Artificial Intelligence, 38 (3), 353-66.

In IJCAI-89, Vol. 1, pp. 290-296.

J. Parallel and Distributed Computing, 6 (3), 649-62.

There are three existing connection::;t models in which network states are assigned a computational energy.

There are three existing connection::;t models in which network states are assigned a computational energy.

We propose an approach for meeting real-time constraints in AI systems that views (1) time as a resource that should be considered when making control decisions, (2) plans as ways of expressing control decisions, and (3) approximate processing as a way of satisfying time constraints that cannot be achieved through normal processing. In this approach, a real-time problem solver estimates the time required to generate solutions and their quality. This estimate permits the system to anticipate whether the current objectives will be met in time. The system can then take corrective actions and form lower-quality solutions within the time constraints. These actions can involve modifying existing plans or forming radically different plans that utilize only rough data characteristics and approximate knowledge to achieve a desired speedup. A decision about how to change processing should be situation dependent, based on the current state of processing and the domain-dependent solution criteria. We present preliminary experiments that show how approximate processing helps a vehicle-monitoring problem solver meet deadlines and outline a framework for flexibly meeting real-time constraints.

Artificial Intelligence, 34 (1), 1-38.

IEEE Computer, pages 100-109, January

To be useful in practice, a factory production schedule must reflect the influence of a large and conflicting set of requirements, objectives and preferences. Human schedulers are typically overburdened by the complexity of this task, and conventional computer-based scheduling systems consider only a small fraction of the relevent knowledge. This article describes research aimed at providing a framework in which all relevant scheduling knowledge can be given consideration during schedule generation and revision. Factory scheduling is cast as a complex constraint-directed activity, driven by a rich symbolic model of the factory environment in which various influencing factors are formalized as constraints. A variety of constraint-directed inference techniques are defined with respect to this model to provide a basis for intelligently compromising among conflicting concerns. Two knowledge-based factory scheduling systems that implement aspects of this approach are described.

Artificial Intelligence, 28 (2), 225-33.