At IRDL, Haroon Rashid has developed advanced solutions for detecting biofouling on hydropower plants.
Haroon Rashid, chercheur·e au sein de IRDL (Institut National Polytechnique de Bretagne).
Thèse soutenue en 2025 à l'école doctorale École doctorale Sciences Pour l'Ingénieur (Lorient ; 2022-....).
Référencée dans le réseau ABES/STAR, cette thèse répond aux critères de rigueur de l'enseignement supérieur français.
Tidal energy represents a promising avenue to meet the growing needs for renewable energy, particularly in a context where environmental concerns and climate issues are becoming increasingly pressing. This form of energy, which harnesses the movements of tides to produce electricity, could play a key role in the energy transition, especially for coastal countries like Benin, which have untapped potential in this area. However, a major obstacle stands in its way: biofouling.
This phenomenon, which involves the accumulation of marine organisms such as algae, mussels, and other species on turbines, compromises not only the efficiency of hydropower plants but also leads to high maintenance costs and frequent service interruptions. For example, studies show that biofouling can reduce the energy conversion efficiency of turbines by up to 30%, which has a direct impact on electricity production and, consequently, on the profitability of tidal energy projects. In response to this issue, research conducted by Haroon Rashid at the National Polytechnic Institute of Brittany proposes solutions based on machine learning to proactively detect and manage biofouling.
The results of this thesis are significant and open new perspectives. The use of pre-trained models such as VGG and ResNet has greatly improved the accuracy of biofouling detection. These models, which have proven effective in other fields such as image recognition, are suitable for analyzing underwater images to identify areas affected by biofouling. Furthermore, the introduction of a Fast R-CNN model for real-time detection provides a quick response to biofouling issues, which is crucial for maintaining turbine performance. Indeed, the ability to quickly detect the accumulation of organisms allows for intervention before damage becomes too severe, thereby reducing repair and maintenance costs.
The RegStack algorithm, for its part, predicts the impact of biofouling on turbine performance, thus facilitating the implementation of preventive maintenance strategies. By integrating historical and real-time data, this algorithm can anticipate periods of high organism accumulation, allowing operators to plan targeted interventions. These technological advancements are not limited to theoretical improvements. They have been validated on real platforms, demonstrating their robustness and applicability in the tidal energy sector.
By reducing downtime and optimizing operational costs, these solutions could transform biofouling management, thus offering better profitability to hydropower operators. For instance, pilot projects in Europe have already shown that applying these technologies can reduce maintenance costs by 20 to 40%, representing substantial savings for companies in the sector.
It is imperative that stakeholders in the sector become aware of these innovations. The transition to more efficient and sustainable tidal energy will depend on the adoption of these advanced technologies. Policymakers must also consider what policies to implement to encourage the integration of these solutions into renewable energy projects. How can we ensure that investments in these technologies are supported by favorable regulations? What collaborations can be established between academic institutions and businesses to maximize the impact of this research?
Partnership initiatives between universities and companies could foster innovation and technology transfer, ensuring that research results are quickly implemented on the ground. Additionally, government grants could be considered to support tidal energy projects that integrate these new biofouling detection and management technologies.
In summary, Haroon Rashid's thesis paves the way for more effective biofouling management, a crucial issue for the future of tidal energy. The proposed solutions could not only improve the performance of hydropower plants but also contribute to a more sustainable energy transition. By integrating these innovations, Benin and other West African countries could not only reduce their dependence on fossil fuels but also position themselves as leaders in the field of renewable energy while preserving their marine environment.
Accéder à l'étude complète
Sources et accès
Haroon Rashid. Tidal stream turbine biofouling detection, extent estimation, and performance prognosis : a machine learning approach. Electric power. Université de Bretagne occidentale - Brest, 2025. English. ⟨NNT : 2025BRES0012⟩. ⟨tel-05499160⟩