Control Theory Applications for Dynamic Production Systems

6 Design of Decision-Making for Closed-Loop Production Systems

As illustrated in Figure 6.1, production systems often have closed-loop topologies in which the performance, status, and condition of key components are observed and data collected are analyzed to detect trends, errors, and deviations from desired goals, commands, and targets. Decisions then are made using appropriate rules and algorithms; these can be made manually, automatically or in a hybrid combination depending upon how quickly decisions must be made, how much human experience and judgment is required, and the level of automation in monitoring and analysis. Action then is taken to implement decisions with the objective of manipulating the production system to effectively respond to trends and eliminate deviations from desired behavior. The closed-loop cycle illustrated in Figure 6.1 can take place continuously, or nearly so, or observation, decision-making, and implementation of decisions can take place at discrete intervals, with a period between decisions that can range from milliseconds to months.

Figure 6.1 Closed-loop decision-making in production systems.

Decisions in production systems often are made using algorithms and rules that can consist of algebraic or statistical calculations, detection of trends and response to them, if-then evaluations, and limits. These often are developed with the assistance ...