Reference architecture for IoT-based predictive maintenance systems using digital twins

Context: Predictive maintenance strategies have become essential to industries where asset quality, product quality, or downtime critically impact costs. Internet of Things (IoT) technologies can be leveraged for predictive maintenance strategies when developing new systems or retrofitting existing systems. Additionally, predictive maintenance strategies are accelerated by real-time synchronized, high-resolution representations of physical objects called Digital Twins. Developing IoT-based predictive maintenance systems is complex, and as such, industrially validated guidelines for the development of such systems are preferred. Objective: The main objective of this study is to develop and empirically validate a Reference Architecture for IoT-based Predictive Maintenance Systems using Digital Twins. Method: We proposed a method to create a Reference Architecture for IoT and Cloud-based Predictive Maintenance Systems using Digital Twins. We have applied a systematic casestudy protocol to validate the reference architecture’s usefulness. We derived application architectures for two real industrial case studies and surveyed practitioners. Result: We demonstrated that the methods of creating a Reference Architecture could be used in the Digital Twin-based predictive maintenance domain and showed how an Application Architecture could be designed in this context. The survey results indicate that the respondents were satisfied with the practicality of the Reference Architecture and derived Application Architecture. Conclusion: The proposed Reference Architecture for IoT-based Predictive Maintenance systems using Digital Twins can be practically applied and used in diverse application domains. Furthermore, the Reference Architecture supports the potential of a predictive maintenance system to enhance operational resilience in manufacturing systems and energy infrastructure.