectory is "really challenging" and that longer flight routes are a delicate balance of how much fuel can be put on the aircraft and how much to offload – stressing the importance of a fast and efficient system.
"It’s so complicated to find an optimum route when the systems in the aircraft are not capable of doing this. You need a system on the ground collecting a lot of data and then trying to find the minimum-cost solution," he explained. "There are so many constraints and parameters that need to be considered and you have to apply machine learning to correctly apply these constraints."
The route planner for Fightkeys is five-dimensional, factoring in latitude, longitude, altitude, time and the "probabilistic dimension."
"The most imprecise data during the planning phase is the actual time of departure — it’s not known if you’ll really depart on schedule. It makes a big difference during the planning phase," Zopp explained. "You have a lot of chances to finetune the flight when you’re in the air because all factors are known much more precisely at that point."