Underwater robots are finding wide applications in military detection, undersea operation, oceanic supervision, aquatic life-form observation and pollution search. Around the world, underwater robotic vehicles such as Remotely Operated Vehicles (ROVs) and Autonomous Underwater Vehicles (AUVs) have been designed and built to replace humans for aforementioned submarine tasks. This research is inspired by nature itself, like biomimetic swimming robots, while the experimental research on swimming animals, mainly fish, shows their strength in marine areas. Fish are highly efficient, manoeuvrable and noiseless. The propulsion system for some types of fish is up to 90 percent efficient, while it is around 40 to 50 percent for a conventional screw propeller. This project seeks two main objectives. First, a novel robot concept will be developed with the aim of agile and fast manoeuvre. Then a novel controller based on Neuro-fuzzy control scheme will be developed. This novel controller will be able to control fish movements in heading, up and down, swift turning, and high speed cruise.