L’intégration des systèmes mécatroniques est la source première de leur complexité de conception. Trois principaux types d’intégration sont en cause : l’intégration multi-domaine qui vise à intégrer en synergie la mécanique, l’électronique, l’informatique et l’automatique ; l’intégration multi-niveau prenant en compte les interfaces entre les différents niveaux d’abstraction du système et l’intégration physique dont l’objectif est d’agencer un maximum de composants dans un volume de plus en plus limité. L’objectif de cette thèse est de proposer un moyen d’évaluer cette intégration dans la conception des systèmes mécatroniques, le plus tôt possible dans le cycle de conception. En effet, la phase de préconception génère de nombreuses architectures qui vont être échangées, analysées, discutées et modifiées, avant d’être évaluées en vue de sélectionner les plus pertinentes au regard du cahier des charges à respecter. Dès lors, les concepteurs ont besoin de moyens et d’outils performants pour comparer les différentes architectures candidates, justifier et tracer leurs choix de conception et ainsi réduire la durée de la phase de décisions. S’appuyant sur les modèles générés par le processus d’ingénierie système basée sur les modèles (MBSE), les métriques que nous avons développées visent à évaluer et comparer quantitativement le degré d’intégration des architectures mécatroniques pour faciliter le choix du concept à retenir. Les objectifs adressés par ces métriques sont l’intégration multi-niveau, multi-domaine et physique, ainsi que le degré d’innovation et la mise-en-œuvre d’un projet de conception des systèmes mécatroniques. Ces métriques ont été implémentées en SysML et expérimentées dans un logiciel de CAO (CATIA) pour le choix d’architectures de deux systèmes mécatroniques, à savoir : le moteur-roue et le moteur hybride. Finalement, une méthode de génération de métriques personnalisées a été établie pour répondre au besoin industriel de prendre en compte les spécificités relatives à la culture, l’organisation, la stratégie de chaque entreprise et à l’expérience de leurs concepteurs. / The integration of mechatronic systems is the main source of their design complexity. Three key types of integration are at stake: multi-domain integration which is intended to synergistically integrate mechanics, electronics, information technology and automation; multi-level integration taking into account the interfaces between the different system abstraction levels, and finally physical integration that aims at embedding a large number of components in an increasingly limited volume. The purpose of our approach is to propose a means of assessing this integration degree in mechatronic systems design, as early as possible in the design cycle. The predesign phase indeed generates numerous architectures that will be shared, analyzed, debated and revised before being assessed in order to select the most relevant one according to the specifications to be met. Designers then require means and effective tools to compare alternative architectures, to justify and trace their design choices and thus facilitate decision-making trade-offs. Based on Model-Based Systems Engineering (MBSE), the metrics we developed aim at evaluating and quantitatively comparing the integration degree of mechatronic architectures, in order to provide systems architects an efficient support to opt for the concept to be selected. The objectives addressed by these metrics are multi-level, multi-domain and physical integrations, as well as the innovation degree and finally the design project implementation of mechatronic systems. These metrics have been implemented in SysML and experimented with a CAD tool (CATIA) on different architectures of two mechatronic systems: a wheel hub motor and a hybrid powertrain. Finally, a method to generate customized metrics has been established to meet industrial needs, by taking into account specificities relating to the culture, organization, strategy of each company and their designers’ knowledge and feedbacks.

This thesis is situated in a moment when the theory and practice of inclusive design appears to be significantly implicated in the social and economic response to demographic changes in Western Europe by addressing the need to reconnect older people with technology. In light of claims that cognitive ageing results in an increasing disconnection from novel digital media in old age, inclusive design is apparently trapped in a discourse in which digital media products and interfaces are designed as a response to a deterministic decline in abilities. The thesis proceeds from this context to ask what intellectual moves are required within the discourses of inclusive design so that its community of theorists and practitioners can both comprehend and afford the enaction of cognitive experience in old age? Whilst influential design scholarship actively disregards reductionist cognitive explanations of human and technological relationships, it appears that inclusive design still requires an explanation of temporal changes to human cognition in later life. Whilst there is a burgeoning area of design related research dealing with this issue—an area this thesis defines as ‘cognitively inclusive design’—the underlying assumptions and claims supporting this body of research suggests its theorists and practitioners are struggling to move beyond conceptualising older people as passive consumers suffering a deterioration in key cognitive abilities. The thesis argues that, by revisiting the cognitive sciences for alternative explanations for the basis of human cognition, it is possible to relieve this problem by opening up new spaces for designers to critically reflect upon the manner in which older people interact with digital media. In taking a position that design is required to support human cognitive enactment, the thesis develops a new approach to conceptualising temporal changes in human cognition, defined as ‘senescent cognition’. From this new critical lens, the thesis provides an alternative ‘senescentechnic’ explanation of cognitive disconnections between older people and digital media that eschews reductionism and moves beyond a deterministic process of deterioration. In reassessing what ageing cognition means, new strategies for the future of inclusive design are proposed that emphasise the role of creating space for older people to actively explore, reflect upon and enact their own cognitive couplings with technology.

