- Robotic Locomotion and Control
- Prosthetics and Rehabilitation Robotics
- Robotic Mechanisms and Dynamics
- Robotic Path Planning Algorithms
- Robot Manipulation and Learning
- Muscle Physiology and Disorders
- Digital Platforms and Economics
- ICT Impact and Policies
- Modular Robots and Swarm Intelligence
- Robotics and Sensor-Based Localization
- Real-time simulation and control systems
- Sports Dynamics and Biomechanics
- Human Pose and Action Recognition
- Advanced Data Storage Technologies
- Peer-to-Peer Network Technologies
- Caching and Content Delivery
- Computational Geometry and Mesh Generation
- Neurogenetic and Muscular Disorders Research
- Data Stream Mining Techniques
- Human Motion and Animation
- Neurology and Historical Studies
- Advanced Data Compression Techniques
- Physical Unclonable Functions (PUFs) and Hardware Security
- Maritime Transport Emissions and Efficiency
- Advanced Bandit Algorithms Research
École Normale Supérieure - PSL
2010-2024
Département d'Informatique
2012-2024
Institut national de recherche en informatique et en automatique
2024
Laboratoire d'Informatique, de Robotique et de Microélectronique de Montpellier
2014-2021
Centre National de la Recherche Scientifique
2014-2021
Union Bank of Switzerland
2020-2021
Hatch (Canada)
2021
Université de Montpellier
2014-2020
Institute of Electrical and Electronics Engineers
2019
Gorgias Press (United States)
2019
In this paper, we study Demand Response (DR) problematics for different levels of information sharing in a smart grid. We propose dynamic pricing scheme incentivizing consumers to achieve an aggregate load profile suitable utilities, and how close they can get ideal flat depending on much share. When customers share all their profiles, provide distributed algorithm, set up as cooperative game between consumers, which significantly reduces the total cost peak-to-average ratio (PAR) system....
We consider dynamic stair climbing with the HRP-4 humanoid robot as part of an Airbus manufacturing use-case demonstrator. share experimental knowledge gathered so to achieve this task, which had never been challenged before. In particular, we extend walking stabilization based on linear inverted pendulum tracking [1] by quadratic programming-based wrench distribution and a whole-body admittance controller that applies both end-effector CoM strategies. While existing stabilizers tend use...
We propose a method for checking and enforcing multicontact stability based on the zero-tilting moment point (ZMP). The key to our development is generalization of ZMP support areas take into account: 1) frictional constraints 2) multiple noncoplanar contacts. introduce investigate two kinds areas. First, we characterize provide fast geometric construction area generated by valid contact forces, with no other constraint robot motion. call this set full area. Next, consider control humanoid...
Humanoids locomote by making and breaking contacts with their environment. Thus, a crucial question for them is to anticipate whether contact will hold or break under effort. For rigid surface contacts, existing methods usually consider several point-contact forces, which has some drawbacks due the underlying redundancy. We derive criterion, Contact Wrench Cone (CWC), equivalent any number of applied forces on surface, we provide closed-form formula. It turns out that CWC can be decomposed...
We report on the results of a collaborative project that investigated deployment humanoid robotic solutions in air-craft manufacturing for several assembly op erations where access by wheeled or railported platforms is not possible. Recent de velopments multicontact planning and control, bipedal walking, embedded simultaneous localization mapping (SLAM), whole-body multisensory task-space optimization contact detection safety suggest humanoids could be plausible solution automation, given...
Improving maritime operations planning and scheduling can play an important role in enhancing the sector’s performance competitiveness. In this context, accurate ship speed estimation is crucial to ensure efficient traffic management. This study addresses problem of prediction from a Maritime Vessel Services perspective area Saint Lawrence Seaway. The challenge build real-time predictive model that accommodates different routes vessel types. proposes data-driven solution based on deep...
We build on previous works advocating the use of Gravito-Inertial Wrench Cone (GIWC) as a general contact stability criterion (a "ZMP for non-coplanar contacts").We show how to compute this wrench cone from friction cones forces by using an intermediate representation, surface cone, which is minimal representation each contact.The observation that GIWC needs be computed only once per stance leads particularly efficient algorithms, we illustrate in two important problems humanoids: "testing...
Capturability analysis of the linear inverted pendulum (LIP) model enabled walking with constrained height based on capture point. In this paper, we generalize to variable-height (VHIP) and show how it enables 3-D over uneven terrains inputs. Thanks a tailored optimization scheme, can compute these inputs fast enough for real-time predictive control. We implement approach as open-source software demonstrate in dynamic simulations.
We present a multi-contact walking pattern generator based on preview-control of the 3D acceleration center mass (COM). A key point in design our algorithm is calculation contact-stability constraints. Thanks to mathematical observation algebraic nature frictional wrench cone, we show that volume feasible COM accelerations always an upward-pointing cone. reduce its computation convex hull (dual) 2D points, for which optimal C(n log n) algorithms are readily available. This reformulation...
