It feels like the tech world has been taken over by jellyfish. We've got jellyfish drones, jellyfish robots and insane-looking giant pollution-destroying jellyfish things. Sadly, the poor prefix-less fish get lost in all that jellyfish madness. MIT scientists must have noticed, because they've created a soft robotic fish that moves just like a real fish, which is quite an achievement for a robot.
Soft robots are exactly what they sound like: robots with soft "exoskeletons," so to speak. There’s a number of advantages to soft robots, the most obvious being that they can’t hurt you too easily. In the words of MIT professor Daniela Rus, "as robots penetrate the physical world and start interacting with people more and more, it’s much easier to make robots safe if their bodies are so wonderfully soft that there’s no danger if they whack you."
More importantly (depending on who you ask), soft robots have a completely differently take on the physical world. Traditional robots are programmed so they can’t bump into anything to avoid them breaking themselves with clunky movements. Soft robots have a bit more freedom, both in range of motion and how they interact with the outside world.
The fish robots are made out of silicon, making it easy for them to bend. The silicon also allows them to twist and swim through water the way a real fish does. This allows the fish to quickly changed directions at angles as extreme as 100 degrees. The robot fish’s tail is split into two parts, and each side has a channel that can fill with carbon dioxide, which causes the tail to move in the opposite direction.
Researcher Andrew Marchese also points at another potential upside to robots like this one. He planned from the start to make the robot look like a fish, and through that, he found that the decoupling of the tail had great functional purpose. By mirroring real fish, he learned something about real fish. In Professor Rus’s words, "if you build an artificial creature with a particular bio-inspired behavior, perhaps the solution for the engineered behavior could serve as a hypothesis for understanding whether nature might do it in the same way."