As nuclear facilities come to the end of their operational lifetime, safe decommissioning becomes a more prevalent issue. In many such facilities there exist ‘nuclear caves’. These caves constitute areas that may have been entered infrequently, or even not at all, since the construction of the facility. Due to this, the topography and nature of the contents of these nuclear caves may be unknown in a number of critical aspects, such as the location of dangerous substances or significant physical blockages to movement around the cave. In order to aid safe decommissioning, autonomous robotic systems capable of characterising nuclear cave environments are desired. The research put forward in this thesis seeks to answer the question: is it possible to utilise a heterogeneous swarm of autonomous robots for the remote characterisation of a nuclear cave environment? This is achieved through examination of the three key components comprising a heterogeneous swarm: sensing, locomotion and control. It will be shown that a heterogeneous swarm is not only capable of performing this task, it is preferable to a homogeneous swarm. This is due to the increased sensory and locomotive capabilities, coupled with more efficient explorational prowess when compared to a homogeneous swarm.
Bridgwater, T. J. F. Characterisation of a nuclear cave environment utilising an autonomous swarm of heterogeneous robots. (Thesis). University of the West of England and University of Bristol. Retrieved from https://uwe-repository.worktribe.com/output/1490801