Reviewed by
Jong W. Kim
University of Central Florida, USA
Implementing a multi-agents system rather than a single-agent system is a step-forward to understand teamwork or cooperative activities in dynamic environments. Unfortunately, the BDI software models presently included insufficient features to look seriously at the case of synthetic agents that interact with other agents, including humans. The authors of this book extended Bratman's theoretical accounts (Bratman 1992) to provide important methodological and practical underpinnings for the modelling of teamwork in dynamic environments and so doing they made an important contribution toward the step-forward mentioned above.One of the important aspects was to understand the awareness of agents, that consisted of capabilities of observation, communication, and reasoning. The authors convincingly did it by expressing agent awareness in terms of beliefs, so solving the problem of dealing with computational limitations of perception. This is a good idea and an important contribution that will motivate future developments.
When authors dealt with collective intentions for teamwork, it was essential to consider how collective intentions in a team were formed, so that a common goal could be achieved by the team. The authors viewed the coordinated team activity from a "practical reasoning" perspective, i.e., it involved a cycle of reasoning processes, including repeated updates on beliefs about the environment, decisions on options available and mechanisms to filter them to determine new intentions, creating commitments based on the newly created intentions, and performing actions in term of commitments. Specifically, from a formal approach, this book fundamentally answered two critical processes: deciding what goals were to be achieved and how they were achieved.
To bridge the gap between collective intentions and team action, the authors introduced the idea of collective commitments, suggesting a tuning mechanism. In so doing, the collective commitment guided teams comprised of collective intentions to goal's achievement.
A considerable contribution of this book is that it provided an introduction to the dynamic aspects of teamwork. The authors extended the static TeamLog to deal with reconfiguration in dynamic environments. This was possible by adopting the aforementioned definitions of intentions and commitments from a teamwork perspective. Furthermore, following the four stages of teamwork suggested by Wooldridge and Jennings (1999), including potential recognition, team formation, plan generation, and team action, a reconfiguration algorithm and its example were thoroughly presented that provided a fundamental differentiation between collective intentions and commitments, with aspects of practical understanding that the readers won't but appreciate.
Another important point is that the authors presented a case study of disaster management. The case required the modelling of continuous processes in a unique circumstance - i.e., adaptation toward dynamic and unpredictable environments - where collaborations of individuals and groups were the key-aspect to the collective capability of effectively responding to the dramatic challenges of disasters.
In sum, the prospective contribution of this book is expected to have an impact on a number of aspects that revolve around the modeling and simulation of teamwork, both from theory and application issues. Therefore, I recommend it to researchers in computer sciences, organization theory and social psychology who are interested in collective thinking, collaborative work and organizational phenomena, as well as to practitioners who deal everyday with the challenge of managing teams in many different settings.
ReferencesWOOLDRIDGE, M, and Jennings, N (1999) The Cooperative Problem-Solving Process. Journal of Logic and Computation, 9, 563-592
Return to Contents
of this issue
© Copyright Journal of Artificial Societies and Social Simulation, 2011