For the 2025–2027 period, the LabEx CominLabs will host four new research projects. These projects cover a wide range of fields, from health to digital technologies, and from the environment to space and underwater technologies.
ICON: Improving clinical diagnosis of COgNitive dysfunction in Parkinson’s disease
Parkinson’s disease causes not only motor symptoms but also disabling cognitive impairments, for which no effective drug treatment currently exists. Identifying cognitive profiles remains a clinical challenge. The ICON project aims to identify electrophysiological signatures corresponding to different profiles of cognitive impairment, in order to improve prognosis and develop personalized treatments.To achieve this, the project combines machine learning techniques applied to high-resolution EEG data with neuropsychological assessments, relying on a unique clinical database collected in Rennes.
VideoImpact: Model the environmental cost of video delivery
Video consumption represents a growing share of digital emissions in a sector still often perceived as immaterial. The VideoImpact project aims to identify the most effective levers to reduce the environmental impact of video delivery systems.
It is structured around three main objectives:
1. Model the environmental footprint of the video delivery chain.
2. Build a network of academic and industrial partners in Rennes around this issue.
3. Conduct experiments in collaboration with the social sciences and humanities to test the feasibility and acceptability of the identified levers.
COINS: AI CO-desIgning for the New-Space
The COINS project investigates the use of COTS processors embedding AI applications for the new-space.
Unhardened components used in the new-space are cheaper but more prone to failures, and could be suitable for certain AI applications provided they are designed to tolerate transient faults such as bit flips.
COINS aims to co-design fault tolerance mechanisms, relying on both hardware and software. The goal is to reduce the cost of space systems while maintaining the performance, predictability and reliability required for specific AI applications.
U-CHIC: Underwater-Static Context Header Compression
The U-CHIC project is part of the growing field of the Internet of Underwater Things (IoUT), with key applications in floating offshore wind energy, environmental monitoring, and underwater threat detection.
It focuses on Underwater Acoustic Sensor Networks (UASNs), which connect sensors through acoustic communications tailored to the marine environment. Although slower, these signals remain robust, long-range, and effective even in turbid environnements.
The project aims to optimize these networks despite the constraints of limited data rates and heterogeneous equipment, in order to meet the increasing needs of the sector, particularly for marine renewable energies.
