A Computational Framework for Designing Interleaved Workflow and Groupware Tasks in Organizational Processes

Abstract

Most organizations have traditionally been organized by function, and most coordination is intrafunctional rather than interfunctional. However, many organizations are finding that they must also manage processes - such as order fulfillment, new product development, and interorganizational supply chain management - that span their separate functional units and that integrate their activities with those of other organizations. These processes are essential to the well-being of organizations in a dynamic competitive environment.In response to this, organizations are deploying large-scale enterprise information systems in order to support operational, tactical, and strategic decision making, along with information management. However, deployment of such information systems has not realized the requisite benefits due to issues such as lack of interoperability among applications due to technological evolution, constant changes to the business processes, evolving organizational structures, inherent complexity in management of distributed knowledge and resources.To ameliorate such issues, a recent technological trend is the adoption of support tools such as Workflow Management Systems (WFMS) and groupware to support coordination between individual and group knowledge worker activities respectively. While WFMSs mostly deal with tasks involving very structured information, groupware tools deal with tasks involving unstructured information. Due to these differences, such tools are used in a fragmented manner, causing information loss. The overall guiding design principles that can be used by such process support systems are minimal, resulting in costly overheads for organizations.This dissertation deals with the problems highlighted above from a organizational process design standpoint. The goal of the dissertation is to provide process designers with guidelines and tools that can assist them in modeling flexible and adaptable processes. The following two research questions are central to the work described in this dissertation: (1) How can organizational processes be designed to be flexible andadaptable in dynamic environments? (2) How can collaborative activities be designed to facilitate integration with individual activities in organizational processes?In this regard, this dissertation reports on the development of a conceptual framework to support design of organizational processes considering both individual and collaboration tasks in a unified manner. A business process is modeled as a problem solving mechanism consisting of a series of steps (also termed as process model, process definition or plan), each of which may be an individual or group activity. The task of designing business processes is considered as the development of an effective plan to solve a business process problem by searching the design space. We employ declarative formalisms from recent advances in Artificial Intelligence (AI) planning to support the task of process design. Similarly, we build on research in the field of Collaboration Engineering (CE), to propose an approach for collaborative task design. The feasibility and benefits of the approach are evaluated by prototyping intelligent build time tools for process design, and utilizing the same in the design of processes such as loan processing, and new drug discovery.