Search

L’un des objectifs de l’Usine du Future est de fabriquer rapidement des produits sur mesure de haute qualité, avec un niveau de consommation énergétique bas en amenant de la souplesse au sein du système manufacturier global. Cet objectif implique de construire des systèmes manufacturiers plus flexibles et plus résilients. L’intégration de nouvelles technologies issues des domaines de la productique et du numérique constitue une piste intéressante. Cependant, cette évolution doit s’accompagner d’une démarche qui maintient l’Homme au centre du système global, tant sur les aspects décisionnels que sur le contrôle des systèmes « intelligents » implémentés. C’est dans ce contexte que HUMANISM propose de mettre en place une approche de conception centrée sur l’Homme. L’idée majeure est d’équilibrer les implications de l’Homme et de la technologie, en profitant des avancées de l’automatisation, mais également des capacités d’adaptation de l’Homme, ainsi que de ses capacités à tirer profit de son expérience. Les systèmes manufacturiers dits « intelligents » (SMI) déploient des Systèmes Auto-Organisés (SAO) efficients, mais qui sont à même de générer des comportements émergeants pouvant conduire à des situations inattendues et dangereuses ; HUMANISM a pour objectif de développer une méthodologie de conception de systèmes d’aide coopératifs, de façon à soutenir la conscience de la situation de l’Homme, ainsi que ses prises de décisions pour le contrôle de tels SMI.L’approche de conception centrée sur l’Homme proposée dans HUMANISM repose, d’une part, sur l’intégration d’une méthode d’Analyse de l’Activité Cognitive (AAC). Il s’agit de l’une des principales méthodes qui prend en considération les facteurs humains pour la conception de systèmes sociotechniques. Cette méthode pourra assister la conception de système d’aide pour les SMI en fonction des besoins actuels et futurs de l’Homme, besoins qui seront redéfinis de par l’implantation de nouvelles technologies telles que les SAO. D’autre part, l’approche de conception centrée sur l’Homme repose sur les principes de la Coopération Homme-Machine (CHM) qui propose des modèles précis permettant d’identifier et d’implémenter des organisation et des partages de tâches optimaux entre l’Homme et les nouvelles technologies. C’est dans le but d’assurer la généricité de nos développements théoriques que HUMANISM testera ses solutions sur trois systèmes manufacturiers. Ils mettront en œuvre des technologies et des modèles de comportement différents au regard de leur capacité d’auto-adaptation et des niveaux d’automatisation qu’il serait possible d’implémenter. Ces trois SAO seront un cobot, une patrouille de robots mobiles et une patrouille de produits intelligents. Ils se différencient également du point de vue de leur interaction avec l’Homme (proche vs. distant) et dans leur comportement plus ou moins prédictif, ou encore leur capacités de communication. Chaque SAO est dès à présent disponible dans leur version réelle ou simulée dans les laboratoires impliqués dans HUMANISM.

Livrable 5

Livrable 4

Livrable 2

Publié en tant queEP3375689A1;EP3375689B1;ES2769330T3;FR3063702A1;FR3063702B1;PL3375689T3; Le dispositif (10) d'optimisation comporte des moyens (12) de détermination d'une vitesse cible, et des moyens (20) d'indication d'instructions de modification d'une vitesse instantanée du véhicule ferroviaire en vue de l'approcher de la vitesse cible, à l'attention d'un conducteur (22) du véhicule ferroviaire. Les moyens de détermination (12) comportent des moyens (30) d'étude du comportement du conducteur (22), la vitesse cible étant déterminée en fonction de ce comportement du conducteur (22).

International audience; Resilience engineering is an important issue for railway domain. Usually, rail resilience concept is limited to the capacity of the system to confront extreme weather or environmental conditions such as earthquakes or floods. The contribution of the rail staff or passengers is rarely taken into account to assess the capacity of the rail system to recover unprecedented situations, i.e. to make the rail system resilient. This paper aims at developing an original approach to detect possible negative or positive events related to interactions between human and machine in railway domain. It will present three case studies in order toillustrate such an interest. The first study involves and a tram driver and acar driver equipped with a Collision Avoidance System. The second case is anempirical example involving a train speed controller (i.e., the KVB) and atrain driver confronted to several environmental conditions (e.g., presence ofdead leaves on the tracks, rail track with important downhill, etc.). The lastcase relates to an experimental protocol for which the tram drivers are invitedto respect a recommended speed given by an eco-driving tool in order to supportthe energy consumption. Each case is analyzed in order to discover operationalsituations with impact on several criteria such as safety, consumption, ortraffic flow.

The driver assistance device (10) comprises means (12) for determining a target speed (Vc), means (20) for comparing an instantaneous speed (Vi) of the railway vehicle with the target speed. (Vc) or with an interval surrounding the target speed, and means (22) for displaying driving instructions to a driver of the railway vehicle, said driving instructions including instructions for changing the instantaneous speed ( Vi) to approach the target speed (Vc), the display means (22) being configured to: display a speed increase instruction pictogram when the instantaneous speed (Vi) is less than a first predefined value, and displaying a speed reduction instruction pictogram when the instantaneous speed (Vi) is greater than a second preset value.

