Search

International audience; Industry 4.0 is leading to rethink how operational decisions are made within companies. In particular, it raises the question of the evolution of employee involvement and autonomy in operational decision-making in a Lean 4.0 context. Dealing with such issues within companies presents high stakes but also involves many risks and difficulties. Therefore, it is necessary to test these new Industry 4.0 autonomy models within our Evolutive Learning Factories by developing a suitable experimental protocol. This article proposes a typology of scenarios and case studies that will serve as a basis for future experiments to study these issues in a standardized work context. This first study framework confirmed that the decisions induced by all the problems and opportunities encountered at the operational level are numerous and varied. This research work is a first step and opens up much broader research perspectives on the contribution of Industry 4.0 technologies in implementing new models of autonomy at work.; L'industrie 4.0 amène à repenser la manière dont les décisions opérationnelles sont prises au sein des entreprises. Se pose notamment la question de l'évolution de l'implication et de l'autonomie des employés dans les prises de décision opération-nelles dans un contexte Lean 4.0. Traiter de telles problématiques au sein des entre-prises présente de forts enjeux mais comporte aussi de nombreux risques et diffi-cultés. Il convient donc de pouvoir tester ces nouveaux modèles d'autonomie 4.0 au sein d'Evolutive Learning Factories en développant un modèle expérimental adapté. Cet article propose une typologie de scénario d'études qui servira de base aux futures expérimentations permettant d'étudier ces problématiques dans un contexte de travail standardisé. Ce premier cadre d'étude a permis de confirmer que les déci-sions induites par l'ensemble des problèmes et opportunités rencontrés au niveau opérationnel sont nombreuses et variées. Ce travail de recherche constitue donc une première étape et ouvre des perspectives de recherche beaucoup plus larges sur l'ap-port des technologies de l'Industrie 4.0 dans la mise en oeuvre de nouveaux modèles d'autonomie au travail.

International audience; Manufacturing industry needs access to the data in order to realise its activities but also to generate new value-added knowledge. Nevertheless, it is confronted with a large and growing volume of heterogeneous data which limits its ability to exploit them optimally. Moreover, the data are distributed within different heterogeneous information systems, which limits the relationship exploration under the information retrieval process. Usually, the challenge is addressed by trying to manage and normalize the data structure in order to faster searching and exploiting them in a manufacturing context. For their part, the authors present i-Dataquest, an information retrieval system supported by (i) a graph-oriented model built from the structured and unstructured data of the company and (ii) a query system answering ‘what’ and ‘about what’ and (iii) generating three different results: a list of items, a list of property values and a list of sentences. The i-Dataquest prototype is built using Neo4J for the graph system generation, ConceptNet for lexical resource management and StandfordNLP for natural language processing. An evaluation of the prototype’s performance is conducted through a data set representing a drone manufacturer. The results show that the transformation of specific content such as tables in the graph and the semantic expansion of queries significantly improves the recall and precision measures. The results also suggest improving filtering less relevant results by considering particularly queries looking for a specific value.

International audience; This paper proposes a new system-level application for monitoring out-of-step (OOS) events in power systems. As already known, amplitude-dependent frequency shift is a nonlinear phenomenon of electromechanical oscillations under large disturbances. The frequency shift indicates the system’s nearness to instability. This new tool utilizes the Normal Form method to identify the named phenomenon, leading to accelerated OOS detection. The proposed strategy is illustrated and compared to the equal-area criterion method in a single-machine-infinite-bus power system. Extensive tests on IEEE 3- and IEEE 50-machine power systems prove the efficacy and potential of the proposed method for online warnings of instability and ranking of vulnerable system modes.

International audience; In this paper, both vibration reduction and task accuracy improvement for flexible redundant manipulators, using the resolution of redundancy, is addressed. The exciting force of flexural motion, which is induced by the motion of the manipulator, causes undesired deflection at the end-effector. The null space inherent in redundancy feature is exploited to damp out the vibration while maintaining the end-effector posture. In order to exploit all the degrees of freedom in the null space, a feedback control design based on torque optimization has been introduced using analytical method for the redundancy resolution. The simulation analysis is presented to test the effectiveness of the proposed approach.

