Coral Reefs in Decline (Source: The Guardian)

In recent years, coral reefs have become severely threatened by a shifting global ecosystem. Damage to reefs has been occurring at an alarming rate, with over 50% being lost in the past few decades.

A fundamental factor in addressing coral decline lies in the development of improved assaying and sampling methodologies. Current assessments of reef health (video transects, diver sampling, or ROVs) face notable limitations in dexterity and deployment. This includes prohibitively high costs, low efficacy, and risk of damage to the marine ecosystem.

In the search for novel solutions in coral reef conservation, the development of a bio-inspired robotic fish has been highly promising. Wang (me, Stanley!) et al. demonstrated the potential capabilities of such a platform in a robot inspired by the Sunburst Butterflyfish (Chaetodon kleinii) with a soft underactuated tail actuator. This system demonstrated agile swimming locomotion and floating body force application–both critical functionalities for the design of a dexterous underwater manipulator.

Tearing forces applied to a compliant prey based on tail actuation parameters (Wang (me, Stanley!) et al.)

Although prior studies have established fundamental mechanical principles quantifying the kinematics and dynamics of soft underwater actuators, control and intelligence of these robots is an emerging field that requires further investigation. In this project, we explore the application of EECS/ME 106A principles toward the design of a comprehensive mechatronics project with sensing, planning, and actuation. Specifically, we seek to improve the ability of an underactuated fish robot to intelligently navigate a complex underwater environment.

Our work is significant in furthering the capabilities of robots for novel underwater applications. The deployment of a soft under-actuated system in a viscous media yields highly nonlinear and nonholonomic dynamics. We successfully demonstrate feedback control of such a system in the execution of swimming trajectories. This is a crucial component in furthering the navigation abilities of fish robots, enabling their future deployment for coral reef conservation.