Internship Master 2 Research
1. Context
Robotics is increasingly present in the industry. Cobots (collaborative robots), working in collaboration with humans, are revolutionizing the industry. However, many technical problems related to the consideration of human factors in the perception, action and execution loop of this type of robot hinder their acceptance and use in an industrial context. Indeed, the cobots still remain little or not flexible unlike human operators. Detecting human reactions, such as fatigue, affects or movements, is a first step towards adapting robots in close interaction with humans. An example of adaptation could be to propose to the operator to program a new behaviour in case of detection of a facial expression or a surprise movement.
2. Objectives
The objective of this Master 2 research is: (1) to identify human reactions relevant for robot programming in cobotics, in particular facial expressions and operator movements will be studied; (2) to propose an experiment to be able to quantify them (the experiment can be carried out at the LIG on the Baxter or Yumi robot), (3) to propose an implementation integrating the observed operator feedback in order to make the robot more flexible and adaptable to the human operator.
A first experiment on the study of facial expressions and movements is already underway. A first task could be to analyse these initial experimental results in order to consider the relevant human reactions for the rest of the work.
This work will be conducted in the context of the contract ACobot.