The DigiMonet project delivered a fully digitalised optical sensor network that combines distributed fibre‑optic sensing with cloud‑based data analytics and digital twin modelling. At its core the system uses a bidirectional incoherent optical frequency‑domain reflectometry (biOFDR) technique, which is capable of measuring temperature along a fibre with high spatial resolution. The biOFDR unit is coupled to a set of fibre Bragg gratings (FBGs) that are inscribed into the sensing fibre using a femtosecond point‑to‑point writing process supplied by Engionic. This inscription method allows the creation of highly precise, non‑destructive markers that give each sensor a unique optical fingerprint. The combination of Raman‑based distributed sensing and FBG interrogation provides a robust, multi‑modal temperature measurement capability that can be deployed over long distances.
The software architecture is built around an OPC UA client‑server model. Siemens Industrial Edge Devices (IEDs) host the local configuration and operation space; the devices communicate exclusively over HTTPS/LAN, eliminating the need for legacy USB or RS‑232 interfaces. The IEDs can be replaced by virtual IEDs running on local virtual machines or in public clouds such as Google Cloud or Amazon Web Services, enabling a fully cloud‑based deployment. OPC UA microservices expose configuration, calibration, and digital twin services. The biOFDR control software, developed by LHFT, calls these services via Python batch scripts that are launched from the MATLAB‑based device control environment. Two key identifiers – a configuration ID set during field test setup and a sensor ID automatically read from the fibre – allow the system to address the correct OPC UA servers and to associate measurement data with the corresponding sensor.
Data handling is designed for resilience. Measurement files are stored redundantly on the local DigiMonet PC and on a network drive attached to the OPC UA platform. Successful cloud uploads are flagged by renaming the file extension; if a transmission fails, the OPC UA server can reset the file and restart the entire processing chain. The system supports both online and offline operation: when network connectivity is lost, the biOFDR unit continues to acquire data and stores it locally until connectivity is restored, at which point the data are automatically uploaded to the cloud.
The cloud side of the architecture is built on Siemens’ Insights Hub (formerly MindSphere). The platform receives the measurement data, applies digital twin models for data correction, and performs automated processing and fusion of sensor streams. The integration with the SIPLU CMS Condition Monitoring System via an OPC UA client allows early detection of machine faults and supports predictive maintenance planning. Visualisation is provided by WinCC Unified, which offers a real‑time dashboard for the industrial edge infrastructure.
The project demonstrated the full stack in a field test on a power‑plant generator, showing that the system can operate reliably under industrial conditions and that the digital twin services improve data quality and enable advanced analytics. The demonstration was presented at Siemens’ internal IoT@Siemens Conference in 2022.
Collaboration in DigiMonet involved several partners. Siemens AG led the development of the digital twin and cloud integration, while AOS contributed to the OPC UA microservice framework. LHFT supplied the biOFDR hardware and control software, and Engionic provided the inscription technology for the optical markers. The consortium also included SAG, which performed risk‑minimisation studies and helped deploy the sensor platform in a real‑world environment. The project ran until 31 May 2023 and was funded by the German Federal Ministry of Economic Affairs and Energy (BMWK) under grant number 0350066A. The final report, issued by Siemens AG, documents the technical achievements and outlines the potential for commercial deployment of the DigiMonet sensor network.