AIMOS

Artificial Intelligence Modeling Optimization Systems (AIMOS)

Context

The work of the AIMOS (Artificial Intelligence, Modeling, and Optimization Systems) team focuses on artificial intelligence, optimization, and distributed computing, as well as their applications, primarily in the fields of healthcare, personal services, and the environment.

Objectives and scientific themes

Against a backdrop marked by the rapid development of artificial intelligence, the growing volume of data, and the diversification of its uses, the AIMOS team’s work addresses several major scientific challenges. By combining artificial intelligence, optimization, and distributed computing, the team is developing methods tailored to complex, massive, and heterogeneous datasets, particularly those from the fields of healthcare, civil security, bioinformatics, and the environment.

This research aims to improve analytical, predictive, and decision-support capabilities, while taking into account the specific constraints of various fields of application. It places particular emphasis on the robustness and reliability of methods, data confidentiality, and the preservation of human expertise in decision-making processes.

Finally, one of the team’s strengths lies in its resolutely interdisciplinary approach, based on close collaboration with scientific, institutional, medical, and industrial partners. This integration of methodological developments and real-world challenges enables the team to contribute both to scientific advances in artificial intelligence and optimization and to the transfer of these advances to applications with significant societal implications.

Implementation

In collaboration with the fire and rescue services ot the Ain depatment and the DTNUM of the French Ministry of the Interior, the team is developing the concept of the “augmented operator,” which uses generative artificial intelligence to assist emergency call center operators. Transcription and summarization of conversations, extraction of key information, machine translation, and decision support: these tools are designed to reduce cognitive load, speed up call handling, and ensure the reliability of responses, while keeping people at the center of decision-making and guaranteeing the confidentiality and reliability of information.

 This research is part of the efforts of the “AI & Emergency and Risk Sciences” (IA-SUR) research group, coordinated by Christophe Guyeux, a member of the AIMOS team. This group brings together researchers, emergency services, and institutional and industrial partners to focus on AI and digital sciences applied to civil safety, particularly for risk analysis, decision support, and emergency communications.
The team is also involved in the ANR PRCE OPTIMOPS project, which aims to develop a digital twin of fire and rescue services, combined with optimization and artificial intelligence methods to improve resource planning and allocation based on risks and operational constraints.

In the field of health, she continues her work on the analysis of cardiac MRI scans, particularly the detection of the R wave in highly noisy environments, in collaboration with the University of Burgundy and Siemens. Her research also covers the analysis of electronic health records, imaging, cardiac signals, eye analysis, and the automated coding of medical data.

In bioinformatics, the team is developing tools for sequencing and analyzing nearly 400,000 Mycobacterium tuberculosis genomes—representing approximately 100 TB of data—to better understand the mechanisms of resistance, virulence, and the evolution of clinical forms of tuberculosis.

Finally, as part of the TALEDZ project, funded by the PEPR PREZODE – France 2030 program, she is using artificial intelligence to detect avian influenza early in the French West Indies. The goal is to cross-reference diverse data sets to anticipate outbreaks and strengthen public health surveillance.

Environment

Within the ANR OCOD project (2025-2028), the team is developing optimization methods to automate the collection of environmental data useful for monitoring agricultural crops. Sensors distributed throughout the fields measure parameters such as temperature and humidity, and their data is then collected by a drone.Sensors distributed throughout the plots measure, among other things, temperature and humidity, and their data is then collected by a drone. The research aims to optimize flight plans using combinatorial optimization methods, evolutionary algorithms, and geometric heuristics, with a focus on machine learning.