multi-robot collision avoidance

Daniel Claes, Daniel Hennes, Karl Tuyls and Wim Meeussen

We developed a multi-robot collision avoidance system based on the velocity obstacle paradigm. In contrast to previous approaches, we alleviate the strong requirement for perfect sensing (i.e. global positioning) using Adaptive Monte-Carlo Localization on a per-agent level. While such methods as Optimal Reciprocal Collision Avoidance guarantee local collision-free motion for a large number of robots, given perfect knowledge of positions and speeds, a realistic implementation requires further extensions to deal with inaccurate localization and message passing delays. The presented algorithm bounds the error introduced by localization and combines the computation for collision-free motion with localization uncertainty. We will soon provide an open source implementation using the Robot Operating System (ROS). The system is tested and evaluated with up to eight robots in simulation and on four dierential drive robots in a real-world situation.

Our team won the Best Demonstration Award at the 11th International Conference on Autonomous Agents and Multiagent Systems (AAMAS'12) with: "CALU: Collision Avoidance with Localization Uncertainty".

mobile telepresence with assisted control

Sjriek Alers, Daan Bloembergen, Daniel Hennes and Karl Tuyls

MITRO (Maastricht Intelligent Telepresence RObot) is a custom-built robot system specifically designed for augmented telepresence with assisted control. Telepresence robots can be deployed in a wide range of application domains, and augmented presence with assisted control can greatly improve the experience for the user.

uav simultaneous localization and mapping

Joscha Fossel, Daniel Hennes and Karl Tuyls

In this project we tackle simultaneous localization and mapping (SLAM) on unmanned aerial vehicles (UAVs) equipped with a 2D laser range finder, attitude and altitude sensors. In contrast to other research in this area, our approach performs SLAM on a 3D instead of a 2D map using planar scans. To that end we use a scan registration algorithm on 3D transforms of planar scans. The transform is computed by combining data from a downward facing ultrasonic sensor and inertial measurement unit. An octree based map is used to represent the 3D environment. Our scan registration algorithm is derived from iterative closest point matching and Hector SLAM. The proposed SLAM system supports both 2D and 3D maps. We evaluate the performance of our system in simulation and on a real multirotor UAV equipped with a 2D laser range nder. The results show that signicant improvements are achieved when using a 3D map for localization in the UAV domain.