The ultimate goal of the work presented in this article is to develop practical frameworks for formally reasoning about multi-agent systems. Such frameworks are particularly important, as multi-agent approaches are inherently complex, and are already being applied in safety-critical domains such as air traffic control. The article is in three parts. The first contains an informal overview of MyWorld, a testbed for experimentation in Distributed Artificial Intelligence (DAI). A MyWorld system contains a number of concurrently executing agents, each of which is programmed along the lines proposed by Shoham in his Agent-Oriented Programming (AOP) proposal . The second part of the article contains a detailed formal model of MyWorld, which rigorously defines the possible states and state transitions of MyWorld agents. The third part develops a logic that can be used to represent the properties of MyWorld systems; this logic is closely related to the formal model of MyWorld, in that the histories traced out in the execution of a system are used as the semantic basis for the logic. We comment on the applications of the logic, and conclude by indicating areas for future work.
Wooldridge, M. (1995). This is MYWORLD: The logic of an Agent-Oriented DAI testbed. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 890, pp. 160–178). Springer Verlag. https://doi.org/10.1007/3-540-58855-8_10