[Image above] Researchers at Harvard University designed these fish-inspired robots that can synchronize their movements without any outside control. Credit: Harvard University’s Self-organizing Systems Research Group
Like many years, 2021 kicked off with a dance party. But instead of humans in Times Square, robots at Boston Dynamics headquarters showed off their fancy footwork.
The eerily human-like movements of the robots creeped some people out because of the similarity to robots in “I, Robot” and the “Terminator” franchise. However, these robots are nowhere close to harnessing that movie level of autonomy, as Aaron Saunders, vice president of engineering at Boston Dynamics, explained in an interview with IEEE Spectrum.
“As a company, we’ve explored a lot of things, but Atlas is not using a learning controller right now,” Saunders says. To create the dance, “we used simulation to rapidly iterate through movement concepts … and make adjustments to find a compatible set of moves that we could execute on Atlas.”
In other words, the robots did not learn to dance and then choose their own moves. Instead, the robots followed a well-defined set of routines given to them by the researchers—no deep learning required.
Even though this specific video of Boston Dynamics robots does not demonstrate robot autonomy, there are many examples of autonomous robots swimming around the internet—literally, in the case of some robots at Harvard University.
Fish-shaped robots demonstrate new dimension of self-organizing capability
At Harvard’s Wyss Institute for Biologically Inspired Engineering is the lab of associate professor Radhika Nagpal. Nagpal, who is also the Fred Kavli Professor of Computer Science at Harvard’s John A. Paulson School of Engineering and Applied Sciences, and her team specialize in self-organizing systems, or systems that feature large numbers of simple agents cooperating to produce complex and robust global behavior.
“Biological systems, from embryos to social insects, depend upon the coordinated behavior of large numbers of individual agents to achieve highly complex outcomes—termites build skyscrapers and embryonic cells develop into healthy babies,” Nagpal’s bio explains. “Radhika is building new types of distributed systems—from multi-modular robots and robot swarms, to vast sensor networks—that have similar capabilities.”
Nagpal has led development of several self-organizing robotic systems, including a termite-inspired robotic construction crew and a thousand-robot Kilobot swarm. Her team’s latest self-organizing system, though, brings the concept to a new dimension.
“…most previous robotic swarms operated in two-dimensional space. Three-dimensional spaces, like air and water, pose significant challenges to sensing and locomotion,” a Wyss press release explains. However, the fish-shaped Bluebots that Nagpal, Ph.D. candidate Florian Berlinger, and former Wyss Technology Development Fellow Melvin Gauci describe in a new paper are able to overcome this challenge using two cameras and three LED lights.
“The on-board, fisheye-lens cameras detect the LEDs of neighboring Bluebots and use a custom algorithm to determine their distance, direction and heading,” the press release says. Using this information, “the Blueswarm could exhibit complex self-organized behaviors, including aggregation, dispersion, and circle formation.”
In addition to investigating these complex self-organized behaviors, the researchers simulated a simple search mission with a red light in the tank. They set the Bluebots to operate using the dispersion algorithm, which led all the Bluebots to spread out across the tank. When a Bluebot got close enough to detect the red light, its LEDs began to flash. This flashing triggered the aggregation algorithm in the rest of the school, and all the Bluebots aggregated around the signaling robot.
“Our results with Blueswarm represent a significant milestone in the investigation of underwater self-organized collective behaviors,” Nagpal says. “Insights from this research will help us develop future miniature underwater swarms that can perform environmental monitoring and search in visually-rich but fragile environments like coral reefs.”
Check out the autonomous Blueswarm in the video below.