This project is developing underwater acoustic tracking strategies for locating and tracking particular species of tagged sharks that have long distance migratory paths. A unique aspect of this project is that receivers will be onboard a team of Autonomous Underwater Vehicles (AUV), enabling them to cooperatively estimate the shark position, thereby allowing the AUVs to track and follow the shark for sustained periods of time. The AUV team's mobility and ability to modify its own path in response to shark movement allows it to obtain in-situ measurements that cannot be obtained from static receiver systems typically used for monitoring shark behavior.
Due to its interdisciplinary nature, this project will make contributions in both Engineering (AUV autonomy and control) and Biological Sciences (shark behavior characterization, life history, and habitat utilization). New approaches to state estimation are being developed. A key to creating a successful estimator for this application will be incorporating kinematic and dynamic models of shark locomotion. Novel control strategies are also necessary given the difficulty of tracking sharks that can travel at high speed. Tracking controllers don't typically consider the additional constraint of maintaining a separation distance to ensure the AUV doesn't affect shark behavior. The end goal of conducting shark tracking experiments will be the first of its kind. Additionally, much merit will be derived from the actual data the experiments yield. Specifically, the engineering within this project is driven by the goal of determining the relationships between local ocean environment variables (e.g. temperatures, current velocities) and shark behavior (e.g. migration path choices, habitat use). While there is considerable knowledge of shark physiology and behavior, there are no fine scale time studies of shark motion and its associated energy use within different behaviors.
This work includes developing new AUV sampling technologies that can be generalized and applied to studying various forms of marine life, and in addition to the ecological contributions, the research is applicable to other tasks associated with safe marine navigation, homeland security, and the military.
(AUVs) with the ability to track and follow sharks in the California coastal regions. The work involved faculty and students from Harvey Mudd College (HMC), California State University Long Beach (CSULB), and the University of Delaware. The project has been truly interdisciplinary with both students and faculty contributing from the fields of Biological Sciences and Computer Science/Engineering. Typical methods to track the movement behaviors of sharks involve catching sharks, attaching acoustic transmitter tags to them, and releasing them. After which, a researcher in a boat will track the tagged shark by listening for the tags with underwater microphones called hydrophones, and driving the boat in the direction from which the tag signal is strongest. To obtain an accurate geoposition of the tagged shark requires the tracking vessel to be located directly over top of the animal, which can result in an alteration of the sharks behavior. In addition, this method can result in less frequent and less accurate shark position measurements. As well, it is taxing on the researcher who must track the sharks for up to 3 days straight. To address these issues, the team proposed the use of multiple AUVs, each equipped with hydrophones and on board computers that can be used to estimate shark positions in real time and drive the AUVs to autonomously follow the tagged shark, while maintaining a safe distance from the shark as to not disturb its behavior. Moreover, using an underwater acoustic communication system, the AUVs can share sensor measurements and coordinate their actions to cooperatively estimate shark positions while following the shark in formation. The intellectual merit of this project includes several contributions to the underwater robotics community, including developments of new ways to coordinate multiple AUVs for tracking marine life. The research team involved was able to create new mechanisms and computer algorithms for fusing multiple types of sensor tag measurements including signal strengths, time-of-flights, and relative bearings to a tag. A new circling controller was also developed that enabled two AUVs to circle a moving target, with AUVs changing their velocity on the fly to position themselves at opposite sides of the circle. This circling controller enables the AUVs to maintain their distance from the shark so as to reduce their effect on the shark’s behavior. By the end of this project, the research team was successful in that multiple AUVs could be deployed to cooperatively and autonomously track and follow a tagged leopard shark for several hours at a time. For the first time, the team calculated 3D estimates of a shark’s position, and compared the estimates with the manual tracking methods typically employed by biologists. The broader impacts of this project have been numerous, including training of undergraduates as researchers, interdisciplinary learning between marine biologists and engineers, and outreach campaigns dedicated to educating elementary school children about robotics and sharks encouraging interest in STEM disciplines. Most importantly, the research is creating new technology that will allow us to better understand the behaviors of marine life in our coast regions. This in turn will provide new data from which policy makers can make informed decisions regarding the protection of our natural resources.
