Lies Benmiloud-Béchet
Director of Digital Excellence Center
With a solid technical background and almost 20 years’ experience in SCADA and Instrumentation & Control systems, Lies has been involved in numerous complex construction projects in a variety of sectors, including transport, nuclear and digital. As a certified Senior Project Manager, he manages high-stakes technical projects with multicultural teams in France and abroad. In 2022, he will become Director of Operations at Assystem’s Digital Factory before taking over the management of Assystem’s Digital Excellence Center in September 2023, where he will oversee digital innovation projects to support the transformation of the Group’s engineering business.
To deal with the challenges of the energy transition, energy production and transmission infrastructure projects are multiplying to meet the growing need for low-carbon electricity. At the same time, these projects are faced with increasing complexity and a growing number of requirements aimed at optimising financial investment and the overall cost of the energy produced, while guaranteeing the highest level of safety.
The integration of solutions based on Artificial Intelligence technologies is a formidable lever for facing these different challenges. They enable informed decisions to be made based on controlled data throughout the infrastructure's lifecycle.
Lies Benmiloud-Bechet, Director of Digital Excellence Center, explains how the development of technologies based on Artificial Intelligence can speed up the execution of complex projects and meet the challenges of the global energy transition.
Artificial intelligence (AI) at the service of project engineering
While AI is growing in popularity, with generative AI at the heart of the news (ChatGPT, Midjourney, etc.), these technologies are still hardly considered by companies in the construction and industrial sectors. In 2022, only 12% of them will be capitalising on AI to boost their results, despite AI adoption more than doubling between 2017 and 2022.
Convinced that engineering has a central role to play in contributing to the global energy transition, and that the use of Artificial Intelligence can be a lever for acceleration, Assystem experts have been capitalising for many years on the development of technologies exploiting Artificial Intelligence*.
* About Artificial Intelligence
Artificial Intelligence offers tremendous opportunities to support the development of large, complex infrastructure programs. Thanks to cutting-edge techniques such as deep learning, computer vision and natural language processing, AI is capable of accelerating engineering projects in all phases of the infrastructure lifecycle. Some AI models use algorithms trained on existing data to automate engineering phases such as quality control, or to simulate different scenarios for predicting system behaviour to facilitate decision-making based on complex data (predictive maintenance and optimisation under constraints). The scope for AI applied to engineering is vast, enabling manufacturers to optimise the management of even the most complex projects.
Towards augmented engineering
Thanks to their ability to process, analyse, model and understand huge volumes of data, AI algorithms represent a real competitive lever for the specific field of engineering. The processing power of AI accelerates business processes, boosting operational performance.
The Artificial Intelligence market in the engineering sector is progressing at full speed, thanks to a worldwide development, innovation and imitation of this technology and the growing number of use cases. For example, AI can be used to automate tasks, anticipate clashes, optimise planning and maintenance, and maintain rigorous quality control regarding current standards and regulations. Automation and rapid access to relevant data optimise the productivity of engineers and technicians. In this sense, AI applied to engineering to the challenges of the massive need for human resources in the future to ensure the energy transition, and even more so for the nuclear sector (major programs to come).
However, while the integration of AI into engineering projects is becoming a matter of course, like any disruptive innovation it must be used by industrial players by assessing and managing the potential risks of its use (RGPD, Cybersecurity). The European community will be regulating the use of AI to control potential misuse, through the European Artificial Intelligence Act currently being drawn up.
Finally, it's important to emphasise that, in the case of critical infrastructures with high safety and certification stakes, human intelligence remains indispensable for interpreting and validating results and making strategic decisions.
«AI thus complements human intelligence, creating a synergy for the success of complex projects.»
Use case: AI at every stage of the project lifecycle
Artificial Intelligence can be used for all or part of the lifecycle of complex projects. With their dual expertise in engineering and digital technology throughout the world, Assystem's teams have already been able to contribute to several areas of use:
- Capitalising on expertise by automating repetitive, low added-value tasks: a key stage in the life cycle of industrial projects, particularly in the case of installations, is to take proper account of the applicable standards and regulations via project and system requirements. In other words, it's about efficiently translating existing generic documentation into a project-specific requirements repository. Assystem has developed the DeepREXT solution, which uses AI algorithms to automate the extraction of requirements. DeepREXT transforms document formats into data using Deep Learning (DL) technologies that analyse the structure of document pages and interpret the content of a paragraph.
- Guaranteeing the optimisation and control of schedules while remaining agile in the face of the hazards inherent in any complex project: the complex nature of projects requires advanced planning based on a multitude of variables, interdependencies, and constraints. AI plays a crucial role in this process. It can analyse vast quantities of data, predict results based on various scenarios and optimise planning by considering numerous factors and constraints. It enables more informed and rapid decision-making, reduces risk and improves overall project efficiency. Among the digital tools developed by Assystem, Optimizio* meets this need by relying on tools based on the latest AI technologies. Optimizio* is capable of optimising highly complex schedules (up to 20,000 scheduling tasks and 12 to 14 constraints) and with a high level of recurrence (more than 10 schedules to be made per year), thus helping to reduce the costs and risks associated with scheduling errors.
- Aggregating and structuring all forms of textual data, in paper or digital form: throughout the lifecycle of industrial and nuclear facilities, operators produce a large quantity of documents in a wide variety of formats and media. This information covers a wide range of areas, and access to it is essential for maintenance, audits, upgrades, extensions, decommissioning operations, etc. To process, structure and exploit this information, the electronic document management (EDM) tools currently in use are inadequate. The DeepFinder solution achieves this by combining traditional techniques for capturing and classifying textual documents with the use of ontologies and new AI technologies applied to natural language processing (NLP).
Thanks to the development of its digital activities, Assystem can combine human and artificial intelligence to support its customers throughout the lifecycle of complex projects.
The use cases and capabilities of AI in engineering are multiplying exponentially, Assystem must continue to develop solutions related to these technologies to continue to accelerate engineering processes, automate repetitive tasks. However, we are not neglecting the existing challenges related to the regulatory environment around AI and data protection.
* Optimizio is an Assystem solution
Key figures
-
46%of organisations plan to implement AI in the next three years. (2022, Deloitte)
-
64%of companies expect AI to increase their productivity (2023, Forbes)
-
37%annual growth rate expected for AI between 2023 and 2030 (2023, Forbes)