This paper considers stochastic bandits with side observations, a model that accounts for both the exploration/exploitation dilemma and relationships between arms. In this setting, after pulling an arm i, decision maker also observes rewards some other actions related to i. We will see is suited content recommendation in social networks, where users' reactions may be endorsed or not by their friends. provide efficient algorithms based on upper confidence bounds (UCBs) leverage additional...
Path-velocity decomposition is an intuitive yet powerful approach to addressing the complexity of kinodynamic motion planning. The difficult trajectory planning problem solved in two separate, simpler steps: first, a path found configuration space that satisfies geometric constraints (path planning), and second, time-parameterization satisfying found. A fundamental requirement first step must be time-parameterizable. Most existing works fulfill this by enforcing quasi-static during step,...
In legged locomotion, the projection of robot's Center Mass (CoM) being inside convex hull contact points is a commonly accepted sufficient condition to achieve static balancing. However, some these configurations cannot be realized because joint-torques required sustain them would above their limits (actuation limits). this article, we rule out such and define feasible region, revisited support region that guarantees both global stability in sense tip-over slippage avoidance existence set...
We propose a method that enables kinodynamic planning in the configuration space (of dimension n) instead of state 2n), thereby potentially cutting down complexity usual algorithms by an exponential factor.At heart this is new technique -called Admissible Velocity Propagation (AVP)which, given path and interval reachable velocities at beginning path, computes exactly efficiently all system can reach after traversing while respecting constraints.Combining with sampling-based methods gives...
We present a model predictive controller (MPC) for multi-contact locomotion where optimizations are realized by time-optimal path parameterization (TOPP). A key feature of this solution is that, contrary to existing planners step timings provided as inputs, here the timing between contact switches computed output fast nonlinear optimization. This appealing locomotion, proper depend on terrain topology and suitable heuristics unknown. show how formulate legged TOPP problem demonstrate...
The ongoing debate over net neutrality covers a broad set of issues related to the regulation public networks. In two ways, we extend an idealized usage-priced game-theoretic framework based on common linear demand-response model [1]. First, study impact "side payments" among plurality Internet service (access) providers and content providers. non-monopolistic case, our analysis reveals interesting "paradox" side payments in that overall revenues are reduced for those receive them. Second,...
We present a real-time pattern generator for dynamic walking over rough terrains. Our method automatically finds step durations, critical issue terrains where they depend on terrain topology. To achieve this level of generality, we consider Floating-base Inverted Pendulum (FIP) model the center mass can translate freely and zero-tilting moment point is allowed to leave contact surface. This equivalent linear inverted pendulum with variable center-of-mass height, but its equations motion...
A novel control scheme for biped robots to manipulate the ZMP three-dimensionally apart from actual ground profile is presented. It shown that linear inverted-pendulum-like dynamics with this can represent a wider class of movements including variation body height. Moreover, also motion in aerial phase. Based on this, foot-guided controller proposed by authors enhanced enable locomote highly uneven terrains and seamlessly transition between walking running without pre-planning overall...
Developments for 3D control of the center mass (CoM) biped robots are currently located in two local minima: on one hand, methods that allow CoM height variations but only work 2D sagittal plane; other nonconvex direct transcriptions centroidal dynamics delicate to handle. This paper presents an alternative controls via indirect transcription is both low-dimensional and solvable fast enough real-time control. The key this development notion boundedness condition, which quantifies...
The variable-height inverted pendulum (VHIP) model enables a new balancing strategy by height variations of the center mass, in addition to well-known ankle strategy. We propose biped stabilizer based on linear feedback VHIP that is simple implement, coincides with state-of-the-art for small perturbations and able recover from larger thanks this This solution "best-effort" pole placement 4D divergent component motion under input feasibility state viability constraints. complement it suitable...
We consider control of a humanoid robot in active compliance just after the impact consecutive to fall. The goal this post-impact braking is absorb undesired linear momentum accumulated during fall, using limited supply time and actuation power. gist our method an optimal distribution between robot's hand foot contact points, followed by parallel resolution Linear Model Predictive Control (LMPC) at each contact. This made possible thanks torque-limited friction polytopes, extension cones...
Recommending items to new or "cold-start" users is a challenging problem for recommender systems. Collaborative filtering approaches fail when the preference history of not available. A promising direction that has been explored recently [12] utilize information in social networks improve quality cold-start recommendations. That is, given are part network, user shows up network with no and limited links, system tries learn user's tastes as fast possible.
Soft soles absorb impacts and cast ground unevenness during locomotion on rough terrains. However, they introduce passive degrees of freedom (deformations under the feet) that complexify tasks state estimation overall robot stabilization. We address this problem by developing a control loop stabilizes humanoid robots when walking with soft flat uneven terrain. Our closed-loop controller minimizes errors center mass (COM) zero moment point (ZMP) an admittance feet based simple deformation...
We propose a method for checking and enforcing multi-contact stability based on the Zero-tilting Moment Point (ZMP). The key to our development is generalization of ZMP support areas take into account (a) frictional constraints (b) multiple non-coplanar contacts. introduce investigate two kinds areas. First, we characterize provide fast geometric construction area generated by valid contact forces, with no other constraint robot motion. call this set full area. Next, consider control...