2018 IEEE Workshop on Advanced Robotics and its Social Impacts (ARSO), 2018
A physical Human-Robot Interaction (pHRI) framework is proposed using vision and force sensors fo... more A physical Human-Robot Interaction (pHRI) framework is proposed using vision and force sensors for a two-way object handover task. Kinect v2 is integrated with the state-of-the-art 2D skeleton extraction library namely Openpose to obtain a 3D skeleton of the human operator. A robust and rotation invariant (in the coronal plane) hand gesture recognition system is developed by exploiting a convolutional neural network. This network is trained such that the gestures can be recognized without the need to pre-process the RGB hand images at run time. This work establishes a firm basis for the robot control using hand-gestures. This will be extended for the development of intelligent human intention detection in pHRI scenarios to efficiently recognize a variety of static as well as dynamic gestures.
HAL (Le Centre pour la Communication Scientifique Directe), Aug 21, 2020
We present a unique framework for manipulating both rigid and deformable objects. • Our framework... more We present a unique framework for manipulating both rigid and deformable objects. • Our framework is model-free and requires a short initialization phase. • Our framework does not require camera calibration, and works with different camera poses.
AbstractThis paper details the development of an adaptive control architecture permitting to imp... more AbstractThis paper details the development of an adaptive control architecture permitting to improve the reliability and robustness of autonomous mobile robot. A continuous monitoring of the significant failures allows dynamically choosing the most relevant ...
This paper details the development of an adaptive control architecture permitting to improve the ... more This paper details the development of an adaptive control architecture permitting to improve the reliability and robustness of autonomous mobile robot. A continuous monitoring of the significant failures allows dynamically choosing the most relevant reaction ensuring the success of the mission. This adaptive behavior is implemented into the control architecture COTAMA. The key points of the specific mechanisms added to COTAMA are addressed and explained. Experimental results are proposed to illustrate the control architecture adaptive behavior.
We present a unique framework for manipulating both rigid and deformable objects. • Our framework... more We present a unique framework for manipulating both rigid and deformable objects. • Our framework is model-free and requires a short initialization phase. • Our framework does not require camera calibration, and works with different camera poses.
HAL (Le Centre pour la Communication Scientifique Directe), Mar 26, 2018
Navigation tasks are often subject to several constraints that can be related to the sensors (vis... more Navigation tasks are often subject to several constraints that can be related to the sensors (visibility) or come from the environment (obstacles). In this paper, we propose a framework for autonomous omnidirectional wheeled robots, that takes into account both collision and occlusion risk, during sensor-based navigation. The task consists in driving the robot towards a visual target in the presence of static and moving obstacles. The target is acquired by fixed-limited field of viewon-board sensors, while the surrounding obstacles are detected by lidar scanners. To perform the task, the robot has not only to keep the target in view while avoiding the obstacles, but also to predict its location in the case of occlusion. The effectiveness of our approach is validated through several experiments.
2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids), Nov 28, 2022
This paper proposes a novel framework for recognizing industrial actions, in the perspective of h... more This paper proposes a novel framework for recognizing industrial actions, in the perspective of human-robot collaboration. Given a one second long measure of the human's motion, the framework can determine his/her action. The originality lies in the use of joint angles, instead of Cartesian coordinates. This design choice makes the framework sensor agnostic and invariant to affine transformations and to anthropometric differences. On AnDy dataset, we outperform the state of art classifier. Furthermore, we show that our framework is effective with limited training data, that it is subject independent, and that it is compatible with robotic realtime constraints. In terms of methodology, the framework is an original synergy of two antithetical schools of thought: modelbased and data-based algorithms. Indeed, it is the cascade of an inverse kinematics estimator compliant with the International Society of Biomechanics recommendations, followed by a deep learning architecture based on Bidirectional Long Short Term Memory. We believe our work may pave the way to successful and fast action recognition with standard depth cameras, embedded on moving collaborative robots.
