Is it a fish? Is it a robot? It’s both.
Adapting the biological designs of an Amazonian fish, researchers at Northwestern University created a highly agile underwater robot that can move in ways most robots can’t. It can switch from swimming forward and backward to swimming vertically, mimicking the omnidirectional black ghost knifefish.
This type of robot can potentially be used for low-light exploration—diving in places unfriendly to scuba divers, said lead researcher Malcolm MacIver, associate professor of mechanical and biomedical engineering at Northwestern University. It could potentially solve a wide range of underwater problems, from finding pollutants in pipes to monitoring invasive species.
Medill Reports and Science in Society spoke with MacIver about the robot’s functions, applications and future.
How would you describe the advantages of the robotic knifefish?
You could think of current underwater robots as submerged bathtubs. They’re really not maneuverable at all, and they’re massive. When they crash into things, like what happened in the BP oil disaster, they cause a lot of damage — so you can’t have them around delicate things like coral reefs.
The way [they] move through the water is with big propellers that spin. That way of movement is good for high speeds, but it’s bad for low-speed control. It really doesn’t give you precise movement. Fish-like propulsion does. Fish are magnificently precise in their movements.
A problem with copying conventional fish is that it’s hard to make mechanisms that flex in robotics. The great thing is that with these fish, [their bodies stay] relatively rigid, and yet they’re super maneuverable.
Why is the development of the robotic knifefish important?
Certain things that animals do are more advanced than what we can do with technology. For example, slow-speed movement through complex spaces is something we can’t do very well in robotics.
[This project is] also part of a whole research program on understanding how sensing and movement get coupled together in animal nervous systems such as our own. This is a fundamental problem in neuroscience that we, as a species, need to know the answer to for a variety of reasons — from basic science to medical issues where the interface goes wrong between sensory systems and movement control. There are all kinds of diseases where that occurs — autism seems to be a problem with sensory processing.
What are the applications of the robot?
One is helping us uncover the answers to [the sensory processing] questions. The other is developing new, highly maneuverable, low-speed underwater robots. [These] are needed for a host of purposes, from fixing broken systems down at the ocean floor (for example, when problems happen with oil drilling), to coral reef monitoring, to exploring a sunken ship. All of these are cases where you need small, agile, highly maneuverable systems, and we don’t have one.
How might this all affect us locally?
This robot could work well for finding structural problems or pollutants within the pipes of municipal or industrial fluid systems, since it doesn’t need to turn around and has great maneuverability. Or [it could] potentially help monitor ecological issues such as invasive species like the Asian carp. If we fielded a hundred of these in an area where the Asian carp are thought to have invaded, we might be able to find the source, and we might be able to use them to take corrective measures.
The work that has caused this interest is essentially on a robot [that] is blind. It has no sensory abilities. However, we have copied the [biological] fish’s sensory system, which uses electric fields to detect and identify objects around the body.
This is a first — nobody else has developed an artificial electrosensory system. We believe it’s a beautifully well-suited mechanism for the way the animal moves. The sensing in all directions goes with the moving in all directions.
Our next step in the lab is to turn on the sensory system and plug it into the movement system. We want to close that loop, so that eventually we can have a free-swimming electrosensory robot with this ribbon fin propulsion system.
What’s the most important thing you’ve learned from this creation?
Animal bodies are beautifully adapted to sense the area space that they can move into. I discovered years ago that these animals can sense in all directions, which is very strange. It’s like having a retina that’s pulled around your entire body and being able to see in all directions at once. When I made that discovery, I asked how their mechanics, their movement abilities, are affected by that crazy way to sense. What I’ve discovered is that they are able to move omnidirectionally. With the robot and this project, we’re uncovering the precise mechanisms.