Borrowing biological designs from the black ghost knifefish, engineers have built a swimming robot that reveals how the animal’s trick of vertical movement works.
Called GhostBot, the robot copies the real fish’s undulating, ribbon-like ventral fin to propel itself through the water. New high-speed experiments show how, when waves travel along the robot’s ribbon from head to tail and meet in the middle, mushroom-cloud-like jets can push it upward.
“These fish are extremely maneuverable, and we knew how they move forward and backward with their fins,” said bioengineer Malcolm MacIver of Northwestern University, who led GhostBot’s design. “What we didn’t know was how they move vertically.”
The black ghost knifefish lives in the rivers of the Amazon Basin, using a self-generating electric field to see through the murky waters. It doesn’t have traditional fins like a bony fish, nor does it sway its body like an eel to move around. Instead, it stays rigid and uses a ribbon-like fin along its belly to navigate through a maze of downed trees, stones and other underwater obstructions with extreme precision.
Emulating such maneuverability in robotic submersibles would create countless opportunities for new or more robust underwater research, says MacIver, co-author of a study detailing the vertical movement mechanics in an upcoming issue of the Journal of the Royal Society Interface. An efficient, submersible hovering robot, for example, could constantly monitor the health of a coral reef without crashing into it (most researchers hire costly divers to do the work).
“I don’t agree that nature always has the best designs, but this is a place where it’s way ahead of human technology,” MacIver said.
When a knifefish moves upward in water, it rolls two opposing waves down its fin that meet in the middle. How that action generates upward thrust — which the dense fish needs, otherwise it’d sink — was a puzzle until the research team, including mechanical engineer Oscar Curet (now at Brown University), decided to build a copycat robot.
“Animals usually never behave the way you need them to do while studying their movements,” Curet said. “With robotics that mimic animal movement, you do something over and over again without hurting the animal.”
To reveal how the black ghost and other knifefish move vertically, the scientists worked with a company called Kinea Design to build GhostBot. It took them about seven months and $200,000 to complete a finished version.
They put the forearm-sized device into a special tank able to reveal the fluid mechanics near the fin. It flowed water laced with shiny particles over the swimming robot, shined a laser-beam plane onto the ribbon fin and took footage with high-speed cameras. Computer algorithms then processed the images to map particle velocities.
“While swimming upward, two opposing jets of water meet in the middle and collide. The merged jets deflect downward and push up on the fish,” Curet said. The animal can modulate the jets to move in diagonal directions as well, he says.
“This kind of ribbon fin is something worth paying attention to, because it has independently evolved around the world many times,” said Noah Cowan, a biologist and mechanical engineer at Johns Hopkins University who wasn’t involved in the study. “It’s a very robust design by nature, and these particular animals can move with very little body bending. It may be good for moving a rigid submersible robot.”
A major robotics company interested in developing autonomous submersibles is already discussing how they can use the design, MacIver says. Attaching ribbon-like mechanical fins to submarines people fit into, however, is another matter.
“It sounds fantastic for autonomous submersibles, but we may never use it. The design seems too fragile to sit on the ocean floor, which is what customers sometimes do with our submarines,” said Patrick Lahey, president of Triton Sumarines. “For safety reasons, we have to adhere to tried and tested technologies.”
In addition to GhostBot’s unusual means of swimming, MacIver says it’s also packed with sensors emulating the real animal’s electrosensing abilities.
“They basically use their body as a big eye,” he said. “We want to connect the robot’s swimming to its sensor network and show it can autonomously find an object we want, then hover to say, ‘here it is.’”