A Sensor-Based Study for Security Events Detection Systems On Board Autonomous Trains

Abstract

The Human wish for autonomy in vehicles goes back to the 15 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sup> century and has been the subject of numerous research. In the past decade in particular, the rise of Artificial Intelligence & Deep Learning has provided new efficient tools for self-driving transportation systems. Most of existing works focus on cars, while trains have attracted less attention. However, railway is the most interesting transportation mode in the optic of sustainability. Given the high number of passengers it may carry, an autonomous train must analyze even more accurately its environment. It must also recognize everything happening in its cars to ensure passengers security. The total absence of railway agents on-board fully autonomous trains brings requirements for a monitoring system. It would include sets of sensors for the acquisition, algorithms for analysis and telecommunication network to transfer either the data or its extracted information. Cameras are the first sensor that comes in mind as they would furnish images of the scenes, copying human vision. In addition, other signals such as sound and air composition may supply complementary or new information. The paper offers a review of sensors and their use through the scope of event detection, in the context of public transportation.