Marine scientists have long marveled at how animals like fish and seals swim so efficiently despite having different shapes. Their bodies are optimized for efficient, hydrodynamic aquatic navigation in order that they can exert minimal energy when traveling long distances.
Autonomous vehicles can drift through the ocean in the same way, collecting data about vast underwater environments. Nevertheless, the shapes of those gliding machines are less diverse than what we discover in marine life — go-to designs often resemble tubes or torpedoes, since they’re fairly hydrodynamic as well. Plus, testing latest builds requires numerous real-world trial-and-error.
Researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and the University of Wisconsin at Madison propose that AI could help us explore uncharted glider designs more conveniently. Their method uses machine learning to check different 3D designs in a physics simulator, then molds them into more hydrodynamic shapes. The resulting model will be fabricated via a 3D printer using significantly less energy than hand-made ones.
The MIT scientists say that this design pipeline could create latest, more efficient machines that help oceanographers measure water temperature and salt levels, gather more detailed insights about currents, and monitor the impacts of climate change. The team demonstrated this potential by producing two gliders roughly the dimensions of a boogie board: a two-winged machine resembling an airplane, and a novel, four-winged object resembling a flat fish with 4 fins.
Peter Yichen Chen, MIT CSAIL postdoc and co-lead researcher on the project, notes that these designs are only a couple of of the novel shapes his team’s approach can generate. “We’ve developed a semi-automated process that might help us test unconventional designs that will be very taxing for humans to design,” he says. “This level of shape diversity hasn’t been explored previously, so most of those designs haven’t been tested in the true world.”
But how did AI give you these ideas in the primary place? First, the researchers found 3D models of over 20 conventional sea exploration shapes, resembling submarines, whales, manta rays, and sharks. Then, they enclosed these models in “deformation cages” that map out different articulation points that the researchers pulled around to create latest shapes.
The CSAIL-led team built a dataset of conventional and deformed shapes before simulating how they’d perform at different “angles-of-attack” — the direction a vessel will tilt because it glides through the water. For instance, a swimmer should want to dive at a -30 degree angle to retrieve an item from a pool.
These diverse shapes and angles of attack were then used as inputs for a neural network that essentially anticipates how efficiently a glider shape will perform at particular angles and optimizes it as needed.
Giving gliding robots a lift
The team’s neural network simulates how a selected glider would react to underwater physics, aiming to capture the way it moves forward and the force that drags against it. The goal: find the perfect lift-to-drag ratio, representing how much the glider is being held up in comparison with how much it’s being held back. The upper the ratio, the more efficiently the vehicle travels; the lower it’s, the more the glider will decelerate during its voyage.
Lift-to-drag ratios are key for flying planes: At takeoff, you should maximize lift to make sure it will probably glide well against wind currents, and when landing, you wish sufficient force to pull it to a full stop.
Niklas Hagemann, an MIT graduate student in architecture and CSAIL affiliate, notes that this ratio is just as useful in the event you want the same gliding motion within the ocean.
“Our pipeline modifies glider shapes to seek out the perfect lift-to-drag ratio, optimizing its performance underwater,” says Hagemann, who can also be a co-lead creator on a paper that was presented on the International Conference on Robotics and Automation in June. “You may then export the top-performing designs in order that they will be 3D-printed.”
Going for a fast glide
While their AI pipeline seemed realistic, the researchers needed to make sure its predictions about glider performance were accurate by experimenting in additional lifelike environments.
They first fabricated their two-wing design as a scaled-down vehicle resembling a paper airplane. This glider was taken to MIT’s Wright Brothers Wind Tunnel, an indoor space with fans that simulate wind flow. Placed at different angles, the glider’s predicted lift-to-drag ratio was only about 5 percent higher on average than those recorded within the wind experiments — a small difference between simulation and reality.
A digital evaluation involving a visible, more complex physics simulator also supported the notion that the AI pipeline made fairly accurate predictions about how the gliders would move. It visualized how these machines would descend in 3D.
To really evaluate these gliders in the true world, though, the team needed to see how their devices would fare underwater. They printed two designs that performed the perfect at specific points-of-attack for this test: a jet-like device at 9 degrees and the four-wing vehicle at 30 degrees.
Each shapes were fabricated in a 3D printer as hole shells with small holes that flood when fully submerged. This lightweight design makes the vehicle easier to handle outside of the water and requires less material to be fabricated. The researchers placed a tube-like device inside these shell coverings, which housed a spread of hardware, including a pump to alter the glider’s buoyancy, a mass shifter (a tool that controls the machine’s angle-of-attack), and electronic components.
Each design outperformed a hand-crafted torpedo-shaped glider by moving more efficiently across a pool. With higher lift-to-drag ratios than their counterpart, each AI-driven machines exerted less energy, much like the effortless ways marine animals navigate the oceans.
As much because the project is an encouraging step forward for glider design, the researchers need to narrow the gap between simulation and real-world performance. Also they are hoping to develop machines that may react to sudden changes in currents, making the gliders more adaptable to seas and oceans.
Chen adds that the team is seeking to explore latest forms of shapes, particularly thinner glider designs. They intend to make their framework faster, perhaps bolstering it with latest features that enable more customization, maneuverability, and even the creation of miniature vehicles.
Chen and Hagemann co-led research on this project with OpenAI researcher Pingchuan Ma SM ’23, PhD ’25. They authored the paper with Wei Wang, a University of Wisconsin at Madison assistant professor and up to date CSAIL postdoc; John Romanishin ’12, SM ’18, PhD ’23; and two MIT professors and CSAIL members: lab director Daniela Rus and senior creator Wojciech Matusik. Their work was supported, partially, by a Defense Advanced Research Projects Agency (DARPA) grant and the MIT-GIST Program.
