The project set out to increase the efficiency and reliability of geothermal drilling by upgrading the existing DrillSIM:600 training simulator, originally developed by the British company Drillling Systems, into a real‑time capable digital twin. The upgraded system incorporates eight high‑performance servers, eight large‑screen displays, a “Cyber Chair” for immersive operator training, and an application programming interface (API) workstation that links the simulator to field equipment via an OPC‑UA interface. This interface, widely used in the drilling industry, enables the seamless transfer of live data from an operational drilling rig at the DSC site into the simulator, allowing the model to reflect current drilling conditions in real time.

Central to the technical achievement was the development and integration of new physical and empirical models that describe key drilling parameters. Machine‑learning algorithms were trained on historical drilling data to predict rate of penetration (ROP), weight on bit (WOB), and equivalent circulating density (ECD). Validation results, presented in the report’s figures, demonstrate that the models achieve high predictive accuracy, with feature‑importance analyses revealing the most influential variables for each parameter. The ROP model, for example, was validated against a large set of drilling trials, showing a strong correlation between predicted and observed penetration rates. Similar validation procedures were applied to the WOB and ECD models, confirming their suitability for real‑time decision support. The simulator now uses these models to generate realistic drilling scenarios, enabling operators to plan critical drilling sections and to identify potential challenges before they occur.

The integration of real‑time data and advanced predictive models transforms the DrillSIM:600 into a digital twin of the drilling process. Operators can now run parallel simulations of multiple drilling scenarios, compare outcomes, and adjust strategies on the fly. The system’s ability to ingest live data from the field and to provide instant feedback reduces the time required for drilling optimization and lowers the risk of costly drilling errors. The project also introduced a collision‑avoidance model and an uncertainty‑ellipsoid representation of measurement points, further enhancing the safety and precision of drilling operations.

Collaboration was a key driver of the project’s success. The simulator manufacturer, Drillling Systems, supplied the hardware platform and technical expertise for the software upgrade. The DSC field operator provided access to real‑time drilling data and helped validate the simulator against actual field conditions. A research team—likely comprising university scientists and industry engineers—led the development of the machine‑learning models and the integration of the OPC‑UA interface. The project was carried out over the course of the funding period, which spanned several years and was supported by a German research grant, presumably from the German Research Foundation (DFG) or the Federal Ministry of Education and Research (BMBF). The combined effort of these partners ensured that the upgraded simulator met both academic and industrial standards.

In summary, the project delivered a state‑of‑the‑art, real‑time capable geothermal drilling simulator that integrates live field data, advanced predictive models, and immersive training hardware. By providing accurate, real‑time drilling forecasts and a digital twin environment, the system promises to reduce drilling costs, improve operational safety, and accelerate the transition to reliable geothermal energy production.