ALEIA, a startup specializing in Artificial Intelligence (AI), and Omexom NDT Engineering & Services, a subsidiary of VINCI Energies, in collaboration with the Laboratoire d'Acoustique de l'Université du Mans (LAUM) UMR CNRS, announces the deployment of the AUTEND project, aimed at accelerating the inspection of nuclear power plants through AI. The project will last 2.5 years (2022-2024) and is supported by the French Ministry of the Economy and Finance and by Bpifrance as part of the Recovery Program.
PARIS, June 30, 2022 (GLOBE NEWSWIRE) -- As the rate and number of inspections on nuclear sites are rapidly increasing, the AUTEND project aims to facilitate and accelerate the work of field analysts, with AI automatic identification of the inspected areas.
The project is currently focusing on Non-Destructive Testing (NDT), which is an inspection process for nuclear infrastructures using eddy current or ultrasonic. In fact, the algorithm developed by AUTEND will identify areas to focus the work of analysts.
The application of AI for these inspections will thus increase the capacity of the analysis and maintain the reliability of the interpretation of the results. Overall, the detection of these zones will reduce the time required for their analyses, and therefore help to respect the restart schedule of the nuclear units. In the long term, AI will contribute significantly to the theoretical reliability of the examinations, specially through a progressive construction of an evolving database.
The AUTEND project is built on the ALEIA platform and sustained by adapted datasets (in quality and quantity) and anonymized test sets. The hosting is secured on a sovereign cloud to guarantee full control of the information processing by the users.
The project is led by three leading partners:
- Omexom NDT Engineering & Services, a subsidiary of VINCI Energies, develops, qualifies, and implements control processes for critical components in nuclear power plants. Omexom is the pilot of the project.
- ALEIA, a French startup specializing in AI, designs and develops the artificial intelligence platform required for the project and industrializes the business application.
- The Acoustics Laboratory of the University of Le Mans (LAUM) contributes to the optimization and implementation of the analysis and data processing within the project.
After the first phase of construction and validation of the datasets in 2022, the AUTEND project enables the integration and testing phase of the algorithm planned for the end of 2022 and the beginning of 2023. The generalization of the experimentation is scheduled for the second half of 2023, before an application to other markets and sectors (aeronautical, oil and gas, or railway) in early 2024.
Jean-François HERR, Omexom NDT E&S company manager, says, "The acquisition rates are increasing thanks to robotization, as well as the complexity of the signals to be analyzed. But the time allocated for the analysis remains the same (from a few days to a few weeks during the nuclear unit shutdown). Faced with this increase in the flow of data to be analyzed, it is essential that our analysts focus on the few inspected areas where their expertise is required. To achieve this goal, the use of AI is an obvious choice."
Antoine COURET, Founder and President of ALEIA, says, "As the availability of the nuclear fleet becomes a major issue of our sovereignty, the selection of ALEIA in this project is a new sign of the legitimacy of our solution to develop a sovereign AI platform and corresponding to critical business needs. The ALEIA platform should enable teams to the successful industrialization of AI and achieve sustainable gains in productivity and time in their missions."
Rachid EL GUERJOUMA and Charfeddine Mechri, project managers for the LAUM, say, "The LAUM develops research activities in non-destructive acoustic testing of materials and complex structures with applications in the fields of transportation, energy, civil engineering ... Alongside the other partners, the LAUM brings to this project its expertise in sensors and instrumentation, signal processing, and DATA. What is more, its particular interest in artificial intelligence (AI) tools that are developed in the laboratory and AI skills should strengthen this collaboration."