ABSTRACT
This chapter discusses research carried out towards the development of safe and effective autonomous decision making in intelligent robotic systems, specifically for mobile robots which are employed in hazardous applications, such as landmine clearance, nuclear decommissioning, excavation for construction sites and archaeological excavations. Introducing various forms of knowledge into a robotic architecture can often make behaviour-based navigation more flexible and general. The hybrid style combines both reactive and deliberative control in a heterogeneous architecture. The chapter describes a generalised approach for designing the robotic applications by implementing real-time control system-reference model architecture (RCS-RMA). RCS is a much-generalised architectural framework and allows various control methods to be incorporated within its hierarchical layout. The selection of a specific action affects the safety management of a robotic system to a great extent. Unsafe operating conditions can arise from scenarios not explicitly related to an internal fault, but rather from the robot environment.
