Mobile robot of logistic assistance for a more effective, more reactive and more safe mobility of groups – BAUDET-ROB

This project follows a meeting PactePME-DGA whose purpose is to present to the DGA technological developments from SMEs associated or not to research laboratories. It is also based on the claims raised by the DGA in his paper "Strategic Research Plan and defence technology and security." From these elements it is apparent that the problem of "mobility aid to the fighter landed" is a priority for the DGA "The fighter landed must deal with a load of major equipment (ammunition, food, sleeping bag ...) that decreases its ability to act by increasing its tiredness and its reaction time and limiting his mobility and availability".
To address this problem, two types of solutions are considered:
1. The use of an exoskeleton
2. Using a robot Sherpa
This project deals within the second solution with the aim to provide answers to scientific problems posed by controlling the movement of a mobile machine (BAUDET) responsible for transporting equipment following automatically, ie without human intervention, a group of people.
The principal function proposed for the BAUDET vehicle concerned automatic tracking of a group of people carrying some of the equipment they need. However, this tracking should be smart to follow a group of people without interfering or disrupting their movements. It should enable to follow the group anywhere including areas with limited access and without imposing excessive stresses surveillance to people followed. BAUDET must be an ally and not a new source of problem!

The mission addressed by BAUDET-ROB, called "intelligent tracking of a group of persons by a mobile robot" poses various scientific issues related to autonomy of the vehicle plunged into a complex and dynamic environment. Also, one of the biggest challenges of BAUDET-ROB is being able to adapt to different contexts by managing its actions while avoiding hazardous situations. To do that BAUDET should be able to estimate its integrity and the confidence he has in his future actions.
So project partners propose new solutions based on a multi-sensor engine which allows the system to self adapt to the current situation by interpreting the scene and estimating the reliability and accuracy of this interpretation. Vehicle control is also fundamental in this project because the robot's behaviour must be adapted on real time to the context. It must also integrate the diversity of land crossed by ensuring control and security of the trajectory at each instant.