Many imaging techniques, particularly in environmental transmission electron microscopy (ETEM), generate images with degraded signal-to-noise ratio, contrast and spatio-temporal resolution, which hamper quantification and reliable interpretation of data. Moreover, the extraction of structural information from these images relies on manual acquisition and local structural identification which does not allow statistical analysis of the data and necessarily introduces a human bias carried out at the post-processing stage. The general aim of the ARTEMIA project is to develop a ground-breaking deep learning-based framework for in situ microscopy in liquid and gaseous media allowing the automated, high throughput, real-time acquisition and analysis of ETEM image sequences.Our framework will integrate aberration-corrected in situ ETEM imaging using windowed liquid/gas nanoreactors with denoising and resolution enhancement scheme set up using convolutional neural network (CNN). For model training, datasets consisting of simulated liquid- and gas-phase TEM images will be generated by by atomistic simulations including instrumental noise and imperfections of the microscope optics. In the ARTEMIA project, the CNN models will be applied to the study of two crystalline samples with complementary structural characteristics and electron beam sensitivity, model gold nanoparticles (Au NPs) and microporous zeolite, in reactive gas and/or liquid environments. Our scientific aim will be to gain further mechanistic understanding ofthe growth of model Au NPs in liquid phase and their reactivity in oxidizing and reducing gas environments on one hand and the steaming process of beam-sensitive zeolite on the other hand. The consortium comprises three academic partners (MPQ, LEM, IPCMS) and an EPIC partner (IFPEN) with complementary expertise in liquid and gas ETEM, data science and image processing with special focus on deep learning approaches, atomic modelling and TEM image simulation.

The decrease of CO2 & particulates emissions is a main challenge of the automotive sector. European OEMs and automotive manufacturers need new long term technologies, still to be implemented by 2030. Currently, hybrid powertrains are considered as the main trend to achieve clean and efficient vehicles. EAGLE project is to improve energy efficiency of road transport vehicles by developing an ultra-lean Spark Ignition gasoline engine, adapted to future electrified powertrains. This new concept using a conventional engine architecture will demonstrate more than 50% peak brake thermal efficiency while reducing particulate and NOx emissions. It will also reach real driving Euro 6 values with no conformity factor. This innovative approach will consequently support the achievement of long term fleet targets of 50 g/km CO2 by providing affordable hybrid solution. EAGLE will tackle several challenges focusing on: • Reducing engine thermal losses through a smart coating approach to lower volumetric specific heat capacity under 1.5 MJ/m3K • Reaching ultra-lean combustion (lambda > 2) with very low particulate (down to 10 nm) emission by innovative hydrogen boosting • Developing breakthrough ignition system for ultra-lean combustion • Investigating a close loop combustion control for extreme lean limit stabilization • Addressing and investigating NOx emissions reduction technologies based on a tailor made NOx storage catalyst and using H2 as a reducing agent for SCR. A strong engine modeling approach will allow to predict thermal and combustion performances to support development and assess engine performances prior to single and multi-cylinder test bench application. An interdisciplinary consortium made of nine partners from four different countries (France, Germany, Italy, Spain) will share its cutting-edge know-how in new combustion process, sensing, control, engine manufacturing, ignition system, simulation & modeling, advanced coating, as well as after-treatment systems.

