(How) Is AI Impacting Manufacturing?

Manufacturing research is broadly concerned with tasks performed throughout the product manufacturing life cycle: Decision making at each stage requires creativity, the ability to deal with complexity and coordination with other activities in the cycle. As such, manufacturing represents a fertile domain in which to apply Artificial Intelligence. The potential of AI for solving manufacturing problems was was recognized in the early 1980s, resulting in the initiation of a large number of application projects [Fox 86]. Some of the first commercially successful systems, such as XCON at Digital [Bachant & McDermott 84 84], OPGEN at Hazeltine [Freedman & Frail 86], and GENAID at Westinghouse [Osborne 86] have arisen out of the manufacturing domain. Given the plethora of manufacturing related AI projects underway in the last decade, it is time to review our experiences. In particular, it would be useful to examine the following questions: 1. Has AI made any contributions to solving manufacturing problems? 2. Are the contributions measurable, if so, how? 3. What can we do now that we could not do in the past? 4. What does it take to build, deliver and maintain "real" AI systems? 5. How much and what types of AI are causing the impact? • a little vs a lot. • sophisticated vs unsophisticated. • expert systems vs knowledge based systems 6. Is the cost worth the effort? Each of the panel members were chosen to represent a different industry.