Task Planning for Auotnomous Multi Robot System Using Market Mechanism

In Autonomous Multi Robot Systems (AMRS) tasks are not always predetermined but are often dynamically introduced in the system through one of the robots or exogenously. This raises the question of task planning and distribution of tasks among robots, i.e., which robot will perform what task, and who will be responsible for making this decision. Multi Agent Technology (MAT) provides a paradigm for such a problem, which can be used in the autonomous multi robot systems, AMRS. MAT provides an environment where collective decision making by loosely connected independent agents, robots in the AMRS case, is made possible. MAT provides the communication and operation protocols for the robots in the system. Task planning and allocation in MAT is usually governed with Market Based Mechanisms (MBM) due to their simplicity and control through a reward function associated to the accomplishment of tasks. Robots assign and trade tasks by bidding on the tasks similar to market based auctions. This paper attempts to provide an overview of the task allocation and planning problems specifically in AMRS applications using MAT paradigm and MBM approaches. The aim is to provide, to the reader, an introduction to the area with suggested links for further in depth readings and future research directions.