UNSELFISH: UNtethered Sensor Enabled Locomotion FISH

Starting date
April 15, 2011
Duration (months)
Computer Science
Managers or local contacts
Fiorini Paolo

This projects develops the first flow sensing autonomous robot based on the biological findings about fish locomotion and lateral line sensing. Though biomimetic underwater robotics has been a popular research topic for a while and first prototypes of lateral line flow sensors are already demonstrated, there is no research or proof of concept device that investigates and uses flow information for an control of an underwater robot. The aim of this proposal is to investigate relations between flow sensing and locomotion of fish. Beyond the specific methods for flow sensing, fish robot design and control, we aim of developing a new design paradigm for sensing-locomotion codesign for reduced complexity, energy efficiency and increased robustness. In particular we are interested in how the geometry of fish, locomotion patterns, lateral line sensor distribution and sensing methods mutually depend on each other and if there are some common underlying principles that can be used when designing continuous sensing robots.
More specifically, we define the following objectives:
1. Understanding of what is the contribution of the lateral line to fish behaviour in various flow regimens.
2. Investigating the kinematics and dynamics of real fish and developing understanding what is the response of the fish to various hydrodynamic events.
3. Developing information theoretic tools for flow analysis using computational fluid dynamics simulations and experiments in controlled hydrodynamic environments.
4. Based on those findings develop an optimally designed fish robot that trades off swimming efficiency and locomotion against the cost of gaining information.
5. Develop the first integrated artificial lateral line capable of distributed sensing of flow velocity and pressure
6. Develop lateral line sensing and control algorithms for the artificial fish
7. Demonstrate the stability, maneuverability and efficiency of the fish robot in various flow regimens.
8. Compare the test results of the fish robot to the biological findings and draw conclusions about the role of active and passive behaviours of a real fish.
9. Based on those conclusions, build a methodology of sensing-locomotion codesign that enables developing robust, efficient and simple robots.

The consortium of this project is composed from experts in biology of fish lateral line sensing and fish swimming, experimental and computational fluid dynamics, underwater mechanical engineering, signal processing and nanotechnology for the artificial lateral line development.


Funds: requested
Syllabus: EUROPA - Progetti Europei

Project participants

Paolo Fiorini
Full Professor
Research areas involved in the project
Sistemi ciberfisici
Computer systems organization - Embedded and cyber-physical systems


Research facilities