Robot swarms use on-board processing to plan strategy
Image credit: University of Bristol
Teams of autonomous robots that work out for themselves how to tackle tasks from environmental monitoring to disaster recovery are a step closer.
The ‘Teraflop Swarm’ groups developed by researchers from the University of Bristol and the University of the West of England can adapt rapidly through a process of artificial evolution thanks to high-performance mobile computing power that is embedded in the robots themselves.
Previous approaches have been limited by the need to perform processing on an external computer before communicating strategies to the robots.
Freed from that constraint, the researchers claim, swarms are able to reach a high level of performance within just 15 minutes. This creates new potential applications in environmental monitoring, disaster recovery, infrastructure maintenance, logistics and agriculture, with robots able to continuously and independently adapt in the wild.
Starting from scratch, a robot swarm could discover a suitable strategy directly in situ, and adapt it when the task or environment changes.
The evolved controllers are more straightforward for humans to understand too, making it easier for them to be queried, explained and improved.
Dr Sabine Hauert, senior lecturer in robotics in the Department of Engineering Mathematics and Bristol Robotics Laboratory, is one of the authors of a paper on the work published in the journal Advanced Intelligent Systems. She commented: “This is the first step towards robot swarms that automatically discover suitable swarm strategies in the wild. The next step will be to get these robot swarms out of the lab and demonstrate our proposed approach in real-world applications.”
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.