Robust autonomy for the youBot

In line with developing the best practise in robotics(BRICS), this project implements one of the most advanced approaches to increase robust autonomy for mobile manipulators. Using the newly announced research platform youBot as an example, a robust controller with a protection layer against joint and wheel failures was developed. The protection layer consists of a set of algorithms implemented in software that is responsible for error detection and system recovery. The system includes a model-based fault-tolerant robust control algorithm to deal with transient faults and a reconfiguration algorithm for permanent failures. In addition to the development of the control algorithm, a walk-trough tutorial on the modeling of a complex robotic system in a reusable object-oriented style is presented. Simulation results show that the proposed system is successful at increasing the robust autonomy for sensing and actuation faults. Control Engineering Douwe Dresscher iv Robust autonomy for youBot Douwe Dresscher University of Twente