• Rural Digital Europe
• 2013-2022close
• Open Access close
• Other literature type close
• Preprint close
• Part of book or chapter of book close
• FR close
• English close
• Hyper Article en Ligne - Sciences de l'Homme et de la Société close
• arXiv.org e-Print Archive close
• INRIA a CCSD electronic archive server close
• HAL-Pasteur close
• HAL Descartes close

The works focus on the identification of industrial robots that belongs to the field of the identification of continuous-time inverse dynamic models in closed loop. First, a generic instrumental approach relevant for the identification of rigid industrial robots is proposed. The set of instruments is the inverse dynamic model constructed from simulated data calculated from the simulation of the direct dynamic model. This algorithm termed the IDIM-IV method validates the inverse and direct dynamic models simultaneously, improves the noise immunity of estimates with respect to corrupted data in the observation matrix and has a rapid convergence. This new approach is experimentally validated and compared with other standard methods. Then, a statistical test able to assess the validity of the set of instruments as well as the consistency of the least-squares estimates is presented. This test is based on the use of the Two-Stage-Least-Squares method and the regressed Durbin-Wu-Hausman test that are commonly used in econometrics. Finally, the perspectives that the IDIM-IV method can offer to the communities of robotics and automatic control are enlightened Ce manuscrit résume les travaux menés sur l'identification des robots industriels conduits à l'ONERA. Dans le premier chapitre, les méthodes usuelles de modélisation et d'identification appliquées aux robots industriels rigides sont présentées. Le deuxième chapitre introduit une nouvelle méthode d'identification basée sur l'utilisation des variables instrumentales alors que le troisième chapitre présente un test statistique capable de valider la construction des instruments. Enfin, dans un quatrième chapitre, les perspectives offerts par ces nouvelles méthodes sont introduites.

The objective of this dissertation is to shed light on some fundamental impediments in learning control laws in continuous state spaces. In particular, if one wants to build artificial devices capable to learn motor tasks the same way they learn to classify signals and images, one needs to establish control rules that do not necessitate comparisons between quantities of the surrounding space. We propose, in that context, to take inspiration from the " end effector control " principle, as suggested by neuroscience studies, as opposed to the " displacement control " principle used in the classical control theory.

In Senegal, as in many developing countries, the agricultural sector plays a key role in the economy. As well as supplying food, agriculture is the most important source of employment, especially for women. Through the Plan for an Emerging Senegal (PES) the Senegalese government is implementing an ambitious financing plan to improve the productivity of the agricultural sector and to improve women's employment opportunities. Our study assesses the impact of two PES measures (investing in the agricultural sector and an increase in production subsidies for the agricultural sectors) on economic growth, women's employment, poverty and inequality using a dynamic computable general equilibrium model linked to a micro-simulation model. While on the one hand, we find that investing in the agricultural sectors has positive impacts on growth, improves women's work opportunities, decreases poverty and inequality both in the short and long term, however increasing production subsidies in the agricultural sectors has positive effects only in the short term and does not reduce inequality or poverty in the long term.

We tackle here the problem of multimodal image non-rigid registration, which is of prime importance in remote sensing and medical imaging. The difficulties encountered by classical registration approaches include feature design and slow optimization by gradient descent. By analyzing these methods, we note the significance of the notion of scale. We design easy-to-train, fully-convolutional neural networks able to learn scale-specific features. Once chained appropriately, they perform global registration in linear time, getting rid of gradient descent schemes by predicting directly the deformation.We show their performance in terms of quality and speed through various tasks of remote sensing multimodal image alignment. In particular, we are able to register correctly cadastral maps of buildings as well as road polylines onto RGB images, and outperform current keypoint matching methods.

The African continent is vast with an unequally distributed population. It has experienced high population growth over the last 50 years, which has corrected its former demographic deficit. If current dynamics continue, the results will be population densification, with serious implications for the relationship between people and natural resources, and changes to territorial configurations.