International audience; The collaborative robot (cobot) is a technology contributing to the industrial revolution Industry 4.0. Indeed, cobots’ flexibility and their easy-to-use solutions fill a gap with traditional robots to robotize manufacturing of products with low volume and high mix profiles. Parts needed to manufacture a product must first be presented to the cobot, so that afterwards it can perform operations such as assembly. The paper classifies the current part presenters. Then, a methodology is proposed to select the best presenter based on the characteristics of the parts and the workstation. To be aligned with markets requiring a high mix of products frequently renewed, the development times for part presenter design and for cobot programming are new data to select the best presenter. Indeed, at every part redesign, the part presenter changes and the cobot is reprogrammed. These two times are depreciated based on the lifespan quantity of the part. Lifespan quantity is the number of the same parts, which are dropped-off on the part presenter during the part’s lifespan, i.e. until its redesign. A concrete industrial use case is explained to test the methodology. Results conclude that a tray pattern is the best part presenter, except for parts with low lifespan quantity. Lifespan quantity of the part appears to be a significant parameter when deciding the best part presenter.

International audience; Truss lattice structures are intricate geometries, whose fabrication has recently been simplified by the development of Additive Manufacturing (AM) technologies. These lightweight geometries present great volume densities and surface-to-occupancy ratios, which makes them ideal for thermal dissipation applications. This paper introduces a new framework for the parametric design of graded truss lattice structures that maximize passive cooling. It exploits the results of a semi-analytic formulation and analysis of the volume density and surface-to-occupancy ratio of state-of-the-art unit cells. In particular, it comes out that any truss lattice structure presents an optimal beam diameter over unit cell size ratio that maximizes its surface-to-occupancy value. This value and the ratio for which it is reached are identified and compared for the most common unit cells. The unit cell with the maximal surface-to-occupancy ratio is then identified, along with its set of optimal parameters, taking into account additive manufacturing constraints. The validation of this optimal geometry is performed by populating pre-defined design spaces of both academic and industrial case studies. An orientation strategy and a parametric gradation approach are also proposed to further optimize the generated heat sinks and maximize passive cooling. These results are very helpful to support decision making during the parametric design of a heat sink and to identify, a priori, the optimal unit cell, its control parameters, its orientation and its gradation strategy. The generated geometries are compared with traditional heat sink structures through static heat dissipation simulations, in order to demonstrate their interest.

International audience; Currently, the content-based retrieval is a problem of major interest in several different fields and, focusing on mechanical engineering, many approaches exist to compare and retrieve single CAD parts, to evaluate shape similarity, to extract features and to segment models. However, most of the proposed approaches do not take into account all the key characteristics of an assembly model, such as the relationships between its components, and the different levels according to which two assembly models can be considered similar, i.e. either globally, partially, or locally. For these reasons, the retrieval of CAD assembly models still faces challenges to fully satisfy designers’ expectations. The aim of this paper is to review the state-of-the-art of works addressing the CAD assembly model retrieval and to identify future challenges and possible research directions. Firstly, the paper highlights the user requirements for CAD assembly model retrieval and proposes a set of criteria for analyzing the available methods grouped into the following macro-categories: objective, assembly characterization, assembly descriptor, query specification and type of similarity. Secondly, it describes and characterizes the available methods by organizing them according to the adopted criteria. Finally, it discusses the open issues and future challenges.

International audience; Tuned dynamic absorbers are usually used to counteract vibrations at a given frequency. Presence of non-linearities causes energy-dependent relationship of their resonance and antiresonance frequencies at large amplitude of motion, which consequently leads to adetuning of the absorber from the targeted frequency. This paper presents a procedure to track an extremum point (minimum or maximum) of nonlinear frequency responses, based on a numerical continuation technique coupled to the harmonic balance method to follow periodic solutions in forced steady-state. It thus enable to track a particularantiresonance. The procedure is tested and applied on some application cases to highlight the resonance and antiresonance behavior in presence of geometrically non-linear and/or inertial interactions.

Technological advances, promoted by the Industry 4.0 paradigm, attempt to support the reconfigurability of manufacturing systems and to contribute to adaptive operational conditions. These systems must be responsive to significant changes in demand volume and product mix. In this paper, a hybrid optimisation approach is suggested to solve sequencing and assignment problems of reconfigurable assembly lines, where mobile robots collaborate with human operators. The objectives are: i) to define a schedule of jobs, ii) to assign tasks to the mobile robots, and iii) to decide the allocation of robots to workstations, in order to minimise the number of robots required. Preliminary results show that the proposed methodology can make an efficient robot allocation under high demand variety. In addition to that, the hybrid optimisation approach can be adapted to other configurations of assembly systems, which demonstrates its applicability to solve problems in other contexts.

