Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9035071
Hirofumi Shin, Tetsuya Ishikawa, Takumi Kamioka, K. Hosoda, T. Yoshiike
To achieve robotic walking, a successful approach is to approximate a robots dynamics as a simplified model. However, the difference between the mechanistic properties of a robot and the simplified model causes a problem of unstable and inefficient walking. To solve this problem mechanically, this paper proposes a design principle for the leg structures of bipedal robots that match the mechanistic properties of a simplified model, specifically the spring-loaded inverted pendulum (SLIP) model. The SLIP model is widely applied to robots because it has passive stability and dynamic properties similar to those of animal gaits. We have analyzed the effects of parameters of five-bar linkages with springs as a part of the leg structure of a bipedal robot. Our analysis showed that the spring parameters can impart the same mechanistic properties as the SLIP model in any configuration of a five-bar parallel mechanism. Moreover, a simplified case of a parallel linkage structure using two springs with the same properties also produced the mechanical properties of the SLIP model. These theoretical analyses were also validated with an experimental model.
{"title":"Mechanistic Properties of Five-bar Parallel Mechanism for Leg Structure Based on Spring Loaded Inverted Pendulum","authors":"Hirofumi Shin, Tetsuya Ishikawa, Takumi Kamioka, K. Hosoda, T. Yoshiike","doi":"10.1109/Humanoids43949.2019.9035071","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035071","url":null,"abstract":"To achieve robotic walking, a successful approach is to approximate a robots dynamics as a simplified model. However, the difference between the mechanistic properties of a robot and the simplified model causes a problem of unstable and inefficient walking. To solve this problem mechanically, this paper proposes a design principle for the leg structures of bipedal robots that match the mechanistic properties of a simplified model, specifically the spring-loaded inverted pendulum (SLIP) model. The SLIP model is widely applied to robots because it has passive stability and dynamic properties similar to those of animal gaits. We have analyzed the effects of parameters of five-bar linkages with springs as a part of the leg structure of a bipedal robot. Our analysis showed that the spring parameters can impart the same mechanistic properties as the SLIP model in any configuration of a five-bar parallel mechanism. Moreover, a simplified case of a parallel linkage structure using two springs with the same properties also produced the mechanical properties of the SLIP model. These theoretical analyses were also validated with an experimental model.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114260528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9035034
Akhil S. Anand, Guoping Zhao, H. Roth, A. Seyfarth
A gait model capable of generating human-like walking behavior at both the kinematic and the muscular level can be a very useful framework for developing control schemes for humanoids and wearable robots such as exoskeletons and prostheses. In this work we demonstrated the feasibility of using deep reinforcement learning based approach for neuromuscular gait modelling. A lower limb gait model consists of seven segments, fourteen degrees of freedom, and twenty two Hill-type muscles was built to capture human leg dynamics and the characteristics of muscle properties. We implemented the proximal policy optimization algorithm to learn the sensory-motor mappings (control policy) and generate human-like walking behavior for the model. Human motion capture data, muscle activation patterns and metabolic cost estimation were included in the reward function for training. The results show that the model can closely reproduce the human kinematics and ground reaction forces during walking. It is capable of generating human walking behavior in a speed range from 0.6 m/s to 1.2 m/s. It is also able to withstand unexpected hip torque perturbations during walking. We further explored the advantages of using the neuromuscular based model over the ideal joint torque based model. We observed that the neuromuscular model is more sample efficient compared to the torque model.
