Inspired by the collective intelligence of termites, the TERMES system designed at Harvard is able to achieve complex tasks with simple agents.
The idea is universal across all swarm robotic projects, find a way to split complex tasks up to be solved by many, inexpensive, replaceable and simple robots, or “agents”, instead of a single system. Something that the TERMES system is able to demonstrate by constructing complex, 3D structures without the need of a central commanding unit.
“The key inspiration we took from termites is the idea that you can do something really complicated as a group, without a supervisor, and secondly that you can do it without everybody discussing explicitly what’s going on, but just by modifying the environment,” says principal investigator Radhika Nagpal, Fred Kavli Professor of Computer Science at Harvard SEAS.
This project works further on the idea of the swarm, with intelligent actions being carried out by an agent depending on the observed changes to the environment by the other agents. By have each bot working in parallel, if one starts messing up or breaks, the rest will continue on without worrying about it. These independent intelligent actions also lends itself to scalability, having a few robots or a few hundred robots will require no changes to hardware or software but would get the job done at a much faster rate. No such thing as “too many cooks spoil the broth” in swarm robotics.
For more information on the project, check out the homepage on the Harvard website. It’s also host to two publicly available publications which go into much greater detail:
Termes: An Autonomous Robotic System for Three-Dimensional Collective Construction, RSS, 2011 – Link (harvard.edu, pdf)
Distributed Multi-Robot Algorithms for the TERMES 3D Collective Construction System, IROS, 2011 – Link (harvard.edu, pdf)