Situation-based Methodology for Planning the Commissioning of Special Machinery Using Bayesian Networks

In German mechanical engineering customized systems and integration solutions are the biggest trends which are mainly applied in special machinery. This paper shows a method to decrease test and commissioning time by using expert knowledge and by considering the risk of failing processes. In literature and practice there is a wide research on virtual commissioning. However, research on methods to optimize production is very rare for complex machinery. In the proposed method, for planning and adapting processes, the authors use heuristics because of their ability to optimize processes using expert knowledge. For the decision of the right application of a heuristic, Bayesian Networks are applied to rate and compare different alternatives. Thus, the result is a method which allows to rate the processes with the needed time and the possible risk for an elimination and a substitution of these processes. Using this method the throughput time of a laser system in production in one single commissioning process is decreased in the validation example by approximately three days.