In real-time safety critical systems, it is of paramount importance to guarantee that computation is performed within certain time bounds, otherwise a critical failure may happen. Avionics and aerospace systems, electronic automotive systems, train control systems, etc, are all examples of real-time safety-critical systems. To guarantee correctness, the designer needs first to compute bounds on the execution time of every block of code, and then to guarantee that, in the worst-case, every block is scheduled by the operating system to complete before its deadline. Today, it is difficult to build efficient and predictable real-time systems on modern processors, because the execution time of a piece of code exhibits a large variability. The worst-case can be hundreds of times larger than the best-case, due to dynamically varying paremeters such as the state of cache memories for instance. Therefore, the designer needs to greatly over-provision the computational capacity of the processors, leading to a higher cost of the system. The continued demand for additional functionalities makes the situation unsustainable in the long term. While some methods have been proposed to deal with such large variations, they are not immediately applicable because they focus on scheduling without considering the functional aspects of the application. The overall objective of this project is to contribute to the design and development of the next generation of safety critical embedded real-time systems. In particular, we aim at solving the problem of the large variability of execution times by using sound and provably correct programming models that combine functional and timing aspects. The main idea can be summarised as follows. First, we will use parametric Worst-Case Execution Time analysis techniques for computing off-line a WCET formula. The formula is parametrised with respect to input values of the code block and to state of the processor cache. Then, we plan to use the formula at run-time to dynamically estimate a tighter upper bound to the execution time. The execution time estimation will be used at run time to dynamically select the application behaviour so as to avoid deadline misses. The designer will specify the behaviour of the system by using a synchronous language to formally guarantee at the same time functional and timing correctness. Finally, we propose to use a design methodology to help the designer configure the system in the best way.

More and more applications in electromagnetic compatibility require to take into account very small elements in the geometry for a correct evaluation of electromagnetic interactions. We are particularly interested in this project in a problem dimensioning in aeronautics, which is the risk of electrical breakdown resulting from a high amplitude current injection (example of lightning), in the areas of overlap by riveting panels constituting the structure. More concretely, one wonders what would be on an aircraft the potential points of electrical breakdown. These breakdowns are indeed sources of degradation of the structure and generators of disturbances or even initiation of fires or explosions in sensitive areas. Today, a simulation of this problem is a real challenge because there is no effective method taking into account both the 3D geometric extent and the local details related to these risk areas. Indeed, to simulate the problem it is necessary to know globally the distribution of currents and surface fields over the entire 3D structure, but also around the inter-panel space in the overlap areas. Despite the fact that we can locally refine a mesh in the current GD schemes, we note, for our problem, that the number of zones to be refined and the size of the detail to be taken into account in these zones require massive multi-scale meshes) that make current solutions too expensive or even inadequate. It is therefore important to consider this multi-scale aspect for global geometry and to propose efficient solutions in cost calculation, in memory load and in precision to correctly overcome the problem. In this project, to have an operating solution, it is necessary to make improvements to the current GD approach, but also to have a good mesh generation strategy for the simulation in terms of calculation costs. This is why we are interested in the following 2 scientific challenges: - The study of methods or diagrams that allow to improve the GD approach in terms of calculation and memory cost to be able to deal with applications where we have to consider several mesh zones with an important scale factor. We then obtain non-compliant hp meshes in which the calculation of the fluxes at the interfaces of the different mesh areas is very expensive and the time step for the stability of the numerical scheme becomes too small. We will therefore look for inexpensive approaches for the calculation of flows at the interfaces of refinement zones and temporal schemes based on stable local time step methods, able to take into account large variations in time steps (greater than 10 or 100) or schemes without CF or locally implicit time; - In order to perform the simulation, it will then be necessary to have a strategy to define an optimal mesh in terms of calculation and memory cost with our improved GD solution. In particular, it will be necessary to be able to define cartesian and unstructured meshing areas according to the geometry under study, such as the number of Cartesian cells and the size of the cells being maximized, while maintaining the stability of the GD scheme. In addition, for some unstructured areas it will be necessary to define local refinements able to handle small details, while ensuring the stability and consistency of the improved GD approach. The solutions proposed in the two previous points, will be validated and quantified on an example of small size by comparison with current solutions.

This project deals with the design and implementation of a laser system and a dedicated focusing module for the ignition of a two-phase flow in the aerobic aeronautic chamber. The ignition, and more particularly the re-ignition of aeronautic engines in flight conditions, is a crucial point for the aeronautic and it directly impacts the volume of the combustion chamber. Aeronautic companies developing engines are actively seeking for systems making possible to ignite at high frequency (~ 100 Hz) combustion chambers requiring active re-ignition, including chambers with rotating detonation or constant combustion volume. Making available a reliable and an efficient ignition device working at high frequency is therefore of great importance for aeronautic engines manufacturers. Hence, the main idea of this project is to develop a system consisting in a laser and focusing module making possible the ignition of the two-phase kerosene / air mixture of an aeronautical injector under altitude flying conditions (-40 ° C, 0.5 bar). The development of this system is made possible thanks to the know-how and innovative results acquired during the ASTRID ECLAIR project. The optical and mechanics parts of the laser system will be designed and build by Fibercryst. A compact electronic device making possible to pump and synchronize this laser system will be designed and build by the LOMA in strong interaction with Fibercryst which will provide the key electronic parameters. This electronic device derived from a system previously designed by the LOMA will be operated under 24 V. The laser system will deliver at 100 Hz repetition rate, laser pulses of 1.5 nanoseconds with quasi-Gaussian profile (M2 <1.5) centered at 1064 nm which energy will be greater than or equal to 40 mJ. These laser pulses will be focused in the combustion chamber by a dedicated focusing module that will be designed and built by LOMA. At the point of focus, these pulses will create plasma that will generate a combustion kernel and ignite the combustion. In parallel, we will characterize the kerosene spray of SAFRAN's Ardiden trisector under different operating conditions. A better knowledge of the parameters of the kerosene spray as a function of the temperature and the pressure will improve notably the predictive character of the computation codes of the dynamics of the fluids and will make it possible to optimize the conditions of ignition of the Ardiden motors developed by SAFRAN HE. Using our laser device and focusing module, we will carry out ignition tests on the ARDIDEN trisector installed on the MERCATO bench of ONERA. This bench is the French reference for the study of the ignition of aeronautic combustion chambers under real flight conditions. The performances of our laser igniter will be determined using optical diagnostics installed on MERCATTO and compared to that of conventional spark plugs that will serve as a reference. Finally, we will also test the ability of our system to ignite a complete annular chamber installed on the bench 5 of Safran HE at Bordes. These latter experiments will make it possible to test the ability of our system to ignite an aeronautic engine under different flight operating conditions.

