I spent the final year of my undergraduate studies at the Tallinn University of Technology in Estonia. I started to work at the Center for Biorobotics, an institute dedicated to multidisciplinary research on the borderline of biology and engineering, where I embarked on a project developing a flexible robotic fish from novel materials. This project ended up turning into my undergraduate thesis.

I have always been intrigued by the social dynamics of animals, and among them fish living in groups and being led by a leader: What are the criteria that separate the leaders from the followers? What are the behavioral and physical characteristics that make a particular specimen likely to be followed by others? As an engineer, I was particularly interested in approaching these questions through designing a robotic fish -- a "carrier" -- that could successfully lead other fish, by matching their expectations of what a good leader is like. Of course, in the design of artificial agents for aquatic environments the biological reference can only serve as an inspiration; the actual implementation of the robotic device is a matter of adapting the functional principles, heavily relying on creativity and intuition. I understood from the start that fully comprehending the features necessary for successful leadership is a very involved task and it could take years of research. For the purposes of my bachelor’s thesis in Mechatronics, the more limited goal was to explore the basic features that such a mechanical carrier fish must have, and to implement a robotic platform that could serve as a carrier. The complexity of the task implied that intuition was to take the main role during mechanical design. The design was put to test with a prototype, which was evaluated qualitatively in terms of smoothness of operation and life-like qualities. Then, ideas for improvement emerged and the design was refined into successive prototypes.

The end product was a biologically inspired robot fish. Instead of a rigid body, I implemented a novel material combination (silicon-polymer) that allows a flexible embodiment. The operation of a compliant tail proved to be very difficult to model theoretically: not only are compliant bodies non-linear, but the fish-flow interaction is of tremendous complexity. With the iterative approach in design, however, the result was excellent. The robot's locomotion was surprisingly fish-like; it even left the same vortexes behind as biological fish do. With help from my colleagues, I also set up a measuring environment, and by using force sensors we evaluated the characteristics of the prototype. The maximum velocity of the robot fish was quite comparable to other robotic fish built by others before.
Pictures of the process:



I examined a real fish and I used its body as a base



I made a positive form from foam from gypsum.



I made a two sided mold around the foam. With epoxi and glassfiber, I created the rigid part of the fish body (about 2/3 of the body length)



I cut and sanded the epoxi shell



I placed the electronics and other components inside. I payed attention to keep it balanced as well bounced



I fixed the positions of teh components (glued, screwed)



I isolated the electronics with glassfiber to make it waterproof



I poured silicon inside



Ariel was born



Ariel attached to a force measuring equipment..



..close-up image

built w indexhibit