PRISM : Predictive Risk-Informed asset Strategic Management
1- MIAI chair
PRISM is one of the chairs of the MIAI cluster, which is an international and interdisciplinary Institute in Artificial Intelligence. The PRISM Chair is a four-year project that began in September 2025 and aims to develop AI capabilities in support of critical decision-making for the management of assets at risk, with applications covering both environmental and Industry 4.0 contexts. The PRISM chair brings together four laboratories at the University of Grenoble Alpes, UGA, working in different research disciplines.
-
LJK conducts research on applied mathematics and informatics;
-
GIPSA-lab conducts research on control, signal and images processing, speech, cognition, robotics;
-
3SR conducts research on solid mechanics, materials and structures for civil engineering, environment, transport, defense and health;
-
IGE conducts research on climate, the pollution of our planet and environmental risks.
2- Context and objectives
The aim of this project is to develop decision support methods for the dynamic management of assets at risk in an uncertain context and in the presence of heterogeneous sources of information. The developed methods aim to integrate all the information (sensor data, control data, expert opinion, physical knowledge) into a stochastic model, and then use this surrogate knowledge model to develop predictive decision-making methods. One of the challenges of the project, in terms of AI, lies in the low-data context associated with risk issues, which requires introducing engineering knowledge, explainability and guaranteeing the decisions from the risk point of view. The project draws on a team of academics from a wide range of scientific fields, as well as established collaborations with industry, local authorities and operational technical services. These collaborations aim to help managing ageing critical infrastructures.
From a practical point of view, the chair project aims to provide a guidance both to industrial, critical infrastructure managers, and protective work owners, who are faced to key questions about asset management such as:
-
What is the lifetime of this system and with what level of uncertainty? Or more generally how will the functional and structural efficiency be reduced in context of degradation?
-
When and where should it be maintained or replaced first?
-
When and how should it be monitored in order to decide on replacement/maintenance?
-
How can we anticipate supply requirements based on the failures/replacements that are expected in the next few years?
-
How does the maintenance strategy influence the global risk level?
3- Collaborations
The developments of the chair PRISM are done with support and collaboration of industrial partners and local authorities, in particular:
-
CNR (Compagnie Nationale du Rhône) under a CIFRE industrial PhD framework which is hosted within the project,
-
EDF R&D with industrial contractual collaboration which aims to apply the chair results to real data set,
-
the department of Isère who has provided data on some infrastructure at risks, such as bridges and natural risk protection works like rockfall protective structures, or retaining walls,