Investigation of the Explicit Reference Governor Implemented in Collaborative Robots.

Project Details

Description

Collaborative robots, that will play a major role in Industry 4.0, are about to enter the manufacturing floor, but still several fundamental research questions are open to be solved. In this research proposal, the goal is to develop a novel navigation scheme in collaborative robots, i.e. robots that have to work together with humans in a cooperative environment, such that these robots will be able to slow down nearby obstacles, to avoid collisions, and to physically and socially
interact with objects and people. The navigation scheme that will be used to dynamically modify the references, is the Explicit Reference Governor (ERG), which is a novel scheme that is inspired by the Reference Governor (RG) philosophy. This ERG scheme can enforce both state and input references of a pre-compensated (nonlinear) system without relying on any form of optimization,
which makes it computationally efficient so that it can be executed in real-time. To make it usable for collaborative robots, this scheme has to be further evolve so that it can deal with non-convex and dynamic constraints. After each theoretical modification, tests will be performed in a simulation environment, whereafter experimental validations will be executed on the collaborative robots Baxter and Franka that are available in the R&MM lab at the VUB. Throughout
this PhD, several sub-research questions will be handled to make it possible to perform some industrial scenario cases experiments in the lab environment.
Short title or EU acronymOZR opvangmandaat
AcronymOZR3161
StatusFinished
Effective start/end date1/10/1730/09/18

Keywords

  • robots