Developing hydrodynamic sensing capabilities for fish-like robots
Arne Bruns
ABI Tuesday Seminar Series
Host The Auckland Bioengineering Institute, University of Auckland |
DateTuesday, June 24, 2025 4:00 PM to 5:00 PM (GMT+12) |
Live eventThe live event will be hosted in person or can be joined virtually via this page. |
In-person locationG-floor seminar room |
Maritime Drones, just as their counterparts on land and in the air, are becoming an increasing reality and a field of rapid development. Used for ocean exploration, asset monitoring and harbour security, they withstand high pressures, salt water and rough conditions to provide information and manipulation capabilities. Biomimetic, fish-like robots offer the possibility of silent, agile and energy-efficient locomotion in this domain. To realize their potential of swimming like a fish, sensor systems are needed which provide hydrodynamic information such as flow speed and direction while being able to withstand the demanding oceanic conditions. Existing sensor systems offer high sensitivity, rivalling the biological analogue but face inherent design challenges such as high cost and energy consumption, brittleness and static pressure sensitivity preventing their deployment and proliferation. I hypothesise that these inherent challenges for open ocean deployment in previous hydrodynamic sensor systems for fish-like robots can be overcome through bioinspired capacitance-based designs, providing robust, low-cost blueprints which can be tailored to provide relevant flow information in diverse conditions. Three sensors are developed to validate this hypothesis. The developed sensors based on electroactive polymers are pressure agnostic, physically robust and shielded from electromagnetic interference. A novel membrane-based rheotaxis sensor makes angle-of-attack measurements up to 1 m/s flow speed with a resolution below 5- degrees possible. Furthermore, the possibility of integrating sensing membranes directly into a bionic pectoral fin propulsors, measuring hydrodynamic loading was demonstrated. Lastly, a dorsal fin sensor for detecting cross-body flow along a robotic fish’s spine using solid state fringe-field sensors enabled roll and yaw estimation related to the actual flow. The real-world capabilities of these sensors were successfully tested through high-pressure exposure and longterm salt-water submersion tests. The developed sensors validate the research hypothesis and present a further step in achieving fish-like hydrodynamic sensing capabilities deployable from the depth of the ocean to rivers and mangrove estuaries benefiting research, commercial and security operations.
- Office of Naval ResearchN00014-21-1-2413