The SAFASNAV project is proposed by a consortium comprising DCNS, Telecom ParisTech, ONERA and SART. The goal is the development of structural materials and multilayer thin coatings that reduce the structures and superstructures electromagnetic reflectivity of civilian and / or military ships. This project is an extension of the results obtained during the previous SAFAS project. It consists of replacing the dielectric layers and metallic grids of the SAFAS absorbent structure (metamaterial) by specific materials appropriate to the naval environment. Thus, this work has two objectives: firstly to incorporate metamaterials in naval structural composite structures and secondly to develop coating metamaterials suitable for these naval applications. In addition to significant advances for military applications, the results of the SAFAS project clearly showed potential applications in the civil sector of the antenna and also the absorbent material. The level of TRL reached at the end of the project was that of a validation in a laboratory environment, that is to say, a TRL = 4. Further developments proposed here aim a TRL = 5/6. They are intimately related to the application and justify the entry of DCNS in the consortium for the development of electromagnetic absorbents for naval application in the context of this Astrid-maturation project. Combining the skills and results of SAFAS with the needs in terms of naval defence, but also the expertise in the field of composite materials and marine environment, offers rich prospects in valorisation and increase in maturity of the SAFAS results, as well as some potential applications (in the military, but to a large extent also in the civil sector). This is the challenge of SAFASNAV that move from laboratory mock-up validating a concept to the walls of functionalized naval structures, designed and implemented in an established industrial process. The innovative field of this project is based on the very special properties of self-complementary surfaces that have a theoretically unlimited operating bandwidth and that allow, when associated with one or more other high impedance surfaces, to produce thin absorbent materials without any lossy dielectric or magnetic material. A first purpose of the study is to integrate these multilayer absorbent materials directly in multilayer structural materials used to construct the superstructures of ships. The second purpose is to make coating materials used to reduce the reflectivity of single or multiple metal superstructures. Applications concern the stealth ship (SER) but also all civil and/or military intra-system and inter-system EMC applications (antennas decoupling, radiation patterns correction, effectiveness of shielding ...). The technologies developed should also be able to find their applications in the field of renewable energy (On/Off shore wind power).

The PARHéRo project explicitly aims to increment synergies between scientific and industrial research to anticipate and control the evolution of heterogeneous robotic platforms in complex, unknown and/or hostile environments. The successful completion of robotic missions is ensured by the high degree of autonomy of the platforms, a central element for their robustness, which is achieved through learning, planning and the supervision of the execution of intelligent behaviour. The project aims firstly at providing autonomous multi-robot platforms with a mission specification language able to express both the objectives and the requirements on the state of the robots and the characteristics of the planning model. In order to test the specification language, the project will aim to generate case studies that are coherent with the applications envisaged for future defence and security systems. The mission specification language is also the means by which the results of the learning, achieved by each of its elements, can be shared in the fleet of heterogeneous robots. Autonomous decision making, or even interactive planning with a human overseeing the mission, strives for platform resilience to unexpected, dangerous, or unpredictable events in the environment. In this context, the use of domain knowledge -- whether prior or on-the-fly during the mission execution phase -- can provide a quicker and better solution to the problems faced. This hybridisation between Automatic Planning in Artificial Intelligence and Machine Learning ensures the robustness, adaptability, and resilience of the fleet of heterogeneous robots, all participating in the same strategic objective. Automated planning and machine learning (in particular Reinforcement Learning) are characterized by complementary views on decision making: the former relies on previous knowledge used to create a model and computation of a solution from this model, while the latter on interaction with the world, and repeated experience. Reinforcement learning, can start without any previous knowledge, and allows robots to robustly adapt to the environment, but often necessitates an infeasible amount of experience. Planning allows robots to carry out different tasks in the same domain, without the need to acquire additional knowledge about the domain or about each one of them, but relies strongly on the accuracy of the model. Furthermore, the search space of a planner with partial knowledge about the environment can grow exponentially in the number of possible states, making the planning process practically unfeasible. However, even a small injection of knowledge from prior learning about the model can greatly improve the performance of the solution search. This a priori knowledge, which can come from learning phases of intelligent behaviour, allows the refinement of meta-heuristics, macro-actions, or even hierarchical task decompositions. Learning techniques can also be used to improve the decisions made by a group of robots, such as mission optimisation against opposing criteria. Whether they are high-level strategies or purely reactive components, the coordination of a fleet of autonomous mobile robots requires the transmission of information learned by each robot based on local information, provided that the robustness conditions of the communication network are maintained. Otherwise, the estimation by each robot of the global situation is necessary to guarantee the autonomy of the robot fleet, and the robustness of the mission.