Reducing Hazards in Multiagent Task Delegation

Delegation of tasks is a basic pattern of numerous interactions within multiagent systems. Since agents are selfish and exploit local knowledge about their actions as well as the environment, delegators face the hazard of losing outcome due to self-serving behavior of the “employed” agent. This research work adopts the economic theory of agency to study delegation relationships in multiagent systems and to design a mechanism for task delegation that prevents delegators from losing outcomes in such relationships. The contribution of this work is a mechanism for task delegation considering the economic situation of agents. The paper proposes the results of a simulation study to demonstrate the usefulness and efficacy of the provided artifact.