The concept of placing robotic ‘animals’ in environments with living peers has been around for quite some time. However, those robots have never been capable of engaging on any level with the live animals, only recording the animals’ reaction to the newcomer. That’s no longer the case, as a robotic fish designed to look like the real thing can now see, interact with, and mimic the movements of the live fish around it.
Maurizio Porfiri, a professor of mechanical and aerospace engineering at NYU’s Tandon School of Engineering, led a team that conceived and deployed the fish. Beyond being a masterpiece of modern robotics, the fish allows a far greater ability to accurately monitor animals’ behaviors from within their natural environment.
The robotic fish represents the first closed-loop control system of its sort that has been successfully deployed in a biomimetic animal replica. The control system allows the robotic fish to be able to visualize and mimic the behaviors of live fish in real time, and the results were precisely as Porfiri’s team had hoped.
The team deployed the robo-fish among several zebrafish and gauged their interaction to the new guy. In every experimental and environmental condition, the live fish exhibited less anxiety and fear – none, in fact – toward the fish as it mirrored their movements. The live fish did exhibit those traits when a robotic fish with a pre-determined set of movements was introduced into the environment.
Mirroring is a social survival tool for humans, and animals are no different. However, a robotic anything with the ability to thoroughly observe and mirror the behavior of live counterparts is a significant deal.
This form of mirroring is a very simple social behavior, in which the replica seeks only to stay as close as possible to the live animal. But this is the baseline for the types of interactions we're hoping to build between animals and robots, Porfiri said.
The NYU robo-fish represents a dualistic feat of robotics and animal observation.
We are learning what really matters in zebrafish social interactions, and we can use this information to help the robot interpret and respond appropriately, rather than just copying what it sees, he said.