In the event of an emergency, the pilots have no help to save the aircraft. Their decisions, sometimes disturbed by the stress of the situation, can cause an accident. Moreover, in some very critical cases, it is almost impossible for the pilot to make the best decision. This observation highlights the importance of developing a tool to help pilots land safely. This tool would predict a safe and optimal trajectory that the pilot would have had difficulty finding alone in a moment of intense stress. It would therefore significantly increase the chance of saving the aircraft. The problem of safe emergency trajectory generation raises two main issues. The first one is the computing time of an emergency trajectory. One well-known case is the landing on the Hudson river of the US Airways Flight 1549 after bird strikes caused dual engine failure. This case is very interesting because, in the space of 30 seconds, the situation went from critical to unmanageable. This accident highlights the importance to propose an efficient algorithm in terms of computing time. The second issue is the diversity of types of emergency (loss of engine power, cabin fire, depressurization, ...). They are grouped into two types. The first type is ASAP (As Soon As Possible). For example, cabin fire and medical emergency are considered ASAP. The other type is ANSA (At Nearest Suitable Airport). In this case, the selected landing site is the safest among those reachable. The work carried out has made it possible to develop an algorithm that efficiently generates a trajectory for each available airport around the aircraft’s position. It takes into account the characteristics of the aircraft, modified by a possible failure, and also the weather. It therefore proposes a solution to help the pilot.

Download the slides for this talk.Download ( PDF, 8580.01 MB)