The future autonomous trains will need to see, hear, and react to their environment to be able to operate. These capacities of detection will call for a technological leap based on sensors, i.e. cameras, radars, and lidars.
Detection refers to all the techniques for measuring or acquiring information without contact (i.e. remotely) about objects and phenomena. It thus provides an understanding of the environment on which to base decisions. Detection requires the use of three families of on-board sensors – cameras, radar, and lidar – that basically work by measuring the electromagnetic radiation emitted or reflected by objects.
The role of these sensors in automated transport and more generally in the future of mobility is to automate and enhance the observation of human beings. Without these sensors, autonomous trains (and self-driving cars) would be “blind” and “deaf”. Without them, the mobility of the future would be at a standstill.
These technologies are improving every year, and the new possibilities they are offering show how they are maturing. One key to their development is hybridisation, that is, merging the data from each sensor and processing them with algorithms.
Thus, a camera will capture an actual scene with suggestions of colour, tone, and luminosity; a lidar will reproduce a cloud of points and thus a 3D environment of this same scene with suggestions of depth and volume; and a radar – a more mature technology – will make the collected data more robust and reliable.
Totally safe decisions cannot be made unless data acquired by several sensors are combined and taken into account.