En conception de produits, l’étape de conception dite architecturale doit aboutir à une solution décrivant les caractéristiques du produit, telles que les dimensions ou le choix de composants. Parmi les nombreuses solutions potentielles, il est nécessaire d’effectuer un choix et de pouvoir le justifier de manière rigoureuse. Ce choix doit en outre intégrer l’éventuelle présence d’une solution de référence, pouvant par exemple être issue de projets antérieurs, ou d’études préliminaires.Dans ce but, nos travaux présentent une méthode d’aide à la décision en vue de sélectionner une solution de conception ; nous la nommons méthode OIA (Observation-Interprétation-Agrégation). Cette méthode emploie deux types de modèle, l’un pour formaliser le comportement objectif du produit (Observation), l’autre pour les préférences subjectives du concepteur (Interprétation). Dans ces préférences, nous incluons le comportement attendu du produit, la satisfaction du concepteur ainsi que les compromis intrinsèquement réalisés lors du processus de conception. Ces compromis sont réalisés au travers de stratégies de conception et de logiques de pondérations (Agrégation). La méthode OIA est employée une première fois pour qualifier les solutions de conception en termes de performance. Cette méthode est ensuite appliquée une seconde fois afin de qualifier la solution par rapport à la solution de référence, en termes de confiance accordée par le concepteur. Nous introduisons la notion d’arc-élasticité en conception comme un indicateur de compromis entre performance et confiance pour une solution de conception.Plusieurs algorithmes stochastiques d’optimisation sont comparés comme méthodes de recherche de solutions candidates pertinentes. Nous les détaillons, les comparons et proposons une méthode de sélection multicritère afin d’identifier l’algorithme le plus pertinent.Notre travail est appliqué au dimensionnement d’une jonction rivetée dans un contexte aéronautique industriel. / In product design, the design phase called "embodiment design" results to define solutions describing main product characteristics, such as dimensions or standard components. Among numerous candidate solutions, it is necessary to select and to rigorously justify the selection. This choice must often take into account preexisting reference solutions, coming from previous projects or preliminary studies.In this context, our works propose a decision support method aiming to select a design solution; it is called OIA method (Observation-Interpretation-Aggregation). This approach uses two types of model, the first one concerns the objective behavior of the product (Observation), the second one concerns the designer's subjective preferences (Interpretation). In these preference several indicators are introduced: expected product behaviors, designer's satisfactions and usual compromises realized through design activity. These compromises are achieved through design strategies and weighting approaches (Aggregation). The OIA method is first used to qualify design solutions though the performance indicator. This method is then used to qualify design solutions related to the reference one, through the confidence indicator. The arc-elasticity is finally introduced as a compromise indicator between performance and confidence.Several stochastic optimization algorithms are compared to compute optimal solutions. Theses algorithms are investigated and compared using three criteria, and a multicriteria selection method is proposed to identify the most suitable one.This work is applied to the dimensioning of a riveted junction in an industrial context.

Afin de déterminer le plus tôt possible dans le processus de conception, les solutions les pluspertinentes, la prise de décisions robuste apparait comme fondamentale pour garantir lesmeilleurs choix. A partir de solutions conceptuelles, l’étape de conception architecturale, ditede pré-dimensionnement, vise à déterminer les principales grandeurs dimensionnantes etpilotantes du système à concevoir, tout en satisfaisant l’ensemble des exigences du cahier descharges. La continuité du processus de conception entre les phases préliminaires et détailléesdépend alors de l’efficacité de la phase de conception architecturale à fournir des solutionsavec un comportement physique validé et une architecture fonctionnelle optimisée. Lesactivités de pré-dimensionnement sont donc fortement tournées vers l’optimisationnumérique. L’utilisation de ces techniques requiert une modélisation précise du problème deconception architecturale. En particulier, l’exploration de vastes espaces de conception, lareprésentation et l’évaluation de solutions candidates, ainsi que la formulation a priori despréférences sont des enjeux majeurs.Les travaux de recherche présentés dans cette thèse concernent le développement deméthodologies et la proposition d’outils pour l’aide à la décision en conception architecturaledes produits et des machines. Plus précisément, l’ensemble de ces travaux vise à fournir auxconcepteurs une démarche adaptée pour structurer et formuler des fonctions objectifs lorsquel’activité de conception est abordée par l’optimisation. Notre approche consiste à relier par lamodélisation de préférences, le comportement physique du système à concevoir avec lescritères et les objectifs de conception, selon des étapes d’observation, d’interprétation etd’agrégation. A partir du concept de désirabilité, cette méthode de modélisation est utiliséepour formuler les objectifs de conceptions et pour quantifier le niveau de satisfaction globalatteint par les solutions candidates. Cette approche est utilisée pour aborder les problèmes deconception robuste où les objectifs de performance et de sensibilité sont mis en balance. Danscette perspective, des mesures de dispersion des performances, ainsi qu’une fonction decompromis spécifique au problème de conception robuste en ingénierie, sont proposés.Enfin, l’application de ces méthodes et outils est illustrée au travers du prédimensionnementd’un évaporateur flash bi-étagé, utilisé pour le traitement des moûts dansl’industrie viticole. L’objectif est alors de trouver des solutions de conception robustes, c'està-dire, des architectures présentant à la fois un niveau de performance globale satisfaisant,incluant la qualité du produit, la transportabilité de la machine ou les coûts, et une faiblesensibilités de la température de sortie du produit, ainsi que de son titre