HAL Portal Cnam (French National Conservatory for Arts and Crafts)
Not a member yet
31010 research outputs found
Sort by
Designing scenarios from stakeholders’ acts: Directions and lessons from extreme strategic action scenarios
No full textInternational audienc
Understanding organizational paths : organizing and strategizing tension in a Fab organization
No full textInternational audienceThe objective of this research is both to propose a framework to reveal and analyze the organizational paths of institutions and companies and to implement this framework to produce knowledge about Fab organizations. The organizational path, inspired of the notion of the career path, appears to be of particular interest to reveal and understand organizational development, a goal of Activity-Centered Ergonomics. We conducted a research-intervention in a Fab organization with the aims of co-constructing the organizational trajectory of the company with workers and of “experiencing it”. Here we identify the main moments in the transformation of the organization corresponding to specific organizational forms, as well as the close interrelation between the service offered machines and space, and a strong need for strategizing, organization of work, and organizing. This questions the paradox of some Fab organizations – bound between a fantasy of autonomy and a need of control – and opens up to a methodological and theoretical research agenda. This agenda involves the exploration of the organizational path as it is articulated in actual work as an instrument of discussion spaces and their potential to support the work of organizing and strategizing
Clustering et réduction de la dimension avec données manquantes
No full textInternational audienceIn the context of high-dimensional clustering, we are interested in the Reduced K-means (RKM) subspace clustering method. However, RKM is not suitable for missing data. We propose two methods inspired by commonly used approaches for clustering with missing data: direct approaches through the K-POD method and multiple imputation (MI). RKM provides the partitions and their associated representation subspaces, thus necessitating an additional phase of aggregation of the associated subspaces, which we propose to perform with AFM. Both methods are evaluated on simulated data under the MCAR (Missing Completely At Random) and MAR (Missing At Random) mechanisms, with different rates of missing data. The evaluation is carried out through various scenarios, depending on the balance of the clusters, their sizes and their separation. The results show that both methods succeed in recovering the original clusters and corresponding subspaces in the various scenarios. However, the MI-RKM method requires more time for data imputation and parameter aggregation, which increases computational complexity.Dans le cadre de la classification en grande dimension, nous nous intéressons à la méthode Reduced K-means (RKM) de subspace clustering. Cependant, RKM n'est pas adaptée dans le cas de données manquantes. Nous proposons deux méthodes inspirées d'approches couramment utilisées pour la classification avec données manquantes : les approches directes à travers la méthode K-POD et l'imputation multiple (MI). RKM fournit les partitions ainsi que leur sous espace de représentation associés rendant donc nécessaire une phase supplémentaire d'agrégation des sous espaces associées, que nous proposons d'effectuer avec l'AFM. Les deux méthodes sont évaluées sur des données simulées sous les mécanismes MCAR (Missing Completely At Random) et MAR (Missing At Random), avec différents taux de données manquantes. L'évaluation se fait à travers divers scénarios en fonction de l'équilibre des classes, de leur tailles et leur séparation. Les résultats montrent que les deux méthodes parviennent à retrouver les classes initiales et les sous-espaces correspondants dans les divers scénarios. Cependant, la méthode MI-RKM nécessite plus de temps pour l'imputation des données et l'agrégation des paramètres, ce qui accroît la complexité computationnelle.</div
TRUSTED: The Paired 3D Transabdominal Ultrasound and CT Human Data for Kidney Segmentation and Registration Research
No full textInternational audienceInter-modal image registration (IMIR) and image segmentation with abdominal Ultrasound (US) data have many important clinical applications, including image-guided surgery, automatic organ measurement, and robotic navigation. However, research is severely limited by the lack of public datasets. We propose TRUSTED (the Tridimensional Renal Ultra Sound TomodEnsitometrie Dataset), comprising paired transabdominal 3DUS and CT kidney images from 48 human patients (96 kidneys), including segmentation, and anatomical landmark annotations by two experienced radiographers. Inter-rater segmentation agreement was over 93% (Dice score), and gold-standard segmentations were generated using the STAPLE algorithm. Seven anatomical landmarks were annotated, for IMIR systems development and evaluation. To validate the dataset's utility, 4 competitive Deep-Learning models for kidney segmentation were benchmarked, yielding average DICE scores from 79.63% to 90.09% for CT, and 70.51% to 80.70% for US images. Four IMIR methods were benchmarked, and Coherent Point Drift performed best with an average Target Registration Error of 4.47 mm and Dice score of 84.10%. The TRUSTED dataset may be used freely to develop and validate segmentation and IMIR methods
Factors associated with the transmission of the delta SARS-CoV-2 variant in households: the Israeli COVID-19 Family Study (ICoFS)