International audience; Since the founding theory established by G. Floquet more than a hundred years ago, computing the stability of periodic solutions has given rise to various numerical methods, mostly depending on the way the periodic solutions are themselves determined , either in the time domain or in the frequency domain. In this paper, we address the stability analysis of branches of periodic solutions that are computed by combining a pure Harmonic Balance Method (HBM) with an Asymptotic Numerical Method (ANM). HBM is a frequency domain method for determining periodic solutions under the form of Fourier series and ANM is continuation technique that relies on high order Taylor series expansion of the solutions branches with respect to a path parameter. It is well established now that this HBM-ANM combination is efficient and reliable, provided that the system of ODE is first of all recasted with quadratic nonlinearities, allowing an easy manipulation of both the Taylor and the Fourier series. In this context, Hill's method, a frequency domain version of Floquet theory, is revisited so as to become a by-product of the HBM applied to a quadratic system, allowing the stability analysis to be implemented in an elegant way and with good computing performances. The different types of stability changes of periodic solutions are all explored and illustrated through several academic examples, including systems that are autonomous or not, conservative or not, free or forced.

International audience; Friction-induced vibrations are known to affect many engineering applications. Here, we study a chain of friction-excited oscillators with nearest neighbor elastic coupling. The excitation is provided by a moving belt which moves at a certain velocity v d while friction is modelled with an exponentially decaying friction law. It is shown that in a certain range of driving velocities, multiple stable spatially localized solutions exist whose dynamical behavior (i.e. regular or irregular) depends on the number of oscillators involved in the vibration. The classical non-repeatability of friction-induced vibration problems can be interpreted in light of those multiple stable dynamical states. These states are found within a "snaking-like" bifurcation pattern. Contrary to the classical Anderson localization phenomenon, here the underlying linear system is perfectly homogeneous and localization is solely triggered by the friction nonlinearity.

International audience; In this paper, a method for the quasi-static simulation of flexible cables assembly in the context of automotive industry is presented. The cables geometry and behavior encourage to employ a geometrically exact beam model. The 3D kinematics is then based on the position of the centerline and on the orientation of the cross-sections, which is here represented by rotational quaternions. Their algebraic nature leads to a polynomial form of equilibrium equations. The continuous equations obtained are then discretized by the finite element method and easily recast under quadratic form by introducing additional slave variables. The asymptotic numerical method, a powerful solver for systems of quadratic equations, is then employed for the continuation of the branches of solution. The originality of this paper stands in the combination of all these methods which leads to a fast and accurate tool for the assembly process of cables. This is proved by running several classical validation tests and an industry-like example.

International audience; Factories with a high production throughput and a low product variety traditionally choose a flow layout for mass production. The future market profile however will be volatile, customers expecting a larger choice in product diversity. Such a variable demand calls a new, flexible and reconfigurable production system, able to reconfigure in order to follow the market. This paper aims to present a method to modelize the production system permitting to identify the ways of improving the facilities on the reconfigurability criterion. Several reconfigurable assembly systems are proposed, targeting full load of the facilities. The specificity of the study consists in the consideration of both multiproduct and volume flexibility.

At present, optical sensors are being widely used to realize high quality control or reverse engineering of products, systems, buildings, environments or human bodies. Although the intrinsic characteristics of such breakthrough technologies may vary, ensuring complete acquisition relies on the definition of the optimal acquisition planning. To this end, the view planning problem (VPP) must be solved to automatically determine the optimal positions and/or trajectories of the acquisition devices to fully cover the part to be digitized. Such an automatization of the entire acquisition process is of considerably interest in the context of Industry 4.0. The aim of this paper is to review the state of the art works addressing the view planning problem and to identify the future challenges and possible research directions. First, the paper introduces a set of criteria to analyze the available methods, grouped into several macrocategories. The categories are presented and formalized to clearly understand the backbone and similarities of the grouped methods. Second, the paper describes and characterizes the available methods, based on their analysis according to the adopted criteria. The results of this extensive analysis clearly highlight the open issues and future challenges.

International audience; Human Robot Collaboration is seen as a significant feature of Industry 4.0 implementation. Collaborative robots (cobots) are supposed to deliver superior process performance, which was so far achieved through the application of Lean Manufacturing techniques. The following case study built around the assembly process of a pneumatic cylinder, tends to analyse not only the actual benefits of cobot implementation, but also the success factors, in conjunction with Lean Manufacturing usage. Finally, this paper suggests a draft method towards the successful integration of cobot.