2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids)
This paper proposes a novel framework for recognizing industrial actions, in the perspective of h... more This paper proposes a novel framework for recognizing industrial actions, in the perspective of human-robot collaboration. Given a one second long measure of the human's motion, the framework can determine his/her action. The originality lies in the use of joint angles, instead of Cartesian coordinates. This design choice makes the framework sensor agnostic and invariant to affine transformations and to anthropometric differences. On AnDy dataset, we outperform the state of art classifier. Furthermore, we show that our framework is effective with limited training data, that it is subject independent, and that it is compatible with robotic realtime constraints. In terms of methodology, the framework is an original synergy of two antithetical schools of thought: modelbased and data-based algorithms. Indeed, it is the cascade of an inverse kinematics estimator compliant with the International Society of Biomechanics recommendations, followed by a deep learning architecture based on Bidirectional Long Short Term Memory. We believe our work may pave the way to successful and fast action recognition with standard depth cameras, embedded on moving collaborative robots.
HAL (Le Centre pour la Communication Scientifique Directe), Oct 30, 2017
Emerging industrial applications involving mobile manipulation in the presence of humans is drivi... more Emerging industrial applications involving mobile manipulation in the presence of humans is driving attention towards steerable wheeled mobile robots (SWMR), since these can perform arbitrary 2D planar trajectories, providing a reasonable compromise between maneuverability (necessary for human avoiding algorithms) and effectiveness. Instantaneous center of rotation (ICR) based kinematic models and controllers are the most suited for such robots, as they assure the existence of a unique ICR point at all times. However, unsatisfactory behavior do exist in numerous applications requiring frequent changes in the sign of the angular velocity command. This is typically the case for robot heading control: moving the ICR point from one border of the 2D ICR space to the other makes it pass by the robot geometric center, where only pure rotations are feasible. This behavior is not desirable and should be avoided. In this paper, we propose a novel complementary route ICR controller, where the ICR can go from one extreme to the other by means of border switching in one sample period. Thanks to this approach, fast response to the velocity commands is achieved with little steering motion. The new algorithm has been tested successfully in simulations and experiments, and is more time efficient with far more satisfactory behavior than the state-of-art direct route based controllers. These results have been also confirmed quantitatively, using a newly developed metric, the command fulfillment index (CFI).
Ce rapport presente la methodologie ContrACT dediee au developpement d'architecture de contro... more Ce rapport presente la methodologie ContrACT dediee au developpement d'architecture de controle logicielle en robotique. La proposition centrale repose sur l'utilisation de modules logiciels temps-reel, interconnectes dynamiquement, afin de mettre en oeuvre la strategie de commande requise en fonction des informations contextuelles a la disposition du robot. Ce document couvre a la fois les aspects conceptuels (concepts metiers, modeles) et les aspects technologiques (middleware, programmation temps-reel) de cette methodologie.
Robotic software is now one of essential part of robotic system development, therefore software c... more Robotic software is now one of essential part of robotic system development, therefore software control architecture design methods and concepts, often inspired by engineering software field, are necessary within a robotic project to enhance evolution, modularity and re-usability, and to avoid redesign costs [1]. Control software architecture design approaches are usually classified into three main categories: Reactive architectures are built by gathering several modules called behaviors. Each behavior reacts continuously to the situation sensed by the perception system [4]. Actuators control values are obtained from a weighted summation of all commands generated by these modules. The complexity of this method lies in weights adjustment allowing a good reaction and keeping global objective [5][8]. Deliberative architectures are built in several levels, usually three. Decisions are taken in the higher level; the intermediate level is in charge of controlling and supervising. The low ...
This paper summarizes recent (2011-2016) research carried out within the LIRMM IDH group, to addr... more This paper summarizes recent (2011-2016) research carried out within the LIRMM IDH group, to address the development of collaborative robots for industrial applications. The presented works have been carried out in the frame of various projects, involving major European industrial actors such as PSA Peugeot Citroen, Airbus, and the Tecnalia Foundation.
