To deal with the critical increase in CO2 emissions into the atmosphere, the capture and storage of CO2 from industrial effluents is an essential decarbonisation strategy. For capture, gas/liquid absorption columns have proved their worth. However, because of the large quantities to be treated, leading to large-scale equipment, high energy costs due to solvent regeneration, and their lack of operational flexibility in on-board processes, their deployment is being held back. The CAMPHRE project proposes to address this environmental issue by designing an innovative intensified process for CO2 capture, mineralisation and solvent regeneration. The process involves the use of two centrifugal technologies (HiGee), the Rotating Packing Bed (RPB) and the Spinning Disk Reactor (SDR), combined in a one single rotating unit. The principle is to use centrifugal force to accelerate the transfer phenomena in order to obtain more compact processes. The RPB has proved its worth for the absorption of CO2 by various solvents. SDRs are particularly interesting for precipitation reactions thanks to the extreme mixing generated. The originality of the project involves combining these two operations in a single hybrid rotating process to minimise the energy cost of the operation. The aim is to produce and model a multi-stage prototype in which an RPB will be positioned at the top to capture CO2 using an amine solution, followed below by the SDR to precipitate the absorbed CO2 in the form of CaCO3 using Ca(OH)2. The major challenges of this process will be to ensure the homogeneous distribution of phases in the rotating elements, to control the flows between the two operations, and to understand the coupled phenomena so that they can be modelled and to propose an extrapolation of the process to a larger scale.

Cavitation phenomena can be used to intensify processes involving physical or chemical transformations. Although acoustic cavitation has been studied for a long time for applications in chemistry, water treatment, food processing, etc., and phenomenological knowledge in this field has progressed considerably in recent years, nevertheless its implementation in industrial applications is still limited. Hydrodynamic cavitation appears promising, as it offers the advantage of easier scale-up and has recently attracted interest. The objective of the CAPRI project is to identify and compare the key phenomena involved in acoustic cavitation and hydrodynamic cavitation and to evaluate their impact on the process to intensify typical physical and chemical transformations. The project will employ a global multi-scale approach to identify which type of cavitation method is best suited to achieve the desired final product properties and to recommend the most suitable operating conditions to achieve this. Energy and economic aspects will also be taken into account.

Numerous problems in signal/image processing, statistics, and machine learning rely on the resolution of optimization problems with sparse or low-rank priors. These problems are very challenging to solve due to their combinatorial nature and can be considered as open to a large extent. Within this context, the promise of EROSION is to push the frontiers of sparse and low-rank optimization by combining the strengths of exact relaxations and local optimization. To that end, EROSION will focus on two high-level research objectives: 1) deriving exact relaxations of the targeted problem with the same global minimizers, less local minimizers and wider basin of attraction, and 2) developing initialization strategies that are guaranteed to lie within a basin of attraction of a global solution of the exact relaxation. Finally, these methodological developments will be applied to several signal processing and machine learning problems.

Markov decisions processes (MDP) and their model free counterpart in reinforcement learning have known a large success in the last two decades. However, these successes often rely on quite exceptional hardware possibilities and cannot be applied in many "usual" context, where, for instance, the volume of data available or the amount of computing power is more restricted. To define the next generation of more "democratic" and widely applicable algorithms, such methods still need to deal with very demanding exploration issues. EPLER proposes to overcome this difficulty by exploiting the underlying knowledge and structure present in many MDPs. We will focus in particular on the so-called (rested) multi-armed bandit and restless multi-armed bandit problems, which provide a powerful optimization framework to model scheduling and resource sharing problems. Theory shows that index policies, which are easy to implement, are either optimal or nearly-optimal for bandit problems. Our first challenge will be to characterize performance guarantees for extensions of the restless bandit control problem and to address the case of correlated bandits, which is ubiquitous in resource sharing problems. In our second challenge, we will leverage structures of the optimal policies, i.e. strong optimality of index policies, to significantly improve both the exploration and the exploitation in the model free setting, as well as defining exploration schemes based on particle systems to tackle use cases with sparse rewards.

Facing the need to reduce the use of petroleum resources, limit greenhouse gas emissions and recover waste, lignocellulosic biorefineries appear as a promising solution for the production of platform molecules. In this context, the GLUXY project aims to develop an innovative process scheme combining enzymatic hydrolysis, fermentation and membrane separation steps for the production of 2 platform molecules: glutamic acid and xylitol identified among the economically interesting molecules. The originality lies in the co-fermentation of the two types of sugar resulting from the hydrolysis of biomass, pentoses (C5) and hexoses (C6), without prior separation, by a mixed culture of two microorganisms, a bacteria that produces glutamic acid from C6 and a yeast that produces xylitol from C5. Several types of grain bran will be used as lignocellulosic feedstocks to prepare C6 and C5 hydrolysates. The separation steps will involve membrane techniques, nanofiltration and electrodialysis. The challenge is to limit the number of process steps, and therefore the cost and environmental impact of the produced molecules. The assessment of the environmental impacts of this process scheme will be carried out via a Life Cycle Assessment (LCA). One of the strengths of this project is the simultaneous study of the entire transformation chain that will take into account the complexity and variability of raw materials but also the interactions between each of these steps.