International audience; The diffusion of next-generation sequencing technologies has revolutionized research and diagnosis in the field of rare Mendelian disorders, notably via whole-exome sequencing (WES). However, one of the main issues hampering achievement of a diagnosis via WES analyses is the extended list of variants of unknown significance (VUS), mostly composed of missense variants. Hence, improved solutions are needed to address the challenges of identifying potentially deleterious variants and ranking them in a prioritized short list. We present MISTIC (MISsense deleTeriousness predICtor), a new prediction tool based on an original combination of two complementary machine learning algorithms using a soft voting system that integrates 113 missense features, ranging from multi-ethnic minor allele frequencies and evolutionary conservation, to physiochemical and biochemical properties of amino acids. Our approach also uses training sets with a wide spectrum of variant profiles, including both high-confidence positive (deleterious) and negative (benign) variants. Compared to recent state-of-the-art prediction tools in various benchmark tests and independent evaluation scenarios, MISTIC exhibits the best and most consistent performance, notably with the highest AUC value (> 0.95). Importantly, MISTIC maintains its high performance in the specific case of discriminating deleterious variants from benign variants that are rare or population-specific. In a clinical context, MISTIC drastically reduces the list of VUS (<30%) and significantly improves the ranking of "causative" deleterious variants. Pre-computed MISTIC scores for all possible human missense variants are available at http://lbgi.fr/mistic.

International audience; Digitalization is a topic that is ubiquitous in everyday life. The technological revolution progresses with high pace. Technology/web/mobile companies all over the world are flourishing but primary schools (in Germany) do not follow the economy's development and has not made computer science part of its curriculum yet. This is either due to technical equipment limitations or to the inability or insecurity of teachers to include technology and the underlying principles in their classroom. Furthermore, it is not officially integrated into the core curriculum which prohibits the implementation of a universal standard for computer science competences that have to be taught in primary schools. The paper will present the underlying project and the development of specific learning environments that are designed to nurture cooperation and interaction between participants to have a closer look at the in-between, where learning occurs. This will help to solidify the claim that computer science in primary school has a right to exist.

International audience; Oyster production has historically taken place in intertidal zones, and shellfish farms already occupy large extents of the French intertidal space. The expansion of French shellfish aquaculture within intertidal areas is therefore spatially limited, and moving production to the subtidal offshore environment is considered to be a possible solution to this problem. Finding new sites along the French Atlantic coast was studied here from the perspective of small oyster companies run by young farmers, who are interested in offshore bivalve aquaculture expansion compatible with their investment capacity. In assessing the feasibility of such offshore production, we considered three main issues: (1) bivalve growth potential and (2) technical feasibility and conflicting uses, both within a spatial framework, as well as (3) the steps and barriers of the administrative licensing process. Oyster spat in an experimental offshore cage showed significantly faster growth, in terms of both weight and length, compared to those in an intertidal cage, mainly due to lower turbidity and full-time feeding capacity (i.e., constant immersion in the water). A combination of Earth Observation data and bivalve ecophysiological modelling was then used to obtain spatial distribution maps of growth potential, which confirmed that offshore sites have better potential for oyster growth than the traditionally oyster-farmed intertidal sites overall, but that this is highly spatially variable. Small-scale producers indicated two technical factors constraining where farms could be located: bathymetry must be between 5 and 20 m and the distance from a harbor no more than five nautical miles. These were included along with maps of various environmental and socioeconomic constraints in a Spatial Multi-Criteria Evaluation (SMCE). Touristic traffic and bottom trawling by fisherman were found to be the two other most restrictive variables. The GIS-based SMCE developed in this study showed that there is almost 400 km 2 of highly-to very highly-suitable area within which to develop offshore aquaculture using simple, low-cost bottom-cage techniques, and can be used to assist the shellfish industry in the Marine Spatial Planning decision-making process, still in progress in this coastal area. However, the complexity of the administrative processes necessary to obtain an offshore license is perceived as a stronger barrier by farmers owning small companies than site selection, technical feasibility, and required investments, and will be crucial to address in order to realistically proceed to offshore cultivation. The process demonstrated here, and the results are relevant to other coastal and offshore locations throughout the world and can be adapted for other species.

International audience; When talking about media in primary school, often, one thinks of the use of computers during everyday classes every once in a while. When realizing that we live in a more and more digitalized world, this seems rather odd. The sole use of media in school is not sufficient to prepare children for a world that expects so much more. The fundamentals of computer science are strongly related to mathematics content and can therefore be linked to competences that children already learn today. The paper will present the evaluation of learning environments, developed with computer science content, to have a closer look at interactional processes that occur when working on a topic that is on the one hand strongly connected to e.g. mathematics but on the other hand fundamentally new for elementary school children and especially relevant in today's society. We will focus on co-cooperative processes and consensus and dissent.