National audience

International audience

International audience

The aim of Deliverable 5.1 of the ITERATE project is to present results of analyses made of the data collected in the ITERATE WP4 driving simulator experiments. The ITERATE WP4 driving simulator experiments consisted of train and car driving experiments carried out on a portable car/train driving simulator platform and in full scale car and train driving simulators. The analyses conducted and presented in this deliverable consist of hypothesis testing, cluster analysis and development of a procedure for estimation of parameters of relationships in the WP6 UMD simulation model. Findings from the analyses show that country was found to be an important factor in both the car and train driving experiment. Significant effects of the factor workload were also established in many of the events analysed. The conclusion of comparisons between the data collected in the portable and full scale simulators is that the ITERATE portable driving simulator platform is a suitable tool to studying underlying factors controlling driver behaviour at least for the car diving. However, there were differences between the portable and full scale train simulators. These differences have to be considered in later stages of the project. The next step within the ITERATE project is to estimate relevant relationships in the WP6 UMD simulation model based on the findings of the analyses presented in this deliverable. The analyses presented will also be important as inspiration for other studies outside of the present project with the ITERATE driving simulator dataset as an important source of data.

Delivrable D3.1; In Deliverable 3.1 of Workpackage 3, we discuss the methodology developed and applied in the European ITERATEproject (IT for Error Remediation And Trapping Emergencies). This methodology has as its objective to design experiments that will provide data to seed the ITERATE model. In the ITERATE project a high-level theoretical model of vehicle operator behaviour has been developed in Workpackage 1, specifying the factors that play a role in the influence of innovative support systems on vehicle operation in potentially dangerous situations. Themodel isapplicable for different surface transport modes: road vehicles, rail transport and ships. Themodel will be calibrated by experiments investigating how the different factors interact. One hundred and sixty car drivers and 160 train drivers in five countries will drive with a static driving simulator, and 64 drivers (both train and car) with full motion simulators. Finally an executable simulation model will be constructed with the aim to predict the effects of support systems on operator behaviour and risk.

Delivrable D2.1

Delivrable D2.2; This Deliverable provides an outline of the process involved in task 2.2 of the ITERATE project forselecting an appropriate set of hypotheses to be tested inWorkpackages 3 (Experimental design andscenario specification), and 4 (Experimental studies).During a workshop with all partners involved in the project, hypotheses were formulated for thethree main system functionalities outlined in D2.1: those that support speed management (SpeedManagement), systems that support system object detection and avoidance (Collision Avoidance)and those that monitor operator state (Operator State). In formulating these hypotheses, partnerswere encouraged to consider systematically the effect of operator state and different operatorgroups on interaction with these three systems. To allow succinct formulation of hypotheses, fouroperator based parameters which are thought to affect operator behaviour with the system, wereidentified: sensation-seeking, hazard perception, fatigue and (high and low) task demand. Informulating hypotheses, partnerswere encouraged to consider the whole process linking a cause toan effect, with a clear mechanism that would link the two.

Delivrable D1.2; The first work package (WP1) contains a critical review and synthesis of human behaviour models ofdrivers of road vehicles, trains and maritime vessels (ships). Based on this review a reference modelof DriverVehicleEnvironment is developed. A variety of approaches to modeling driver behaviourare possible as options. The literature review covers the more widely cited of these. Generally, thesemight be categorized as either 'Descriptive' models which can only describe the driving task in termsof what the driver has to do or 'Functional' models which are able to explain and predict drivers'performance in demanding situations and drivers' behaviour in typical ones. It seems that theoptimal approach might be a hybrid of several types of models. In recent years, a variety of driversupport and information management systems have been designed and implemented with theobjective of improving safety as well as performance of vehicles. While the crucial issues at atechnical level have been mostly solved, their consequences for driver behaviour remain to be fullyexplained. To reach this goal predictive models of the interaction of the driver with the vehicle andthe environment are necessary. The aim of the European Project AIDE was to integrate all in vehiclesupport and information systems in a harmonized user interface (Saad, 2006). The ITERATE projectwill take this further by developing it into a unified driver model that is also applicable to othertransport domains.The first deliverable in this work package (D1.1) presented a critical review of Driver-Vehicle-Environment (DVE) models and most relevant drivers' parameters and variables to be implementedin such models, in different surface transport modes and in different safety critical situations. Theaim of this deliverable (D1.2), succeeding D1.1 is to describe and detail the Unified Model of Driverbehaviour (UMD), define the environmental parameters to be implemented and their relationshipswith the driver variables. The proposed model will be used to support design and safety assessmentof innovative technologies and make it possible to adapt these technologies to the abilities, needs,driving style and capacity of the individual drivers. The model will also present the environmentalparameters, different road and traffic scenarios with different weather and visibility conditions to besimulated in the test phases. The scenarios of traffic that are independent of the activities carried outby the vehicle and driver will be simulated. The model is simplified in the sense that traffic conditions(density, complexity) are not sensitive to the 'test' driver and vehicle behaviour, but remain fixed in agiven trial. Thus, within the constraints of this pioneering effort, only the behaviour of the test driveris variable, while the environment and vehicle are defined as parameters with fixed values.