{"title":"A deep reinforcement learning based approach towards generating human walking behavior with a neuromuscular model","authors":"Akhil S. Anand, Guoping Zhao, H. Roth, A. Seyfarth","doi":"10.1109/Humanoids43949.2019.9035034","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035034","url":null,"abstract":"A gait model capable of generating human-like walking behavior at both the kinematic and the muscular level can be a very useful framework for developing control schemes for humanoids and wearable robots such as exoskeletons and prostheses. In this work we demonstrated the feasibility of using deep reinforcement learning based approach for neuromuscular gait modelling. A lower limb gait model consists of seven segments, fourteen degrees of freedom, and twenty two Hill-type muscles was built to capture human leg dynamics and the characteristics of muscle properties. We implemented the proximal policy optimization algorithm to learn the sensory-motor mappings (control policy) and generate human-like walking behavior for the model. Human motion capture data, muscle activation patterns and metabolic cost estimation were included in the reward function for training. The results show that the model can closely reproduce the human kinematics and ground reaction forces during walking. It is capable of generating human walking behavior in a speed range from 0.6 m/s to 1.2 m/s. It is also able to withstand unexpected hip torque perturbations during walking. We further explored the advantages of using the neuromuscular based model over the ideal joint torque based model. We observed that the neuromuscular model is more sample efficient compared to the torque model.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114298434","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9035012
Yuya Koga, Kento Kawaharazuka, Moritaka Onitsuka, Tasuku Makabe, Kei Tsuzuki, Yusuke Omura, Yuki Asano, K. Okada, M. Inaba
In recent years, some research on musculoskeletal humanoids is in progress. However, there are some challenges such as unmeasurable transformation of body structure and muscle path, and difficulty in measuring own motion because of lack of joint angle sensor. In this study, we suggest two motion acquisition methods. One is a method to acquire antagonistic relations of muscles by tension sensing, and the other is a method to acquire correct hand trajectory by vision sensing. Finally, we realize badminton shuttlecock-hitting motion of “Kengoro” with these two acquisition methods.
{"title":"Modification of muscle antagonistic relations and hand trajectory on the dynamic motion of Musculoskeletal Humanoid","authors":"Yuya Koga, Kento Kawaharazuka, Moritaka Onitsuka, Tasuku Makabe, Kei Tsuzuki, Yusuke Omura, Yuki Asano, K. Okada, M. Inaba","doi":"10.1109/Humanoids43949.2019.9035012","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035012","url":null,"abstract":"In recent years, some research on musculoskeletal humanoids is in progress. However, there are some challenges such as unmeasurable transformation of body structure and muscle path, and difficulty in measuring own motion because of lack of joint angle sensor. In this study, we suggest two motion acquisition methods. One is a method to acquire antagonistic relations of muscles by tension sensing, and the other is a method to acquire correct hand trajectory by vision sensing. Finally, we realize badminton shuttlecock-hitting motion of “Kengoro” with these two acquisition methods.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134467270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9034992
M. Dežman, T. Asfour, A. Ude, A. Gams
The wrist pronation and supination movement is important in everyday manipulation tasks. Users with limitations in this particular movement have severe impairment. While advanced upper-arm exoskeletons can assist in the pronation/supination movement, typically, the resulting exoskeleton frame that combines both the elbow joint and pronation/supination mechanism becomes heavy and bulky with a large volume. We propose a new arm pronation supination mechanism that is integrated into the exoskeleton frame and has a reduced weight and volume penalty. The mechanism functions via a double rod system, where the rods are guided through a set of specially shaped grooves that finally result in the rotation of the wrist component. The paper presents a plastic rapid prototype built using 3D additive technologies. The mechanism is actuated via a Bowden cable transmission. Its underlying kinematics are experimentally evaluated using an external motion capture system to identify its advantages and disadvantages.
