Discrete event control of mobile robotic systems with potential applications in healthcare

Mobile robots have the potential to be very useful within healthcare environments such as hospitals and care homes. Operating semi-autonomously, mobile robots may be used as intelligent mobility aids, navigation systems, or delivery vehicles. A typical scenario envisioned is that of a patient pressing a bedside button to request an associative robotic device. The robot would then navigate to the patient autonomously. This scenario presents a number of research challenges. The robots must operate in a semi-structured environment and therefore require flexible control strategies. Many potential applications require that the robots share control with human operators. If multiple robots are to be used, for example, in supporting evacuation procedures, cooperative behaviour is necessary. Much of the research in mobile robotics for healthcare has been toward improving the local intelligence of the robots. For example, there has been work to equip robots with improved sensing systems, obstacle avoidance techniques, and navigation capabilities. As these technologies mature, there will be an increasing need for a means of ensuring cooperative behaviour among the robots. This paper describes the concept of using a Petri net based discrete event formalism in the control of mobile robotic systems. The utility of the formalism in ensuring collision- and deadlock-free behaviour with task-preserving intervention is investigated. Two experimental platforms have been used to test the concept of Petri net on-line control. Although the initial research has been carried out in the context of industrial applications, the concepts could be applied to mobile robotic systems in healthcare