The current generation of electric vehicles have made significant progress during the recent years, however they have still not achieved the user acceptance needed to support broader main-stream market uptake. These vehicles are generally still too expensive and limited in range to be used as the first car for a typical family. Long charging times and uncertainties in range prediction are common as further barriers to broader market success. For this reason the CEVOLVER project takes a user-centric approach to create battery-electric vehicles that are usable for comfortable long day trips whilst the installed battery is dimensioned for affordability. Furthermore the vehicles will be designed to take advantage of future improvements in the fast-charging infrastructure that many countries are now planning. CEVOLVER tackles the challenge by making improvements in the vehicle itself to reduce energy consumption as well as maximizing the usage of connectivity for further optimization of both component and system design, as well as control and operating strategies. This will encompass measures that range from the on-board thermal management and vehicle energy management systems, to connectivity that supports range-prediction as a key element for eco-driving and eco-routing driver assistance. Within the project it will be demonstrated that long-trip are achievable even without further increases in battery size that would lead to higher cost. The driver is guided to fast-charging infrastructure along the route that ensures sufficient charging power is available along the route in order to complete the trip with only minimal additional time needed for the overall trip. The efficient transferability of the results to further vehicles is ensured by adopting a methodology that proves the benefit with an early assessment approach before implementation in OEM demonstrator vehicles.

Electrification of vehicles and improved efficiency of internal combustion engines (ICE) are the main levers to reduce greenhouse gas emissions. Recent studies indicate that in 2040 thermal cars sales will still remain an important part of the market and the spark-ignition engine (SIE) is seen as the most interesting ICE technology. However, technological challenges must be tackled before meeting real driving emissions expectation due to the diversification and complexity of hybrid applications. For flow aerodynamics, mixing and combustion down to the individual engine cycle, challenges are for example associated to robustness of concepts on a cycle basis, rapid variations of engine loads observed in hybrid technologies during transients, the occurrence of extreme cycles for a wider range of operating conditions. Numerical, experimental and analyzing tools have made significant progress in recent years for the analysis of spatial and temporal scales of the unsteady in-cylinder flows. Large-Eddy Simulation (LES) is an essential tool for the design of robust concepts. While LES has been validated against well-defined experiments, the prediction of internal turbulent dynamics and combustion during a cycle is affected by epistemic uncertainties. Therefore, progress is still needed to obtain optimal and robust design. The main objective of ALEKCIA is to develop game-changing tools for augmented prediction and analysis of turbulent reactive flows with a focus on real SIE operations to better capture time-resolved events and increase understanding and control of the origins of undesired behaviors. The key hypothesis is that future progress and success is tied to the synergistic, strong combination of experimental and numerical tools at every stage of the project, which will provide advancement in the analysis of physical scales and boundary conditions (BCs). The major scientific challenges addressed by ALEKCIA are to 1/ quantify and reduce uncertainties (UQ) due to model parameters and BCs, 2/ develop new Data Assimilation (DA) approaches for coupling LES with experimental measurements, 3/ develop new decomposition methods to analyse big data generated by LES and high-speed PIV, 4/ combine them with UQ and DA methods for detailed analysis of individual SIE cycles during steady operations and fast transients. We stress that this methodology could also be used more widely for industry and energy applications. To achieve its ambitious objectives, work in ALEKCIA is structured into one management task (T0) and three technical tasks (T1 to T3). We will address non-cyclic phenomena under transient and fired operations and develop novel analysis from the acquired experimental and LES databases of a SIE performed respectively at PRISME (T1) and IFPEN (T3) laboratories. The partners of the project will also collaborate on the development of crank-angle resolved spatio-temporal EMD decomposition (T1 and T3) for engine flows to obtain an unprecedented detailed understanding of the mechanisms involved in the generation of in-cylinder flow, turbulent dynamics and their impact on combustion. The development of UQ tools to quantify and reduce uncertainties in complex LES of SIE flows is also targeted (T3). Finally, the capabilities of DA methods to calibrate realistic BCs on-the-fly is investigated by PPRIME (T2 and T3). This task is particularly relevant when assimilating experimental data (in the form of BC and in-cylinder large-scale flow patterns from EMD) obtained in extreme cycles. EMD obtained from a selected number of measured cycles presenting very slow or fast combustion rates will be coupled with UQ and DA tools for their inclusion in LES (T3). In this scenario, LES will be able to properly follow the assimilated aerodynamic behaviour of these cycles while turbulent dynamic will be modelled. Finally, the application of the developed tools will allow to identify the main key parameters controlling internal aerodynamics.