One of the primary benefits associated with robots and autonomous machines is their ability to access places and perform tasks that are either difficult or unsafe for humans. However, one reoccurring challenge is maneuvering uneven terrain and other obstacles that really can’t be accounted for or designed around. This led University of Buffalo researchers to look more closely at animals like beavers and termites in building mechanical responses to navigating demanding environments.
With applications in areas like search-and-rescue operations, space exploration, or even marine research, the team focused on developing new algorithms for these machines that will allow them to make sense of their environment and solve problems. Due to the ever-changing dynamic associated with these environments, the team looked to stigmergy, a discipline that dissects patterns, in helping to explain a wide variety of biological behaviors.
The team, led by Nils Napp (Ph.D. and assistant professor of computer science and engineering), outfitted a mini-rover vehicle with a camera, custom software, and a robotic arm capable of lifting and moving objects. They then created terrain with randomly placed rocks, bricks and broken bits of concrete to simulate an environment that might occur after a disaster such as a tornado or an earthquake.
The robot used the algorithms to monitor and scan its environment continuously, even picking up other objects in order to fill gaps so it can more easily navigate the area and reach its destination. This behavior mirrors that of a beaver building a dam or termites building a nest. By adjusting to the clues offered by its surroundings, the robot modifies its environment in order to reach the specified goal.
Image Credit: University of Buffalo News Center/http://www.buffalo.edu/news/releases/2018/06/017.html