Mixture Distributions in Autonomous Decision-Making for Industry 4.0

Industry 4.0 is expected to revolutionize product development and, in particular, manufacturing systems. Cyber-physical production systems and digital twins of the product and process already provide the means to predict possible future states of the final product, given the current production parameters. With the advent of further data integration coupled with the need for autonomous decision-making, methods are needed to make decisions in real time and in an environment of uncertainty in both the possible outcomes and in the stakeholders’ preferences over them. This article proposes a method of autonomous decision-making in data-intensive environments, such as a cyber-physical assembly system. Theoretical results in group decision-making and utility maximization using mixture distributions are presented. This allows us to perform calculations on expected utility accurately and efficiently through closed-form expressions, which are also provided. The practical value of the method is illustrated with a door assembly example and compared to traditional random assembly methods and results.