Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Autonomous Robot Exploration and Measuring in Disaster Scenarios using Behavior Trees

: Rosas, F.G.; Hoeller, F.; Schneider, F.E.


Sgurev, V. ; IEEE Systems, Man and Cybernetics Society -SMC-:
10th IEEE International Conference on Intelligent Systems, IS 2020. Proceedings
Piscataway, NJ: IEEE, 2020
ISBN: 978-1-72815-456-5
ISBN: 978-1-72815-457-2
International Conference on Intelligent Systems (IS) <10, 2020, Online>
Conference Paper
Fraunhofer FKIE ()

Autonomous exploration and environmental measuring using robotic systems is highly relevant in natural or industrial post-disaster scenarios. Challenging conditions and presence of unknown hazardous substances might disturb rescuers, affecting their performance or even endanger their lives. Under these circumstances, robotic systems can be deployed to reduce the workload of rescuers and improve the situational awareness of emergency services.To accomplish this, several algorithms need to be synchronized on the robots. Behavior trees offers a simple and modular structure for task execution, able to provide a deep insight of complex task flow at first glance. To cover the scenario described in this work, we combine robot navigation, exploration and complete coverage planning, 2D/3D obstacle mapping and localization of hazard substances. Our main contribution is a behavior tree-based robot execution frame, which coordinates the sub-tasks of all these applications. Our syst em was successfully tested at the European Robotics Hackathon 2019, where a hazardous materials incident response operation was staged. With this implementation, we aim to simplify the system design and application of robotic systems used for autonomous operation.