A digital twin architecture for the provisioning of heterogeneous IoT devices

As IoT ecosystems expand in scale and complexity, enterprises are challenged in managing the provisioning of heterogeneous IoT devices with varying capabilities, protocols, and onboarding requirements. Provisioning, a necessary phase in the IoT device life cycle, enables unique identity assignment, network connectivity, and service authorisation. However, IoT systems currently reported in the literature often rely on vendor-specific tools or lack scalability and maintainability, which are required to support the current diversity of IoT devices. To address these limitations, this paper presents a digital twin architecture for the provisioning of heterogeneous IoT devices. The architecture presented here was designed to support diverse IoT device types, each with unique provisioning requirements, while enabling efficient provisioning in an industrial context. The architecture enables interoperability with diverse device communication protocols, provisioning methods, and characteristics. By integrating, as is typical for digital twins, the IoT device’s representation with decision-making and control capabilities, the digital twin acts as a contextually aware orchestrator. Using a proof-of-concept implementation, the paper demonstrates the maintainability of the architecture by supporting both request/response (gRPC) and publish/subscribe (MQTT) communication patterns. The proof-of-concept implementation also exhibits scalability, simultaneously provisioning up to 200 representative IoT devices with minimal impact on the provisioning interval. While the system shows promise in maintainability, further optimisations are required to reduce complexity when adapting to new onboarding and provisioning processes.