Delivrable D8.1b; This document contains the revised plan for using and disseminating knowledge and foreground developed within the ITERATE Project. The Deliverable contains five main Chapters and an Appendix. The first Chapter describes the purpose of the document, its structure, and introduces the other sections. Chapter 2 and 3 define the dissemination strategy of the ITERATE project and provide a classification of dissemination activities. For each type of dissemination action, the corresponding implementation approach is proposed. Then, for each type of dissemination activity, the actions performed to date and those planned are described in details. The dissemination materials already produced by the project and their usage are briefly described. Materials and products already completed, as well as planned, are described, even though a dedicated Deliverable is foreseen in the future that will contain copies of the actual products provided for dissemination purposes. This is particularly true for the Proceedings of the Workshops organised by ITERATE in collaboration with other EU co-funded Projects, which will result in dedicated books and monographs. The Exploitation plan is discussed in the last Chapter of the Deliverable. The two different natures and typology of partners , i.e., academic and industrial/consultancy, are considered. In particular, for each partner, a market and competition analysis is performed and the objectives and guidelines for subsequent exploitation of the results is preliminarily discussed. A further aspect of exploitation of results is discussed in relation to possible developments of EC directives and initiatives in terms of transport safety. Finally, the appendix contains, for completeness, the Dissemination and Exploitation Questionnaire utilised to collect information among partners.

Delivrable D8.1a; This document contains the initial plan for using and disseminating knowledge and foreground developed within the ITERATE Project.The Deliverable contains five main Chapters and an Appendix.The first Chapter describes the purpose of the document, its structure, and introduces the other sections. Chapter 2 and 3 define the dissemination strategy of the ITERATE project and provide a classification of dissemination activities. For each type of dissemination action, the corresponding implementation approach is proposed. Then, for each type of dissemination activity, the actions already performed and those planned are described in some details. The dissemination materials already produced by the project and their usage are briefly described. Materials and products already completed, as well as planned, are described, even though a dedicated Deliverable is foreseen in the future that will contain copies of the actual products provided for dissemination purposes. The Exploitation plan is discussed in the last Chapter of the Deliverable. The two different natures and typology of partners , i.e., academic and industrial/consultancy, are considered. In particular, for each partner, a market and competition analysis is performed and the objectives and guidelines for subsequent exploitation of the results is preliminarily discussed. Finally, the appendix contains, for completeness, the Dissemination and Exploitation Questionnaire utilised to collect information among partners.

Method and distributed simulation tool for the control accomodation of corporate networks; Aujourd'hui, la plupart des entreprises s'organisent en réseaux de partenaires, de fournisseurs et de sous traitants pour réaliser de nouveaux projets. Cependant, les outils classiques de planification n'offrent pas toutes les fonctionnalités nécessaires pour garantir la conduite de telles organisations. Dans ce mémoire, une démarche et un outil sont proposés pour évaluer la faisabilité des plans multi-sites de production ou de transports avant leur exécution. La méthode présentée permet d'adapter la configuration de ces réseaux afin de maintenir de bonnes performances pour les projets évalués. Intégrant le protocole HLA, l'outil gère la synchronisation des simulateurs en garantissant la causalité des événements simulés. La bonne exécution des plans multi-sites associés aux configurations d'ateliers et de logistique est ainsi évaluée. Les performances de l'outil révélées par les résultats de simulation de différents scénarios montrent l'intérêt et l'effectivité de l'approche.

Aujourd'hui, la plupart des entreprises s'organisent en réseaux de partenaires, de fournisseurs et de soustraitants pour réaliser de nouveaux projets. Cependant, les outils classiques de planification n'offrent pas toutes les fonctionnalités nécessaires pour garantir la conduite de telles organisations. Dans ce mémoire, une démarche et un outil sont proposés pour évaluer la faisabilité des plans multi-sites de production ou de transports avant leur exécution. La méthode présentée permet d'adapter la configuration de ces réseaux afin de maintenir de bonnes performances pour les projets évalués. Intégrant le protocole HLA, l'outil gère la synchronisation des simulateurs en garantissant la causalité des événements simulés. La bonne exécution des plans multi-sites associés aux configurations d'ateliers et de logistique est ainsi évaluée. Les performances de l'outil révélées par les résultats de simulation de différents scénarios montrent l'intérêt et l'effectivité de l'approche.