A dynamic and prospective model to predict and prevent errors in production processes

In the industrial production field, a highly competitive sector, companies are increasingly facing complex decision-making problems that require the use of effective decision support tools. In this research area, multi-criteria decision-making (MCDM) methods are particularly useful; methods used as an effective tool to support decision makers (DMs) in critical decision-making processes [1]. Our approach aims to propose the integration of a dynamic and prospective technique, already proposed in the literature, with a scenario analysis and a knapsack-like mixed-integer model for the optimal selection of processes on which to operate in order to reduce the expected impact of errors. In particular, the work analyzes the dynamic and prospective model formulated by [2] who proposes the Parsimonious AHP method [3] in a dynamic and prospective way. Furthermore, this work proposes the optimal selection of interventions on processes, for which the prediction of errors and their impact is more critical. The proposed approach has been tested on historical data from an international company in manufacturing sector [2]. The results demonstrate that the adoption of a dynamic and prospective MCDM model allows us to provide accurate forecasts on the possible future rankings of errors in production processes since, through a feedback system deriving from past information, the ordering of errors is continuously updated. Moreover, the integration of forecasts with an optimal selection model allows to define the intervention planning that better meets the DM requirements. The model we propose can be used at a management level as a complete decision support tool for choices relating to the planning and control of production processes that can be easily implemented in companies' Internet of Thinks (IoT) systems [4]