Role-Based Ad Hoc Teamwork

Abstract

An ad hoc team setting is one in which teammates must work together to obtain a common goal, but without any prior agreement regarding how to work together. In this paper we present a role-based approach for ad hoc teamwork, in which each teammate is inferred to be following a specialized role that accomplishes a specific task or exhibits a particular behavior. In such cases, the role an ad hoc agent should select depends both on its own capabilities and on the roles currently selected by the other team members. We formally define methods for evaluating the influence of the ad hoc agent's role selection on the team's utility, leading to an efficient calculation of the role that yields maximal team utility. In simple teamwork settings, we demonstrate that the optimal role assignment can be easily determined. However, in complex environments, where it is not trivial to determine the optimal role assignment, we examine empirically the best suited method for role assignment. Finally, we show that the methods we describe have a predictive nature. As such, once an appropriate assignment method is determined for a domain, it can be used successfully in new tasks that the team has not encountered before and for which only limited prior experience is available.

A Formalization of the Coach Problem

“…Esteva et al [16] and Dignum [14] introduce the notion of norms in combination with the agent roles already used by Odell et al [33], Ferber and Gutknecht [17] and Wooldridge [47]. Roles can be used to decompose the tasks to be performed by the MAS into sub-objectives to increase efficiency [44,18,42]. The role-based approach to ad hoc teamwork by Genter et al [18] determines role selection on the team's utility.…”

“…Roles can be used to decompose the tasks to be performed by the MAS into sub-objectives to increase efficiency [44,18,42]. The role-based approach to ad hoc teamwork by Genter et al [18] determines role selection on the team's utility. In robot soccer, roles have been based on absolute position in the field [32], position relative to the ball [2,3,31] and robot trajectories, sometimes also considering positions of other players [48].Other work on team coordination through roles and dynamic positioning can be found for example in [28,27,13].…”

“…There's related work on the use of roles to infer a robot's plans or intentions [4,5,7,18]. 'Plans' can be interpreted in various ways, ranging from a sequence of actions currently performed according to some behaviour [7,40] to a robot's entire internal state (e.g.…”

Weighted synergy graphs for effective team formation with heterogeneous ad hoc agents

“…Role assignment in an ad hoc team is considered, where an ad hoc agent has to select a role, such that it maximizes the team's overall utility based on its observations of its teammates [15]. An ad hoc agent in the pursuit domain has to vary its behavior to better suit the team's objective of capturing the target, by modeling its teammates and choosing a best response [4].…”

Using Opponent Modeling to Adapt Team Play in American Football

“…One of the other book chapters describes a new direction in Robocup research, the creation of ad-hoc teams [13]. In role-based ad hoc teams, the new agent studies the behavior of its future team mates before deciding on a policy based on marginal utility calculations.…”