Image credit: Airbus A350 cockpit
The aviation industry is no stranger to the benefits of artificial intelligence. Advances in computing, processing power, and access to data have further fueled the use of AI and machine learning across the industry. Its impact can be seen in airline design, engineering, manufacturing, supply chains, maintenance, and, of course, customer interactions.
On the aircraft itself, most modern commercial planes contain 2,000+ IoT sensors that generate massive amounts of critical operational data each day. AI is used to check temperature and weather, to monitor flight plans, and to fly planes on autopilot. Now AI-based chatbots will talk with pilots. Airbus is creating conversational AI models to take on some pilot tasks. The natural language processing algorithms are being trained using interaction models, domain classification techniques, and data content.
Alexa, Fly Our Plane
The Airbus AI software can automatically contact air traffic control prior to landing, for example, to determine how busy the airport is. Machine learning will be used to assess other aspects of landing, such as the force with which the plane touches down. The chatbots enable natural conversation and veer away from rules-based actions and reactions, resulting in more direct and informative pilot interactions. The goal is to relieve pilot stress, which is on the rise.
Airbus already has tested the use of AI and image recognition to fly a plane. In January 2020, the airline manufacturer’s AI-based Autonomous Taxi, Take-Off and Landing (ATTOL) system controlled the fully automated take-off and landing of an A350 jetliner. Airbus states the technology will improve operations and overall performance, and the company notes that ATTOL is designed to aid, rather than replace, pilots.
Airbus will put its conversational AI capabilities to other uses as well. The software will help Airbus develop chatbots for dashboarding and visualization and to assist in onboarding application users. Airbus plans to use conversational AI in its manufacturing facilities as well, allowing employees to ask technical questions on the factory floor. The company currently uses its EcoDesignBot chatbot to provide operations managers with sustainability information.
AI in the Sky
Airbus has several AI-based initiatives for its airline customers as well. The aviation manufacturer uses AI for computer vision, anomaly detection, time series analysis, hybrid modeling, and decision making. The company’s data analytics platform, Skywise, uses AI and machine learning to collect and analyze operational data from the aircraft. The data provides predictive maintenance, health monitoring, and reliability information to aircraft owners. Skywise has been guiding Airbus owners’ operational response to COVID-19, analyzing traffic patterns and flight restrictions and swapping planes accordingly.
The use of AI is growing in Air Traffic Management, particularly in Europe where airspace is near capacity. The ability to assess historical data and current trends can provide support to air traffic controllers. It also can model ways to integrate drones in the shared skies and will lead to the development of “green” flight paths.
Image credit: Airbus
Airbus recently tested a fello’fly flight, which relies on AI to ensure a specific distance is maintained during partner flights. The partner planes fly in formation, similar to the V-shaped pattern of migrating geese, enabling the second aircraft to benefit from the wake updraft of the leader plane. The close but staggered positioning was tracked by AI on a test flight from France to Canada. Airbus reported the fello’fly flight saved six tons of CO2 emissions and predicted it could save more than 5 percent in fuel consumption. That kind of savings will really help AI take off.