Multiple target tracking based on particle filtering for safety in industrial robotic cells

Human-robot collaboration is a key issue for the development of factories of the future, a space where humans and robots can work and carry out tasks together. Within this collaborative human-robot interaction paradigm, safety is one of most critical subjects. The presented paper describes a security framework for industrial robotic cells based on laser rangefinders able to detect and track obstacles around the robot, information that is used to modify robot's behaviour and guarantee safety. The system includes a Joint Probability Data Association Particle Filter to enhance the tracking process. The obtained results, implemented and tested in a real industrial robotic cell, show the suitability of the approach.