As the OptiDrill project nears completion, we continue our ‘Beneath the Surface’ blog series, spotlighting the consortium partners and their unique contributions to this groundbreaking initiative.
In this instalment, we explore the role of Fraunhofer Institution for Energy Infrastructures and Geothermal Systems (IEG), the coordinator of the OptiDrill Project and one of Europe’s leading research organisations, in developing and testing OptiDrill’s innovative technologies.
We spoke with Dr Shahin Jamali, Head of Monitoring and Artificial Intelligence at Fraunhofer IEG and also the Project Coordinator, to discuss their role, the challenges faced, and the technological breakthroughs achieved during this transformative journey.

Fraunhofer IEG: A Leader in Geothermal Innovation

Fraunhofer IEG is at the forefront of applied research, specialising in geothermal systems, energy infrastructure, and artificial intelligence. With world-class laboratories, drilling rigs, and testing facilities across Bochum, Aachen, and Weisweiler in Germany, Fraunhofer IEG’s role was essential in advancing geothermal energy solutions for the OptiDrill Project.
As the coordinator of the OptiDrill project, Fraunhofer IEG led efforts to develop and validate Down-The-Hole (DTH) hammers and AI-driven monitoring systems. Their expertise in integrating machine learning, sensor technologies, and real-world testing has been critical in aligning research outputs with practical drilling applications.

Aligning Technologies: The Integration Challenge

Coordinating the project while managing multiple technologies—AI, machine learning, and sensor-based systems—posed significant challenges. Dr. Jamali reflected on these complexities, “One of the biggest challenges was the fundamentally different nature of the AI-driven components and the sensor-based hardware systems. While AI and machine learning relied heavily on historical drilling data and computational power, the sensor systems required specialised hardware components and stable supply chains. Additionally, data transfer from downhole sensors to surface-level computing systems posed technical challenges early on.”
These hurdles led to the AI and sensor systems initially evolving as separate subsystems. Dr. Jamali noted that more proactive resource allocation, early supplier engagement, and clearly defined technical standards could have accelerated the integration process.

Methodologies and Innovations

Fraunhofer IEG’s approach relied heavily on data-driven machine learning, iterative model refinement, and rigorous cross-validation. The team leveraged state-of-the-art tools like TensorFlow, PyTorch, and scikit-learn, ensuring a flexible and transparent development process.
Dr. Jamali emphasised that the trial-and-error process of testing multiple models and data strategies was time-consuming but invaluable, “We learned that not all historical drilling data can be treated as one universal dataset. Tailoring models to specific regions and refining data preprocessing strategies significantly improved performance.”
These insights shaped the direction of their work, leading to robust, region-specific predictive models that offer real-world benefits.

Progress, Breakthroughs, and Lessons Learnt

Despite these challenges, Fraunhofer IEG made consistent and meaningful progress. The team’s initial work was slowed by data availability issues, as high-quality historical drilling data was difficult to access and required extensive cleaning and structuring. However, once this bottleneck was addressed, progress accelerated. A key breakthrough emerged when the team applied advanced neural network architectures to predict and optimise the Rate of Penetration (ROP). Dr Jamali described this as a turning point:
“After overcoming the data challenges, we successfully developed AI models for ROP prediction, lithology classification, and drilling problem detection. Testing these models during shallow drilling campaigns demonstrated their practical feasibility, validating the system’s performance in real-world scenarios.”

Field Testing: Bridging Research and Real-World Application

One of the project’s milestones was deploying the OptiDrill prototype for real-world testing. Dr Jamali described the moment as a culmination of years of collaborative effort:
“Seeing the software operational in the field validated years of effort invested in data processing, AI model development, and system integration. Field testing not only proved the system’s feasibility but also provided valuable insights for further optimisation.”
The integration of AI modules for ROP prediction, lithology classification, and problem detection showcased the potential of the OptiDrill system to deliver actionable results for drilling teams.

Collaboration: The Foundation of Success

Collaboration within the consortium played a vital role in overcoming challenges. Fraunhofer IEG worked closely with partners like Geolorn, BGS, ON Power, and Bestec to gather data, understand geological conditions, and align AI developments with real-world drilling needs.
“Having access to the expertise of our partners, particularly on drilling operations and geological properties, was invaluable. Their support ensured we could develop AI models that were both accurate and practical,” Dr. Jamali noted.

A Game-Changer for the Geothermal Industry

When asked about OptiDrill’s lasting impact, Dr. Jamali highlighted the system’s ease of deployment and affordability, “Compared to existing optimisation tools from larger drilling companies, OptiDrill’s simplicity and minimal intervention make it highly attractive—particularly for smaller contractors who may not have the resources to adopt more expensive systems.” While still a prototype, the system’s reliance on historical data and state-of-the-art machine learning techniques gives it enormous potential to revolutionise drilling operations.

Broader Applications and Key Takeaways

Beyond geothermal drilling, Dr. Jamali sees potential for OptiDrill’s methodologies in other sectors, including oil and gas and mining. Industries that rely on real-time data analysis and decision-making under complex conditions stand to benefit significantly.
Reflecting on the project, Dr Jamali shared, “The importance of high-quality data cannot be overstated. Successful deployment of AI tools requires careful attention to regional variability, iterative testing, and strong collaboration with domain experts. This project highlighted that innovation thrives at the intersection of technical expertise and real-world application.”

Interviewee Profile: Dr. Shahin Jamali

Dr Shahin Jamali is the Head of Monitoring and Artificial Intelligence at Fraunhofer IEG and the Coordinator of the OptiDrill Project. With expertise in AI-driven monitoring systems, anomaly detection, and predictive analytics, Dr. Jamali is a leader in applying machine learning to energy and geothermal solutions. His work has been instrumental in developing advanced tools for optimising drilling operations.

The OptiDrill project is funded by the European Union’s Horizon 2020 Research and Innovation Action under grant agreement No. 101006964