No full textInternational audienceUnderstanding how interpersonal interactions and immunological factors shape SARS-CoV-2 transmission in households is crucial for designing control measures. We developed a Bayesian data augmentation transmission model to evaluate the effects of isolation, parental care, and vaccine-induced immunity on Delta variant transmission from the follow-up of 1,093 Israeli households (July–August 2021). Among the 2883 household contacts, 1096 (38%) were infected. Children were 38% (CI: 7-81) more likely to be infected than adults. Isolation measures reduced transmission by 52% (CI: 46-57). Transmission was 39% (CI: 11-76) higher between children and adult females than males. Vaccine effectiveness was 78% (CI: 54-90), 85% (CI: 70-94), and 73% (CI: 49-88) for one, two, and three doses of recent vaccination (< 90 days), respectively but dropped to 18% (CI: (-6)-36) for two doses administered more than 90 days ago. Household member interactions significantly shaped transmission, and isolation measures effectively reduced transmission
Residual stress control in large format polylactic acid additive manufacturing via fast thermomechanical simulation and in-operando imaging techniques
No full textInternational audiencePolymer-based Large Format Additive Manufacturing (LFAM) is an extrusion-based technology utilizing a robotic arm-mounted nozzle to deposit large-diameter polymer beads from heated polymer pellets. However, technical challenges arise due to slower cooling rates and heat accumulation, significant deformation that should be accounted for updating the nozzle path, as well as the development of residual stresses from thermo-chemical shrinkage leading to debonding. To overcome these challenges, the study proposes to combine recent and fast thermal and mechanical approaches. This computationally efficient digital twin of the process is validated experimentally on a thin-wall structure using polylactic acid as a feedstock material. To do so, anisotropic material properties are characterized, and in-operando temperature and displacement field measurements are performed using an infrared thermal camera and backward Digital Image Correlation techniques. Numerical results are in satisfying agreement with experimental data. The validated digital twin is then utilized to characterize the effect of process parameters on the number of layers above the glass transition temperature, the formation of residual stresses and the position offset between the top surface of the structure and the nozzle. This paper presents a fast numerical tool to better design fabrication conditions and improve the quality and fabricability of LFAM-produced parts
Les intérimaires des entrepôts : des salariés subalternes au cœur des recompositions des mondes ouvriers
No full textSES-ENS. Ressources en Sciences économiques et sociale
Facing the challenges of absenteeism and budget constraints in public hospitals. Contribution of management tools in a context of institutional pluralism
No full textInternational audiencePublic hospitals are facing two pernicious challenges: increased budgetary constraints on the one hand, and exacerbated absenteeism and scarcity of paramedical human resources on the other. Healthcare managers are faced with this dual challenge, which they must manage by taking into account two institutional logics in tension: the care logic and the managerial logic. We examine how the use of workforce management tools can enable managers to manage this institutional pluralism in public hospitals, and these contradictory issues. It appears that the instrumental uses of these tools enable managers to embody the managerial logic alone, while the socio-political uses of these tools, implemented in multi-professional discussion forums, promote the hybridization of institutional logics. We detail the conditions that foster the hybridization of institutional logics through the uses of management tools, and the spaces in which it can occur
The dynamics of silica sediment at the ECUSED site in the Baie de Seine based on the COVA methodology: (1) glauconite traced by arsenic
No full textInternational audienceThe English Channel, and specifically the Baie de Seine, is one of the maritime areas hosting the most human activities in Europe. The ECUSED project focuses on the cumulative impact of marine aggregate extraction activities ("Granulats Marins du Havre" concession) and dumping of dredged sediment by the Normand harbours ("Grands Ports Maritimes") for the maintenance of navigation channels (MACHU site), which are concentrated at the outlet of the Seine estuary. To achieve this goal, we used an innovative methodology (ELCOVA) on 114 surface sediments samples according to a regular grid of 2 stations per km². Samples were separated by sieving into 16 particle size fractions and elemental geochemical analyses were carried out on the separated granulometric fractions using X-ray fluorescence spectrometry. The present work focuses on identifying the movements of the various grain sizes naturally present at the site. The mineral considered is glauconite, an alumino-silicate, traced in the elemental analysis by arsenic. The study zone reveals a complex history in terms of granulometry, geography and time. We identify two sources of material, underwater erosion, river input, transports on the seabed and in suspension and a dynamic equilibrium sedimentary state. An essential point of the results obtained is that each granulometric fraction has its own history. Grains belonging to the same sediment sample do not come from the same sources, have not been transported in the same direction, and have not been deposited at the same time under the same energy conditions. It is possible to find grains of the same size and mineralogy from different sources in the same surface sample. This directly challenges the view of sedimentologists on sedimentation