Virtual walking in virtual environments (VEs) requires locomotion interfaces, especially when the available physical environment is smaller than the virtual space due to virtual reality facilities limitations; many navigation approaches have been proposed according to different input conditions, target selection and speed selection. With current technologies, the virtual locomotion speed for most VR systems relies primarily on rate-control devices (e.g., joystick). The user has to manage manual adaptation of the speed, based on the size of the VE and personal preferences. However, this method cannot provide optimal speeds for locomotion as the user tends to change the speed involuntarily due to non-desired issues including collisions or simulator sickness; in this case, the user may have to adjust the speed frequently and unsmoothly, worsening the situation. Therefore, we designed a motion protector that can be embedded into the locomotion system to provide optimal speed profiles. The optimization process aims at minimizing the total jerk when the user translates from an initial position to a target, which is a common rule of the human motion model. In addition to minimization, we put constraints on speed, acceleration and jerk so that they do not exceed specific thresholds. The speed protector is formulated mathematically and solved analytically in order to provide a smooth navigation experience with a minimum jerk of trajectory. The assessment of the speed protector was conducted in a user study measuring user experience with a simulator sickness questionnaire, event-related skin conductance responses (ER-SCR), and a NASA-TLX questionnaire, showing that the designed speed protector can provide more natural and comfortable user experience with appropriate acceleration and jerk as it avoids abrupt speed profiles. China Scholarship Council: No. 201708390014

International audience; The design of complex engineered systems highly relies on a laborious zigzagging between computer-aided design (CAD) software and design rules prescribed by design manuals. Despite the emergence ofknowledge management techniques (ontology, expert system, text mining, etc.), companies continue tostore design rules in large and unstructured documents. To facilitate the integration of design rules andCAD software, we propose a knowledge graph that structures a large set of design rules in a computableformat. The knowledge graph organises entities of design rules (nodes), relationships among design rules(edges), as well as contextual information. The categorisation of entities and relationships in four sub-contexts: semantic, social, engineering, and IT – facilitates the development of the data model, especiallythe definition of the “design context” concept. The knowledge graph paves the way to a context-awarecognitive design assistant. Indeed, connected to or embedded in a CAD software, a context-aware cog-nitive design assistant will capture the design context in near real time and run reasoning operationson the knowledge graph to extend traditional CAD capabilities, such as the recommendation of designrules, the verification of design solutions, or the automation of design routines. Our validation experi-ment shows that the current version of the context-aware cognitive design assistant is more efficientthan the traditional document-based design. On average, participants using an unstructured design rulesdocument have a precision of 0.36 whereas participants using our demonstrator obtain a 0.61 precisionscore. Finally, designers supported by the design assistant spend more time designing than searching forapplicable design rules compared to the traditional design approach.

To facilitate the transfer, storage and manipulation of intricate parts’ geometry whose fabrication has been made possible thanks to the rise of Additive Manufacturing (AM) technologies, an encoding framework reducing the resulting file size has been developed. This approach leverages the fact that many AM parts are presenting repetition patterns, by encoding the repetition of similar geometry chunks. The decomposition of the part into chunks is a complex optimization problem, whose identification as a Weighted Exact Cover (WEC) problem allowed to develop a new heuristic algorithm dedicated to its fast resolution in linear time O ( n ) . The encoding strategy is implemented through a variation of the AMF file standard (for quick adoption of the format by existing software), and also through a new ad-hoc hybrid file format. To demonstrate the efficiency of the approach, the encryption of lattice and support structures through these two encoding strategies are compared to the results of several state-of-the-art encoding approaches. The way this data weight lightening strategy preserves the overall accuracy is discussed while considering different floating points encoding precisions with respect to the AM process requirements. This comparison exhibits file size reductions up to -84% in comparison with file sizes generated by state-of-the-art approaches. Not only the proposed repetition pattern encoding framework allows file size reductions, but it could also be exploited to optimize and speed-up some steps of the Product Development Process (PDP), including process planning phases.

International audience; In this work we apply and compare two algorithms for setting up Harmonic Balance equations and numerical continuation of the solution path for harmonically driven mechanical systems. The first algorithm relies on a predictor-corrector scheme and an Alternating Frequency-Time approach (AFT-PreCo). The second algorithm relies on a high-order Taylor series expansion of the solution path (asymptotic numerical method) and classical Harmonic Balance formulated entirely in the frequency domain (cHB-ANM). We conclude that the cHB-ANM is suited for smooth nonlinearities, for instance geometrically nonlinear finite element models. Here, cHB-ANM avoids aliasing errors and convinces with a numerically robust adjustment of the continuation step length and a continuous representation of the solution path. For non-smooth nonlinearities such as stick-slip friction or unilateral constraints, AFT-PreCo is better suited, and convinces with high numerical robustness and efficiency.