The global manufacturing scenario is continuously posing new requirements on production system adaptability. Increasing volatility in the global and local economies, shortening product life cycles, increasing degree of product customization, call for production systems that comply with these changing demands in all their basic functions, including quality and production control. Current Zero Defect Manufacturing approaches are local solutions, in the sense that they are focused on single production stages. They are also static and sequential, in the sense that when a problem is analysed and solved at a specific stage, the company tends to consider the process as ‘frozen’ and moves the attention to a new critical stage that may have aroused. This long sequential strategy prevents the company from quickly adapting its production operations to changing production targets, thus undermining its competitiveness on the global market.
The proposed ForZDM project aims at developing and demonstrating a next generation Zero Defect Manufacturing (ZDM) strategy capable of dynamically achieving the production and quality targets grounding on an integrated quality and production control solution for multi-stage systems. Tools for efficient and fast multi-sensor data gathering and knowledge capturing, grounding on advanced big data management and data analytics, will be developed with the objective to identify significant inter-stage quality correlations among product, process and resource variables in multi-stage systems, also considering codified human feedback in the system, as a precious source of data. Inter-stage correlations are hidden enemies for manufacturers using complex multi-stage systems, because they are extremely difficult to be identified and effectively managed, without the support of advanced tools. They cannot be identified at single process level, as the effect of variation in one stage is only visible at downstream stages. Meeting this first objective in ForZDM will translate into a fast and effective identification of product variation patterns and root causes for defects, and will open a wide range of new opportunities for manufacturers.
ForZDM has the second main objective to develop and provide the tools to enable defect generation and propagation avoidance solutions, proposing new cyber physical systems, integrated in a distributed control platform, which will elaborate the sensorial data and, according to the inter-stage correlation models, will trigger alarms and will adjust the process, grounding on an bi-directional horizontal, machine-to-machine, information flow.
However, modelling the inter-stage correlations and managing solutions for zero defect generation and propagation is not sufficient to enable European companies to deliver on time the required quantity of high quality products satisfying customer requirements, in highly dynamic demand contexts. Indeed, the risk is to polarize the limited company efforts on quality goals, overlooking due-time performance, especially in small lot productions. Avoiding local improvement actions that are detrimental at system level is the third main objective of ForZDM. The selection of the most suitable defect prevention/correction and inspection policy will be based on the analysis of the impact of the action on the overall economic, production logistics and quality performance of the entire process-chain, thus supporting flexible adaptation to evolving product variants and production targets. Long ramp-up times while switching the production among product variants will be avoided. For achieving this development, ForZDM will ground on a bi-directional vertical integration of information, linking the shop floor level to the decision making and strategic level, thus achieving the proper integration of quality and production control, at system level.
