AbstractWith advances in actuation and sensing materials and devices, there is a growing interest in developing underwater robots that propel and maneuver themselves as real fish do. Such robots, often known as robotic fish, could provide an engineering tool for understanding
fish swimming. Equipped with communication capabilities and sensors, they could also serve as economical, dynamic samplers of aquatic environments. In this paper we discuss some of the major challenges in realizing adaptive, cost-effective, mobile sensor networks that are enabled by resource-constrained
robotic fish. Such challenges include maneuvering in the presence of ambient disturbances, localization with adequate precision, sustained operation with minimal human interference, and cooperative control and sensing under communication constraints. We also present potential solutions and
promising research directions for addressing these challenges, some of which are inspired by how fish solve similar problems.
Experiments with Cooperative Control of Underwater Robots