Autonomous underwater vehicles (AUVs) for nuclear environments

Remotely Operated Vehicles (ROVs) are now being used in the nuclear industry for the monitoring of fuel storage facilities and in the clean-up of disaster areas. This project will investigate how autonomous underwater vehicles (AUVs) can identify areas of high radioactivity by tracking sensitive radiation sensors along surfaces of interest.
The student will work with a research platform (Avexis) to achieve automatic control within sensible specifications within nuclear environments such as fuel ponds. Closed-loop control of Avexis will require proper characterisation of the actuators and a design of a data fusion algorithms to achieve this target. The use of novel algorithms in data fusion and machine learning will be investigated. The student will develop both practical simulations and experimental setups to test his algorithms.
Suitability of the current actuators and their allocation must be analysed. Feedback linearization techniques must be modified to cope with positive actuation only, if the current actuators are used. Advantages and drawbacks of other suitable actuators must be explored, and the possibility of a new mechanical design should be considered to increase the performance of the system.
EPSRC research areas (alphabetical order): Artificial intelligence technologies, Control Engineering, Nuclear fission, Robotics, Sensors and Instrumentation.