AI and its Application in Aircraft Runway Excursion Reduction

A. BEN ALI TCHEIKH SAID

Unknown

According to ICAO (International Civil Aviation Organization), a runway excursion (RE) is a veer off or overrun from the runway surface of an airport. In 2013, an IATA report has reported that 23% and 13% of aircraft accidents are associated to runway excursion and runway debris(RB) respectively (1). Runway excursion events occur while an aircraft is taking off or landing, and these phenomena involve many factors ranging from unstable approaches to the condition of the runway. It is important that all aviation organizations involved in the aircraft operation (Pilots, Air Traffic Controllers, Airport Authorities, etc.) work together to mitigate the hazards that result in RE and RB. ATO's Office of Runway Safety is committed to reducing RE risk through analysis, awareness, and action. Runway risk continues to increase these last years. However, runway accident is one of the major challenges of the last decades. These events should be the results of external weather conditions management, such as slush, dry snow, wet snow, frost, contacted snow, water...present on a runway surface. Basically, aircraft navigation system is defined as a method for deriving navigation parameters such as position and velocity by using radio-signals transmitted from satellites. This expensive satellite based technology currently used in the aviation industry does not permit to get in real-time, the behavior of the airport surface; and cannot detect foreign objects present in the runway surface. Different systems are currently used to interpret the pilot precision approach aid. Some system such as ILS (instrument landing system), MLS (Microwave Landing system) could provide a specified quality of guidance data from the feet limit height. Additionally, Runway visual range (RVR) is an important factor to take into consideration during a safe landing. Given the possibilities opened by some current technologies, such as the Internet of Things, Big Data, Deep Learning and its framework Convolutional Neural Network, Dynamic Vision Sensor(DVS), Weather sensor, Temperature sensors…, the aviation industry could build robust system which will allow the pilot to inspect in real-time the behavior of his future landing runway. This mobile embedded system (App) does not need to change the landing system methodology currently used by the aviation, and the main objectives of this work are to develop an app which will reduce the risks related to runway excursion-incursion and foreign objects. ...learn more

Project status: Published/In Market

Mobile, Robotics, RealSense™, Networking, Internet of Things, Artificial Intelligence

Code Samples [1]