|
|
|
With the amazing growth of Internet-based communication and cooperation within and between enterprises, stronger requirements for accurate modeling and effective managing of the related processes are emerging. The research areas of computer supported cooperative work (CSCW) and workflow systems are faced with new challenges in this context. This paper gives a nice contribution in time management of workflow systems, a topic which has not received much attention till now, but that, in my opinion, is going to become more and more relevant. Workflows will have to model, for example, e-commerce complex processes, where different tasks may not only be heterogeneous and distributed, but they may also require specific time windows for their execution and their synchronization.
With respect to what current commercial systems do for time management (which is actually very little), the paper stresses the importance of temporal reasoning not only at run time, but also when processes are designed and when they are instantiated. In the paper, time constraints allow to impose lower and upper bounds on time intervals between activities in a workflow, as well as binding activity execution to certain fixed-dates (e.g., first business-day of the month). Technically, the algorithms to manipulate the constraints are derived from the operations research literature, and, more specifically from the critical path method (CPM) used for project management. The paper is well-written, even if it gets quite technical when discussing algorithms for constraint manipulation.
I see at least two interesting extensions to this work: (i) relaxing the current restriction of a fixed duration for each activity, to a range of values, leaving more freedom to the agents in charge of the task; (ii) generalizing fixed-dates to granularities and allowing distances to be expressed in terms of these granularities (e.g., shipment must begin the same or next business-day of payment). Among recent related work, I recommend [2] which partly addresses (i), [3] which is not specifically workflow related but may help in (ii), and [4], as a nice general overview of current issues in workflow management.
Copyright © 2000 by the author(s). Review published with permission.