Challenges and Opportunities of Computer Vision Applications in Aircraft Landing Gear

Computer vision in the recent decade has seen tremendous increase in widespread adoption, ranging from facial recognition used in consumer electronics such as smart-phones to navigation in autonomous vehicles. Part of this success can be attributed to advances in artificial intelligence (AI) or deep learning. As usual, it is expected that the transition of such technologies will be slowest in the aerospace industry, in particular for airborne applications, because of the stringent regulations. Certifying deep learning based algorithms is even more difficult because of its indeterministic nature. However, there exist certain benefits and opportunities when we consider some elementary yet trivial applications such as positional sensing and steering angle detection in the landing gear. Being also one of the harshest environments accompanied by brakes that can reach up to 800 degrees Celsius, it is yet another challenge to introduce these delicate hardware used in computer vision such as camera or LIDAR (Light Detection and Ranging). This paper aims to present the various constraints, challenges, opportunities and examples related to computer vision applications in landing gear. Together with the recent work from Airbus demonstrators, it is hoped to raise awareness and stimulate innovations for more computer vision applications in the aerospace industry.