A multi-agent tuple-space based problem solving framework

Abstract

Many real-life problems are inherently distributed. The applications may be spatially distributed, such as interpreting and integrating data from spatially distributed sources. It is also possible to have the applications being functionally distributed, such as bringing together a number of specialized medical-diagnosis systems on a particularly difficult case. In addition, the applications might be temporally distributed, as in a factory where the production line consists of several work areas, each having an expert system responsible for scheduling orders. Research in cooperating intelligent systems has led to the development of many computational models ( Huhns and Singh 1992, Hewitt 1991, Gasser 1991, Polat et al. 1993, Polat and Guvenir 1993, 1994, Shoham 1993 Sycara, 1989) for coordinating several intelligent systems for solving complex problems involving diverse knowledge and activity. In this paper, a system for coordinating problem solving activities of multiple intelligent agents using the tuple space model of computation is described. The tuple space based problem solving framework is implemented on an Intel Hypercube iPSC/2 allowing multiple rule-based systems performing their dedicated tasks in parallel. The tasks are interrelated, in other words, there exists a partial order among the tasks which have to be performed.