Engineers Introduce ScaFi, a Scalable Fish-Like Robot for Aquatic Exploration
Engineers have introduced ScaFi, a new robot designed to emulate fish like cod and mackerel. This aquatic machine addresses limitations inherent in conventional propeller driven underwater vehicles and existing fish inspired robots. ScaFi permits significant size adaptation from a single design, aiming to reduce development efforts for varied aquatic environments.
Propeller powered underwater vehicles assist scientists in aquatic exploration and monitoring. Their mechanical operation, however, presents distinct limitations. Spinning blades can snag on vegetation, disturb sediment, and startle marine life. This renders them unsuitable for shallow waterways, dense plant growth, or close encounters with fish. Roboticists developed machines that mimic fish, bending bodies instead of propellers.
Most fish inspired robots are built for a specific size and task. Scaling them typically means starting development from scratch. So, the engineers behind ScaFi sought a solution. Their robot is modelled on fish that concentrate body bending towards the tail for propulsion, a natural swimming method observed across a wide range of body sizes.
Nana Obayashi, an assistant professor of mechanical and aerospace engineering at NYU Tandon and a faculty member of the NYU Center for Robotics and Embodied Intelligence, headed the project as a doctoral researcher at EPFL. Obayashi highlighted the need for adaptable tools, stating that monitoring distinct environments, such as creeks and lakes, currently requires two separate robots.
ScaFi incorporates a rigid front section and a flexible fibreglass rod tail. A single motor actuates two tendons that cross near the tail's end. This system creates the S shaped bend necessary for fish like swimming motion. This mechanism forms the basis of the robot's scalability, according to a paper published in npj Robotics.
The only design parameter changing with the robot's size is the diameter of the tail rods. These rods become proportionally thicker as the robot grows, preserving consistent tail bending behaviour. And the core motor mechanism and crossed tendon system remain unchanged, simplifying the overall design process.
This design reduces engineering effort for fish like robots in diverse aquatic settings. It also provides a standardised platform for researchers to systematically study how swimming performance changes with scale, a question difficult to investigate in live animals or custom made robots.
The team constructed three robots, approximately 0.6 metres, 1.1 metres, and 2.9 metres long, to assess their swimming capabilities. The smallest robot generated swirling water patterns consistent with those produced by real fish. Operations in various environments proved successful.
Swimming motion across all three sizes closely aligned once adjusted for body scale. The authors presented this as evidence that their scaling preserved the fish like gait, despite the robots' nearly fivefold increase in length. But energy efficiency did not scale with the same precision across all models.
Field tests involved deploying the medium sized robot in a Swiss stream. The largest operated on Lake Geneva, and the smallest was placed in creeks only 15 to 30 centimetres deep. The stream test showed the robot continued swimming even after a GPS signal loss.
Energy efficiency proved harder to scale. The two smaller robots performed similarly. However, the largest was consistently less efficient, necessitating a different, more powerful motor. Researchers suggest drag and inertia may be factors, though the precise cause remains undetermined.
This indicates the team scaled the swimming motion more effectively than the energy consumption required for it. So, a similar trade off appeared in disturbance tests. The smallest robot was the most agile, but recovered slowest after being knocked off course.
Conversely, the larger robots were less nimble but offered greater stability. The researchers propose this approach, scaling based on one key structural parameter, could extend to other compliant robots, including those for non aquatic applications. Whether energetic performance can scale as successfully as motion remains an open question.
ScaFi is a fish inspired robot that can scale its length significantly from a single design.
Its flexible tail, made of fibreglass rods with a consistent motor and tendon system, allows for proportional scaling.
The robot successfully mimicked fish swimming patterns and operated in various aquatic environments.
Scaling of energy efficiency proved more challenging than scaling physical motion.
The approach may extend to other compliant robotic systems.
Source: TechXplore
