On 14 February the journal Science published an article by a Harvard robotics team, Designing Collective Behavior in a Termite-Inspired Robot Construction Team. The article, by Werfel, Petersen and Nagpal, is a wonderfully readable introduction to one of the ways in which researchers strive to expand the capabilities of robotics: by taking inspiration from biological systems, and then carefully engineering and innovating new robots that can complete a new task. In this case the authors take one what we call emergent phenomena- the powerful ability for many simple systems to collaborate together, without centralized control, to yield overall behavior that appears very deliberate and powerful. By designing robots that can carry building bricks just tall enough to surmount, the group has created a team of robotic builders that can build their own ramps, then use those ramps to build any specified, “legal” architectural structure. The robots each have local sensing, using markings on the floor and on each building brick to maintain orientation, and sensing nearby robots. The bricks are very cleverly designed so they magnetically couple without drift, and so they provide just the right feedback to the robot for it to maintain correct orientation. The result is quite fun to watch: you can see three robots building a miniature castle– and there is no big camera system overhead directing the robots.
Perhaps the most important result of this publication is the fact that emergent phenomena require innovation- and that innovation in this case includes really careful design of the robot’s physicality, from wheels to gripper, and equally carefully innovation in designing just the right building brick. Our robot future may be built by careful steps such as this, where innovators design just the right robots and just the right robo-environments where robot and environment together yield power that is hard to imagine today.