Modelling and Management of Uncertainty in Production Systems : from Measurement to Decision

Abstract: The advanced handling of uncertainties arising from a wide range of sources is fundamental in quality control and dependability to reach advantageous decisions in different organizational levels of industry. Es-pecially in the competitive edge of production, uncertainty shall not be solely object of estimation but the result of a systematic management process. In this process, the composition and utilization of proper in-formation acquisition systems, capability models and propagation tools play an inevitable role. This thesis presents solutions from production system to operational level, following principles of the introduced con-cept of uncertainty-based thinking in production. The overall aim is to support transparency, predictability and reliability of production sys-tems, by taking advantage of expressed technical uncertainties. On a higher system level, the management of uncertainty in the quality con-trol of industrial processes is discussed. The target is the selection of the optimal level of uncertainty in production processes integrated with measuring systems. On an operational level, a model-based solution is introduced using homogeneous transformation matrices in combination with Monte Carlo method to represent uncertainty related to machin-ing system capability. Measurement information on machining systems can significantly support decision-making to draw conclusions on man-ufactured parts accuracy, by developing understanding of root-causes of quality loss and providing optimization aspects for process planning and maintenance.