Adaptive control for robotic prefabrication

The European Commission (EU) proposes to increase Europe’s offshore wind capacity from its current level of 25 GW to between 230 and 450 GW by 2050 (approx. 21 250 new turbines of 20 MW capacity and 750 turbines annually) as one of the strategies to meet the EU’s goal of climate neutrality by 2050. To reach this goal, rapid and large-scale deployment of offshore wind farms has become an undeniable need. In addition, the highest energy potential lies far from shore at deep waters where large bottom-fixed (e.g., jacket type) and floating substructures are needed. However, the daunting reality is that Europe lacks sufficient production capacity for such structures, and the cost of production is high. The AdaPfab project aims to develop a sustainable and cost-effective prefabrication process, enabling the mass production of offshore wind substructures. The AdaPfab project addresses three key technical challenges related to prefabrication of substructures, i.e., 1) physics-based cutting and welding planning tool, 2) digital prefabrication supervision and adaptive control methods and 3) automated non-destructive testing (NDT). Since the project has started, a modelling framework has been developed that generates a digital representation of a weld based on process physics. When used for planning, the model will bring the welder’s know-how into adaptive welding process control to produce high-quality welds. The framework is currently under development by applying artificial intelligence (AI) in better welding planning. The first results are currently being implemented into the existing Weld Planner tool that is used in production. A method for adaptive robotic grinding with sensor-based path generation and advanced robot control has been developed, which can replace the current manual and time-consuming grinding work in production. Accurate scans of the weld groove are critical for planning of welding and associated process parameters, and an algorithm has been developed to obtain noise-free scan data. Current welding codes require that one should wait 48 hours after welding to perform NDT, to ensure no cold cracking will occur. In this project, we have found that it is possible to reduce the so-called holding (waiting) time while avoiding cold cracking through extensive literature review and numerical analysis. Testing will be performed to verify our findings. Based on preliminary results, a documented input has been submitted to relevant welding code revision committee for potential revision and modification of the current requirement. This will have significant impact on production, with substantially increased productivity and reduced manufacturing cost when it is adopted. Ultrasound testing (UT) of weld of cylindrical structures tends to be quite challenging due to oblique interfaces and difficulties to access standardized cross section. It is even more challenging when the testing should be performed by a robot. Our work in this project will supply important guidance for robotic NDT through developing advanced robotic UT method, simulation of wave propagation and advanced signal processing method. The project aims to quickly implement the results developed in this project into production considering the urgent need to increase renewable energy production, accelerate the energy transition, and achieve the national and global climate goals.

Rapid and large-scale deployment of offshore wind farms (e.g., 50-100 turbines) has become an undeniable need to increase renewable energy production to accelerate the energy transition and achieve the national and global climate goals. However, a high levelized cost of electricity (LCOE) hinders the accelerated development of offshore wind farms, and a significant cost reduction is urgently needed. In addition, the highest energy potential lies far from shore at deep waters where large bottom fixed (e.g., jacket type) and floating substructures are needed. Today, Europe lacks sufficient production capacity for such substructures, and in addition, fabrication has been identified as one of the areas with the largest potential for cost reduction. The AdaPfab project aims to shift the paradigm of the current production technology: from one-off production to mass production. The AdaPfab project gathers Norwegian stakeholders to revolutionize the current production process and solve key technical challenges related to the prefabrication process, i.e., cutting, welding and automated NDT, which are crucial for building the future World Class Production Line (WCPL), representing a flexible, fully automated, resource-efficient, and digitalized production system. The R&D challenges that will be addressed in this project include developing: 1) predictive welding and cutting planning methods to integrate physics and structural performance into the process planning; 2) technology, methods, and algorithms for automated and adaptive process control and 3) automated NDT methods with the support of modelling and AI-enhanced signal processing and defect detect method. The new prefabrication process developed in AdaPfab will result in 1) increased production capacity and efficiency, 2) improved product quality and decreased cost and 3) enhanced sustainability through reducing production-related CO2 emission and promoting local production.