|
useful cases from previous editions Dalian Dyestuff Plant: Integrating an Expert System into a Decision Support System The Dalian Dyestuff Plant is one of the largest chemical dye plants in China. It includes 11 workshops that produce about 100 kind of dyes and other chemical materials sold as semifinished and finished products in domestic and foreign markets. With a governmental shift toward decentralization, in 1983 Dalian's management had to face the new problem of setting production goals for each product. To support these decisions, Dalian built an integrated decision support system, including five subsystems that monitored costs, inventories, and working capital, and supported production planning. The production planning system used a linear programming optimization model for annual production tradeoffs between different products. Monthly production planning creates an output plan by product that attempts to satisfy existing orders plus estimates of additional demand. Because a mathematical optimization would be too complicated, Dalian chose to use an expert system containing four rule sets: rules about which products should be produced regularly, rules related just to finished products, rules covering relationships between products, and rules related to adjusting the production plan. An example of a rule used: if the demand for semifinished product exceeds the amount available, then reduce the varieties of end products except the products that need chlorine as a raw material. The plan is usually developed in two stages, first creating a rough plan and then refining the answer using rules for that purpose. The overall system was first installed in 1986 and has worked well since then. The plant manager estimates that the system accounted for around 10 percent of the plant's $10 million profit. Questions:
Source: Yang, De-Li, and Weiqin Mou. "An Integrated Decision Support System in a Chinese Chemical Plant." Interfaces, Nov.-Dec. 1993, pp. 93-100.
Back to Useful Cases from Previous Editions
|