Abstract. This paper presents the OOA design step in a methodology which integrates automata-based model checking into a commercially supported OO software development process. We define and illustrate a set of design rules for OOA models with executable semantics, which lead to automata models with tractable state spaces. The design rules yield OOA models with functionally structured designs similar to those of hardware systems. These structures support modelchecking through techniques known to be feasible for hardware. The formal OOA methodology, including the design rules, was applied to the design of NASA robot control software. Serious logical design errors that had eluded prior testing, were discovered in the course of model-checking. 1

This paper presents an algorithm for learning the meaning of messages communicated between agents that interact while acting optimally towards a cooperative goal. Our reinforcement-learning method is based on Bayesian filtering and has been adapted for a decentralized control process. Empirical results shed light on the complexity of the learning problem, and on factors affecting the speed of convergence. Designing intelligent agents able to adapt their mutual interpretation of messages exchanged, in order to improve overall task-oriented performance, introduces an essential cognitive capability that can upgrade the current state of the art in multi-agent and human-machine systems to the next level. Learning to communicate while acting will add to the robustness and flexibility of these systems and hence to a more efficient and productive performance.

Learning to communicate is an emerging challenge in AI research. It is known that agents interacting in decentralized, stochastic environments can benefit from exchanging information. Multiagent planning generally assumes that agents share a common means of communication; however, in building robust distributed systems it is important to address potential miscoordination resulting from misinterpretation of messages exchanged. This paper lays foundations for studying this problem, examining its properties analytically and empirically in a decision-theoretic context. We establish a formal framework for the problem, and identify a collection of necessary and sufficient properties for decision problems that allow agents to employ probabilistic updating schemes in order to learn how to interpret what others are communicating. Solving the problem optimally is often intractable, but our approach enables agents using different languages to converge upon coordination over time. Our experimental work establishes how these methods perform when applied to problems of varying complexity. 1

Abstract. We present automated techniques for the explanation of counter-examples, where a counter-example should be understood as a sequence of program statements. Our approach is based on variable dependency analysis and is applicable to programs written in Cimple, an expressive subset of the C programming language. Central to our approach is the derivation of a focus-statement sequence (FSS) from a given counter-example: a subsequence of the counter-example containing only those program statements that directly or indirectly affect the variable valuation leading to the program error behind the counter-example. We develop a ranking procedure for FSSs where FSSs of higher rank are conceptually easier to understand and correct than those of lower rank. We also analyze constraints over uninitialized variables in order to localize program errors to specific program segments; this often allows the user to subsequently take appropriate debugging measures. We have implemented our techniques in theFocusCheck model checker, which efficiently checks for assertion violations in Cimple programs on a per-procedure basis. The practical utility of our approach is illustrated by its successful application to a fast, lineartime median identification algorithm commonly used in statistical analysis and in the Resolution Advisory module of the Traffic Collision Avoidance System. 1