A number of modeling perspectives are relevant to the design of intelligent, adaptive, and collaborative support for humans who work in complex sociotechnical systems.We first describe the separate strands of normative models of activity and communication and self-organizing m o d e l s o f o r g a n i z a t i o n a l a d a p t a t i o n t o technology, and then braid them together to arrive at a coherent framework for the analysis and modeling of collaborative work in complex systems, From this integrated framework, we propose requirements for the design of collaborative support and develop these ideas in the context of the System Workbench for Integrating and Facilitating Teams (SWIFT) architecture.
I. ~NTRODUCI-IONThe Team Engineering Analysis and Design (TEAM) pruject is a collaboration between engineers and social scientists that seeks a principled, model-driven approach to the analysis of work groups in complex problem solving situations and the design of computer support systems for such groups. To design appropriale support systems, a fundamental understanding of activities and associated media and tools is required. TEAM proposes a comperencecentered approach to modeling interaction that incorporates both analysis of current practices (a user-centered perspective that is situated and empirical) and a specification of normative models of activity and communication (a normative task-centered perspective).In this paper we discuss three separate strands of work that reflect three interrelated perspectives on analysis and modeling of activity: Activity requirements (a normative model of activity and associated system, information, and artifacts), activity production (in terms of attention, inference, and performance), and activity evolution (in term5 of activity, norms, affordances, and networks). These intenrelated perspectives are then braided together to account for team performance in a particular context The rest of this paper is organized a5 follows. Section XI describes the process of engineering design review (('ase et al., 1992) that will be used as a specific context for analysis. Section 111 describes the three different perspectives Manuscript received July 1, 1994 on activity modeling and their integration. Section IV proposes a methodology for data collection, analysis, and modeling, and Section V speculates on how such an approach can inform the design of the System Workbench for Integrating and Facilitating Teams (SWIFT), an architecture for the construction and deployment of collaborative support systems (Lu et al., 1993).
II. ENGINEERING DESIGN REVIEW AND ARMS1The specific engineering design context for this research is the design and consuuction of a facility requested by the Air Force and carried out by the Army Corp of Engineers. The following sections elaborate on the participants, the design process, and the electronic review management system that mediated interactions between designers, quality assurance teams, and the Air Force "customer."
A. Participants in the Design ProcessAn Air ...