2018 IEEE Workshop on Advanced Robotics and its Social Impacts (ARSO), 2018
A physical Human-Robot Interaction (pHRI) framework is proposed using vision and force sensors fo... more A physical Human-Robot Interaction (pHRI) framework is proposed using vision and force sensors for a two-way object handover task. Kinect v2 is integrated with the state-of-the-art 2D skeleton extraction library namely Openpose to obtain a 3D skeleton of the human operator. A robust and rotation invariant (in the coronal plane) hand gesture recognition system is developed by exploiting a convolutional neural network. This network is trained such that the gestures can be recognized without the need to pre-process the RGB hand images at run time. This work establishes a firm basis for the robot control using hand-gestures. This will be extended for the development of intelligent human intention detection in pHRI scenarios to efficiently recognize a variety of static as well as dynamic gestures.
HAL (Le Centre pour la Communication Scientifique Directe), Aug 21, 2020
We present a unique framework for manipulating both rigid and deformable objects. • Our framework... more We present a unique framework for manipulating both rigid and deformable objects. • Our framework is model-free and requires a short initialization phase. • Our framework does not require camera calibration, and works with different camera poses.
AbstractThis paper details the development of an adaptive control architecture permitting to imp... more AbstractThis paper details the development of an adaptive control architecture permitting to improve the reliability and robustness of autonomous mobile robot. A continuous monitoring of the significant failures allows dynamically choosing the most relevant ...
This paper details the development of an adaptive control architecture permitting to improve the ... more This paper details the development of an adaptive control architecture permitting to improve the reliability and robustness of autonomous mobile robot. A continuous monitoring of the significant failures allows dynamically choosing the most relevant reaction ensuring the success of the mission. This adaptive behavior is implemented into the control architecture COTAMA. The key points of the specific mechanisms added to COTAMA are addressed and explained. Experimental results are proposed to illustrate the control architecture adaptive behavior.
We present a unique framework for manipulating both rigid and deformable objects. • Our framework... more We present a unique framework for manipulating both rigid and deformable objects. • Our framework is model-free and requires a short initialization phase. • Our framework does not require camera calibration, and works with different camera poses.
HAL (Le Centre pour la Communication Scientifique Directe), Mar 26, 2018
Navigation tasks are often subject to several constraints that can be related to the sensors (vis... more Navigation tasks are often subject to several constraints that can be related to the sensors (visibility) or come from the environment (obstacles). In this paper, we propose a framework for autonomous omnidirectional wheeled robots, that takes into account both collision and occlusion risk, during sensor-based navigation. The task consists in driving the robot towards a visual target in the presence of static and moving obstacles. The target is acquired by fixed-limited field of viewon-board sensors, while the surrounding obstacles are detected by lidar scanners. To perform the task, the robot has not only to keep the target in view while avoiding the obstacles, but also to predict its location in the case of occlusion. The effectiveness of our approach is validated through several experiments.
2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids), Nov 28, 2022
This paper proposes a novel framework for recognizing industrial actions, in the perspective of h... more This paper proposes a novel framework for recognizing industrial actions, in the perspective of human-robot collaboration. Given a one second long measure of the human's motion, the framework can determine his/her action. The originality lies in the use of joint angles, instead of Cartesian coordinates. This design choice makes the framework sensor agnostic and invariant to affine transformations and to anthropometric differences. On AnDy dataset, we outperform the state of art classifier. Furthermore, we show that our framework is effective with limited training data, that it is subject independent, and that it is compatible with robotic realtime constraints. In terms of methodology, the framework is an original synergy of two antithetical schools of thought: modelbased and data-based algorithms. Indeed, it is the cascade of an inverse kinematics estimator compliant with the International Society of Biomechanics recommendations, followed by a deep learning architecture based on Bidirectional Long Short Term Memory. We believe our work may pave the way to successful and fast action recognition with standard depth cameras, embedded on moving collaborative robots.