alcoolémique. / In order to converge as soon as possible toward the most preferable design solution, takingrobust decisions appears as a topical issue to ensure the best choices in engineering design. Inparticular, started from a selected concept, embodiment design consists in determining themain dimensioning and monitoring parameters of the system while meeting the designrequirements. The continuity of the design process between the preliminary and detailedphases strongly depends on the efficiency of the embodiment design phase in providingembodied solutions with a validated physical behaviour and an optimized functional structure.Embodiment design problems are thus generally turned toward numerical optimization. Thisrequires an accurate modelling of embodiment design problems, and in particular,investigation of large design spaces, representation and evaluation of candidate solutions anda priori formalization of preferences are topical issues.Research works presented in this thesis deal with the development of methodologies andtools to support decision making during embodiment design of industrial systems andmachines. In particular, it aims to provide designers with a convenient way to structureobjectives functions for optimization in embodiment design. This approach consists in linkingthe physical behaviour of the system to be designed, with the design criteria and objectivesthrough the modelling of designer’s preferences according to observation, interpretation andaggregation steps. Based on the concept of desirability, this modelling procedure is used toformulate design objectives and to quantify the overall level of satisfaction achieved bycandidate solutions. In the scope of robust design, this method is applied first to formulatedesign objectives related to performances, and then, to formulate design objectives related tothe sensitivity of performances. Robust design problems are thus tackled as a trade-offbetween these two design objectives. Measurement methods for performance dispersion andoriginal trade-off function specific to robust design are proposed.Finally, an application of the modelling methodology through the embodiment design of atwo-staged flash evaporator for must concentration in the wine industry is presented.Objective is to find robust design solutions, i.e. configurations with simultaneously adesirable level of performance, including the quality of the vintage, the transportability of thesystem and the costs of ownership, and a low sensitivity of some performances, namely thetemperature of the outlet product and the final alcoholic strength.

Design decisions have a large impact early in the design process. Optimization methods can help engineers improve their early decision making, however, design problems are often ill-posed for optimization at this early stage. This thesis develops engineering methods to use optimization during embodiment design, despite these difficulties. One common difficulty in designing mechanical systems is in handling the effects that design changes in one subsystem have on another. This is made more difficult in early engineering design, when design information is preliminary. Increased efforts have been made to use numerical optimization methods in early engineering design – because of the large impact early decisions have on subsequent development activities. One step toward executing meaningful optimizations in early design is the development of an optimization framework to be used when conditions are expected to change as the design progresses and new information is gained. This thesis presents a design framework that considers such change by subjecting the parametric updating of CAD models to optimization criteria specific to the problem at hand. Under the proposed framework, a part or subassembly is parametrically modeled in CAD; when changes are made to the subsystems that interact with the part or subassembly, it is then updated subject to design objectives and constraints. In this way, the updated part or subassembly satisfies system and subsystem level optimization criteria, reducing the need for the designer to react to design changes manually. It is used to reduce the weight of a Formula SAE suspension rocker by 18%, demonstrating the utility of this framework. Next, we develop methods to help engineers by giving them options and helping them explore during configuration generation. The design of multiple-bend, progressive-die-formed springs typically comprises four steps: (i) functional specification, (ii) configuration generation, (iii) configuration selection, and (iv) detailed shape and size optimization. Configuration generation fundamentally affects the success or failure of the design effort. This presents an important problem: by not generating potentially optimal configurations for further development in detailed design, the designer may unknowingly set the design on track for sub-optimal performance. In response, a method is developed that improves configuration generation. Specifically, an optimization-based spring configuration generator – without which, the generation would typically be based solely on designer creativity, experience, and knowledge. The proposed approach allows the designer to explore numerous optimization-generated spring configurations, which feasibly satisfy the functional specifications. The feasibility study is carried out before a final configuration is chosen for detailed development. Thus streamlining the designer's efforts to develop a design that avoids sub-optimality. We use the feasibleconfiguration generator to identify twenty-two electrical contact spring configurations. All twenty-two of the configurations satisfy the design's functional specifications. Two important concepts that improve decision making in early design were chosen. First, is the concept of a paremetric CAD based framework. Second is the concept of generating iso-performing design solutions. A numerical computer-based application is explained that takes advantage of these two ideas. A genetic algorithm topology optimization framework with the ability to converge to iso-performing solutions was integrated with CATIA V5. This application is demonstrated on a Formula SAE frame where it develops a pareto frontier of designs, expands upon one compromise design by producing iso-performing solutions, and automatically produces designs with the same performance after a parametric suspension change.