Taskability Graph: Towards analyzing effort based agent-agent affordances

Affordance analysis, what something/someone can afford or offers, is an important aspect for day-to-day interaction and decision-making. In this paper, we will enrich the notion of affordance by incorporating agent-agent affordance: what does an agent afford for another agent in terms of a task. Further, we will present an effort hierarchy and derive the concept of Taskability Graph, which encodes: what all agents could do for all other agents, with which levels of mutual-efforts and at which places. This makes the robot more aware about agents' abilities and facilitates to develop better interaction and decision-making capabilities. We will discuss the potential application in effort based shared cooperative planning.