This month, scientists published Rare material one of the most elusive Arctic sharks. The findings show that despite major technological advances in recent years, documenting marine life up close remains a tough task.
But computer scientists at MIT believe they have a possible solution: using robots.
In today’s article, a team from MIT Laboratory of Computer Science and Artificial Intelligence (CSAIL) unveiled “SoFi,” a tender robotic fish that can independently swim in the ocean alongside real fish.
During test dives at Fiji’s Rainbow Reef, SoFi swam at depths of more than 50 feet for up to 40 minutes at a time, nimbly handling currents and taking high-resolution photos and videos using (what else?) a fisheye lens.
Using its wavy tail and unique ability to control its own buoyancy, SoFi can swim in a straight line, turn or dive up and down. The team also used a waterproof Super Nintendo controller and developed a custom acoustic communication system that allowed the SoFi to change its speed and make it perform certain movements and turns.
“To our knowledge, this is the first robotic fish that can swim untethered in three dimensions for extended periods of time,” says CSAIL PhD student Robert Katzschmann, lead author of a modern paper published today in the journal . “We are excited about the opportunity to use a system like this to get closer to marine life than humans can get on their own.”
Katzschmann worked on the project and wrote the paper with CSAIL director Daniela Rus, graduate student Joseph DelPreto, and former graduate student Robert MacCurdy, who is now an assistant professor at the University of Colorado at Boulder.
How it’s working
Existing autonomous underwater vehicles (AUVs) have traditionally been tethered to boats or powered by bulky and exorbitant propellers.
In contrast, SoFi has a much simpler and lighter setup, with a single camera, motor, and the same lithium-polymer battery found in consumer smartphones. To make the robot swim, the engine pumps water into two balloon-like chambers in the fish’s tail, which act like a set of pistons in the engine. As one chamber expands, it bends and curves to one side; when the actuators push water into the second channel, it flexes and flexes in the opposite direction.
These alternating actions create a sideways movement that imitates the movement of a real fish. By changing flow patterns, the hydraulic system enables different tail maneuvers that result in different swimming speeds, with an average speed of about half its body length per second.
“The authors present a number of technical advances in manufacturing, power and waterproofing that enable the robot to move underwater without tethering,” says Cecilia Laschi, professor of biorobotics at the Sant’Anna School of Advanced Studies in Pisa, Italy. “Such a robot could help explore the reef more thoroughly than current robots, both because it can get closer to the reef more safely and because it may be better accepted by marine species.”
The entire rear half of the fish is made of silicone rubber and pliant plastic, and several parts, including the head, which houses all the electronics, were 3D printed. To reduce the risk of water entering the camera, the team filled the head with a diminutive amount of baby oil because it is a liquid that does not compress due to pressure changes during dives.
Indeed, one of the team’s biggest challenges was getting SoFi to swim at different depths. The robot has two fins on the sides that adjust the position of the fish as it dives up and down. To adjust its vertical position, the robot has an adjustable ballast chamber and a “buoyancy control unit” that can change its density by compressing and decompressing air.
Katzschmann says the team developed SoFi with minimal disruption to the environment in mind, from minimal engine noise to the ultrasonic emissions of the team’s communications system, which sends commands using wavelengths of 30 to 36 kilohertz.
“The robot is able to make close observations and interact with marine life and does not appear to disturb real fish,” says Rus.
The project is part of CSAIL’s broader work focusing on tender robots that can be safer, more tough and more agile than their hard-bodied counterparts. Supple robots are in many ways easier to control than stiff ones because researchers don’t have to worry as much about having to avoid collisions.
“Collision avoidance often leads to inefficient movement because the robot must settle for a collision-free trajectory,” says Rus, the Andrew and Erna Viterbi Professor of Electrical Engineering and Computer Science at MIT. “On the other hand, a soft robot is not only more likely to survive a collision, but can use it as information to develop a more efficient movement plan next time.”
In the next steps, the team will work on several improvements to SoFi. Katzschmann plans to escalate the fish’s speed by improving the pumping system and improving the design of its body and tail.
He says they also plan to soon operate the onboard camera to allow SoFi to automatically follow real fish, as well as build additional SoFi for biologists to study how fish respond to various changes in their environment.
“We see SoFi as a first step towards creating something like an underwater observatory,” Rus says. “It has the potential to become a new kind of ocean exploration tool and open up new possibilities for unlocking the secrets of marine life.”
This project was supported by the National Science Foundation.