Testing data analysis and web application for a nuclear site

Context and challenges

This client is a leader in the nuclear research sector.

The project concerned the modernisation of a liquid effluent treatment plant and more particularly the construction of a new building on this nuclear site. New requirements in terms of security mean that certain units of the plant need to be displaced and other units installed.

The client was seeking to undertake a genuine digital transition within its activities, integrating innovative artificial intelligence and a zero paper approach to optimise its data processing.

It wished to create a digital twin for the construction of the new building. The commissioning test phase represented the first stage of supplying data to the digital twin.

In this context, the client appointed Assystem to implement a data science approach with a digital procedure to monitor and perform the commissioning tests.

Project scope

Analysis of operational data and modelling of commissioning test processes
Development of a web application enabling:
- Collection of documents and forms concerning tests
- Data structuring to perform very precise analysis
- Document tracking and verification throughout the testing phase
- Real-time verification of test status and progress
- Data visualisation on test progress directly in the application (number of test forms per type of test, change in number of forms etc.) and extraction of descriptive statistics for performance evaluation
- Feeding the digital twin with commissioning results
Integration of predictive models to support operators during tests

Client benefits

Optimisation of time required for the testing phase through the implementation of real-time tracking models of form completion
Reduced loss of data through the widespread digitalisation of testing forms
Improved efficiency and ease of work operations for client teams through better coordination and communication between teams enabled via this platform
Snapshot of the initial state of health of systems using a capture of the state of systems in the start-up phase in the digital twin