2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids)
This paper proposes a novel framework for recognizing industrial actions, in the perspective of h... more This paper proposes a novel framework for recognizing industrial actions, in the perspective of human-robot collaboration. Given a one second long measure of the human's motion, the framework can determine his/her action. The originality lies in the use of joint angles, instead of Cartesian coordinates. This design choice makes the framework sensor agnostic and invariant to affine transformations and to anthropometric differences. On AnDy dataset, we outperform the state of art classifier. Furthermore, we show that our framework is effective with limited training data, that it is subject independent, and that it is compatible with robotic realtime constraints. In terms of methodology, the framework is an original synergy of two antithetical schools of thought: modelbased and data-based algorithms. Indeed, it is the cascade of an inverse kinematics estimator compliant with the International Society of Biomechanics recommendations, followed by a deep learning architecture based on Bidirectional Long Short Term Memory. We believe our work may pave the way to successful and fast action recognition with standard depth cameras, embedded on moving collaborative robots.
HAL (Le Centre pour la Communication Scientifique Directe), Oct 30, 2017
Emerging industrial applications involving mobile manipulation in the presence of humans is drivi... more Emerging industrial applications involving mobile manipulation in the presence of humans is driving attention towards steerable wheeled mobile robots (SWMR), since these can perform arbitrary 2D planar trajectories, providing a reasonable compromise between maneuverability (necessary for human avoiding algorithms) and effectiveness. Instantaneous center of rotation (ICR) based kinematic models and controllers are the most suited for such robots, as they assure the existence of a unique ICR point at all times. However, unsatisfactory behavior do exist in numerous applications requiring frequent changes in the sign of the angular velocity command. This is typically the case for robot heading control: moving the ICR point from one border of the 2D ICR space to the other makes it pass by the robot geometric center, where only pure rotations are feasible. This behavior is not desirable and should be avoided. In this paper, we propose a novel complementary route ICR controller, where the ICR can go from one extreme to the other by means of border switching in one sample period. Thanks to this approach, fast response to the velocity commands is achieved with little steering motion. The new algorithm has been tested successfully in simulations and experiments, and is more time efficient with far more satisfactory behavior than the state-of-art direct route based controllers. These results have been also confirmed quantitatively, using a newly developed metric, the command fulfillment index (CFI).
Ce rapport presente la methodologie ContrACT dediee au developpement d'architecture de contro... more Ce rapport presente la methodologie ContrACT dediee au developpement d'architecture de controle logicielle en robotique. La proposition centrale repose sur l'utilisation de modules logiciels temps-reel, interconnectes dynamiquement, afin de mettre en oeuvre la strategie de commande requise en fonction des informations contextuelles a la disposition du robot. Ce document couvre a la fois les aspects conceptuels (concepts metiers, modeles) et les aspects technologiques (middleware, programmation temps-reel) de cette methodologie.
Robotic software is now one of essential part of robotic system development, therefore software c... more Robotic software is now one of essential part of robotic system development, therefore software control architecture design methods and concepts, often inspired by engineering software field, are necessary within a robotic project to enhance evolution, modularity and re-usability, and to avoid redesign costs [1]. Control software architecture design approaches are usually classified into three main categories: Reactive architectures are built by gathering several modules called behaviors. Each behavior reacts continuously to the situation sensed by the perception system [4]. Actuators control values are obtained from a weighted summation of all commands generated by these modules. The complexity of this method lies in weights adjustment allowing a good reaction and keeping global objective [5][8]. Deliberative architectures are built in several levels, usually three. Decisions are taken in the higher level; the intermediate level is in charge of controlling and supervising. The low ...
This paper summarizes recent (2011-2016) research carried out within the LIRMM IDH group, to addr... more This paper summarizes recent (2011-2016) research carried out within the LIRMM IDH group, to address the development of collaborative robots for industrial applications. The presented works have been carried out in the frame of various projects, involving major European industrial actors such as PSA Peugeot Citroen, Airbus, and the Tecnalia Foundation.
Uploads
Papers by robin passama