AI research is conducted at a number of academic and research units at the Georgia Institute of Technology. Some of this research is basic in nature, and some has an applied character to it. This article briefly describes basic AI research in the College of Computing at Georgia Tech.

This article describes a causal expert system based on hypothetical reasoning and its application to the maintenance of the lower hoist of a Mark 45 turret gun. The system, Hoist, performs fault diagnosis without the use of a repair expert or shallow rules. Its knowledge is coded directly from a structural specification of the Mark 45 lower hoist. The technology reported here for assisting the less experienced diagnostician differs considerably from normal rule-based techniques: It reasons about machine failures from a functional model of the device. In a mechanism like the lower hoist, the functional model must reason about forces, fluid pressures, and mechanical linkages; that is, it must reason about qualitative physics. Hoist technology can be directly applied to any exactly specified device for the modeling and diagnosis of single or multiple faults. Hypothetical reasoning, the process embodied in Hoist, has general utility in qualitative physics and reason maintenance.

It (McDermott 1988), and the idea of generic also helps to explicitly focus on how to go tasks and task-specific architectures (Chandrasekaran from the knowledge level to the symbol or 1983). These various proposals are program level. I call this in-between level the obviously related to each other, which makes knowledge-use level. At the knowledge-use it desirable to construct a synthesis that combines level, we focus on issues such as how the their strengths. Such a synthesis is presented overall task will be decomposed into manageable here in the form of a componential subtasks, what ordering will be imposed framework. The framework stresses modularity on the tasks, what kind of access to knowledge and consideration of the pragmatic constraints will be needed (and, consequently, what of the domain.

Stochastic simulation is a method of computing probabilities by recording the fraction of time that events occur in a random series of scenarios generated from some causal model. This paper presents an efficient, concurrent method of conducting the simulation which guarantees that all generated scenarios will be consistent with the observed data. It is shown that the simulation can be performed by purely local computations, involving products of parameters given with the initial specification of the model. Thus, the method proposed renders stochastic simulation a powerful technique of coherent inferencing, especially suited for tasks involving complex, nondecomposable models where “ballpark” estimates of probabilities will suffice.

The development of an online turbine generator diagnostic system is described from conception to initial field verification. The system is composed of a data center located in the power plant that collects data from online measurement devices and communicates these data to a centralized diagnostic facility in Orlando, Florida, where the actual diagnosis is done. The resulting diagnosis and recommended actions are transmitted to the power plant where they are displayed to the operator by the data center. The market-place need, initial approaches to the product, system field verification are described. The artificial intelligence (AI) diagnostic program has been diagnosing seven large utility generators since July 1984 and has correctly diagnosed a significant number of generator and instrumentation problems. Issues such as a centralized approach, rule base quality control, and the range of resources needed for a successful product are discussed.

Most high-performance expert systems rely primarily porations, describe the reasoning styles currently used by on an ability to represent surface knowledge about associations people, and show how some of these assessments can be between observable evidence or data, on the one addressed by extending existing AI techniques. Although the present generation of practical systems qualitative causal models in an expert system-remains a shows that this architectural style can be pushed speculative subject. The larger firms are subject to intense captured in the second model would be selected to complement scrutiny by armies of financial analysts, and even the the associational knowledge represented in the first smaller corporations have creditors of various sorts who module. The details of Simulation models have been especially attractive the procedures used to make assessments vary according choices for the complementary representation because of to the specific objective of the analyst. It might be that an the causal relations embedded in them (Brown & Burton, equity investment is under consideration, that a loan request 1975; Cuena, 1983).

This paper describes a device-independent diagnostic program called dart. The resulting generality allows it to be applied to a wide class of devices ranging from digital logic to nuclear reactors. Although this generality engenders some computational overhead on small problems, it facilitates the use of multiple design descriptions and thereby makes possible combinatoric savings that more than offsets this overhead on problems of realistic size.

Among the difficulties in evaluating AI-type medical diagnosis systems are: the intermediate conclusions of the AI system need to be looked at in addition to the "final " answer; the "superhuman human" fallacy must be guarded against; and methods for estimating how the approach will scale upwards to larger domains are needed. We propose to measure both the accuracy of diagnosis and the structure of reasoning, the latter with a view to gauging how well the system will scale up.

Among the difficulties in evaluating AI-type medical diagnosis systems are: the intermediate conclusions of the AI system need to be looked at in addition to the "final " answer ; the "superhuman human" fallacy must be guarded against ; and methods for estimating how the approach will scale upwards to larger domains are needed. We propose to measure both the accuracy of diagnosis and the structure of reasoning, the latter with a view to gauging how well the system will scale up.

Consider a class of computing problem for which all bananas is left as an exercise for the reader, or the sufficiently short programs are too slow and all sufficiently monkey. When it has been possible to couple causal models problems of this kind were left strictly alone for the first with various kinds and combinations of search, twenty-years or so of the computing era. There were two mathematical programming and analytic methods, then good reasons. First, the above definition rules out both evaluation of t has been taken as the basis for "high road" the algorithmic and the database type of solution. In "low road" representations Second, in a pinch, a human expert could usually be s may be represented directly in machine memory as a set found who was able at least to compute acceptable A recent pattern-directed allocation, inventory optimisation, or whatever large heuristic model used for industrial monitoring and control combinatorial domain might happen to be involved.