Reliability-based design methodology for reusable modular assembly systems

The era of mass production has introduced products with shorter lifecycles which has reduced the utilisation of the production systems. The reuse of production systems can exploit its remaining useful life while positively impacting the environmental, social and economic welfares. The environmental, social and commercial impact of a production system can be forecasted during its design phase. Therefore, it is important for the design phase to consider certain additional features that can facilitate the long-term system sustainability. In this direction, the vision of this work is to incorporate reliability-based reuse policies for generating production system designs. Literatures highlight that the lack of systematic and strategic life cycle planning has been the major obstacle for the reuse of production systems. Furthermore, the historic operational and maintenance information of the used productions systems are not being organised and stored appropriately for reuse considerations. Data-driven approaches are becoming more realistic with historic operational information of the production systems becoming more accessible with the advent of new paradigms such as; ‘Industry 4.0’ (I 4.0) and ‘Industrial Internet of Things’ (IIoT). Therefore, the advancements in I 4.0 and IIoT can be used for predicting the remaining useful life of the production equipment. This predicted information can be used during the decision-making process for generating new production system designs/ configurations and reconfigurations.This work was motivated by the current need for a holistic assembly system design methodology that encourages the reuse of second-hand assembly equipment while designing new production lines. The primary agenda behind this motivation is to exploit the remaining useful life of the second-hand equipment by extending their life over several production life cycles. It is difficult to achieve this mainly because of the lack of methods that considers the impact of maintenance actions and the corresponding costs estimated as the result of reliability planning for the production system. The current design approaches are mainly focusing on the investment costs and some average operating costs without considering the reliability degradation of individual equipment modules that are included in the design problem.Towards this end, this work proposes a new methodology with a comprehensive model to support the adaptive configuration and reconfiguration of assembly systems. The proposed model has two novel aspects and has been developed in two stages. The first stage is focused on defining a formal mathematical model for the configuration of Reusable Modular Assembly Systems (RMAS). The linear mathematical model is built to generate RMAS design solutions based on a set of assembly product requirements. The novelty of this model lies in the use of skill-based concepts for the definition of constraints, which makes this model more generic and extendible for design of both simple and complex RMAS solutions. Furthermore, the constraints are classified into appropriate categories to stress the focus on the various aspects of designing RMAS. The second stage is focused on defining a mathematical model to estimate the optimal maintenance strategy for production equipment based on its reliability degradation over time. The novelty of this model lies in the method used for the definition of constraints that simultaneously focuses on the selection of maintenance actions and the corresponding reliability improvements based on the rate of reliability degradation of the assembly equipment. The predicted maintenance actions and the corresponding costs can help the model defined in the first stage to compare them against the cost of other modules to build the optimal RMAS design. The comparison of maintenance actions and its corresponding costs between the new and used assembly equipment has a significant impact on the design of RMAS. Furthermore, the proposed model has also been used to calculate the optimal (least expensive) maintenance schedules for the RMAS design solutions generated in the first stage.‘AutomationML’ (AML) a neutral data format based on XML has been proposed for the storage and exchange of information throughout the design process. The use of AML has been proposed as a framework for the definition of assembly Product, Process and Equipment or Resource (PPR) domains. The novelty of this framework lies in the addition of new skill-based libraries that are defined within the AML context to facilitate the description of the PPR domains. All the current and previously defined skill-based libraries within AML do not provide enough details on how these concepts can be useful for the process of designing RMAS.The models proposed in this methodology has been tested with an industrial use case to verify its applicability and appropriateness. Moreover, the proposed models serve as the foundation for the prototype environment developed within the ReBORN project to enable the collaborative design of RMAS.