{"title":"Exoskeleton Arm Pronation/Supination Assistance Mechanism With A Guided Double Rod System","authors":"M. Dežman, T. Asfour, A. Ude, A. Gams","doi":"10.1109/Humanoids43949.2019.9034992","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9034992","url":null,"abstract":"The wrist pronation and supination movement is important in everyday manipulation tasks. Users with limitations in this particular movement have severe impairment. While advanced upper-arm exoskeletons can assist in the pronation/supination movement, typically, the resulting exoskeleton frame that combines both the elbow joint and pronation/supination mechanism becomes heavy and bulky with a large volume. We propose a new arm pronation supination mechanism that is integrated into the exoskeleton frame and has a reduced weight and volume penalty. The mechanism functions via a double rod system, where the rods are guided through a set of specially shaped grooves that finally result in the rotation of the wrist component. The paper presents a plastic rapid prototype built using 3D additive technologies. The mechanism is actuated via a Bowden cable transmission. Its underlying kinematics are experimentally evaluated using an external motion capture system to identify its advantages and disadvantages.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131569661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9035067
Elisa Maiettini, Giulia Pasquale, V. Tikhanoff, L. Rosasco, L. Natale
Research in Computer Vision and Deep Learning has recently proposed numerous effective techniques for detecting objects in an image. In general, these employ deep Convolutional Neural Networks trained end-to-end on large datasets annotated with object labels and 2D bounding boxes. These methods provide remarkable performance, but are particularly expensive in terms of training data and supervision. Hence, modern object detection algorithms are difficult to be deployed in robotic applications that require on-line learning. In this paper, we propose a weakly supervised strategy for training an object detector in this scenario. The main idea is to let the robot iteratively grow a training set by combining autonomously annotated examples, with others that are requested for human supervision. We evaluate our method on two experiments with data acquired from the iCub and R1 humanoid platforms, showing that it significantly reduces the number of human annotations required, without compromising performance. We also show the effectiveness of this approach when adapting the detector to a new setting.
{"title":"A Weakly Supervised Strategy for Learning Object Detection on a Humanoid Robot","authors":"Elisa Maiettini, Giulia Pasquale, V. Tikhanoff, L. Rosasco, L. Natale","doi":"10.1109/Humanoids43949.2019.9035067","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035067","url":null,"abstract":"Research in Computer Vision and Deep Learning has recently proposed numerous effective techniques for detecting objects in an image. In general, these employ deep Convolutional Neural Networks trained end-to-end on large datasets annotated with object labels and 2D bounding boxes. These methods provide remarkable performance, but are particularly expensive in terms of training data and supervision. Hence, modern object detection algorithms are difficult to be deployed in robotic applications that require on-line learning. In this paper, we propose a weakly supervised strategy for training an object detector in this scenario. The main idea is to let the robot iteratively grow a training set by combining autonomously annotated examples, with others that are requested for human supervision. We evaluate our method on two experiments with data acquired from the iCub and R1 humanoid platforms, showing that it significantly reduces the number of human annotations required, without compromising performance. We also show the effectiveness of this approach when adapting the detector to a new setting.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131665883","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9035053
George Mesesan, Johannes Englsberger, Gianluca Garofalo, C. Ott, A. Albu-Schäffer
This paper presents a complete trajectory generation and control approach for achieving a robust dynamic walking gait for humanoid robots over compliant and uneven terrain. The work uses the concept of Divergent Component of Motion (DCM) for generating the center of mass (CoM) trajectory, and Cartesian polynomial trajectories for the feet. These reference trajectories are tracked by a passivity-based whole-body controller, which computes the joint torques for commanding our torque-controlled humanoid robot TORO. We provide the implementation details regarding the trajectory generation and control that help preventing discontinuities in the commanded joint torques, which facilitates precise trajectory tracking and robust locomotion. We present extensive experimental results of TORO walking over rough terrain, grass, and, to the best of our knowledge, the first report of a humanoid robot walking over a soft gym mattress.
{"title":"Dynamic Walking on Compliant and Uneven Terrain using DCM and Passivity-based Whole-body Control","authors":"George Mesesan, Johannes Englsberger, Gianluca Garofalo, C. Ott, A. Albu-Schäffer","doi":"10.1109/Humanoids43949.2019.9035053","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035053","url":null,"abstract":"This paper presents a complete trajectory generation and control approach for achieving a robust dynamic walking gait for humanoid robots over compliant and uneven terrain. The work uses the concept of Divergent Component of Motion (DCM) for generating the center of mass (CoM) trajectory, and Cartesian polynomial trajectories for the feet. These reference trajectories are tracked by a passivity-based whole-body controller, which computes the joint torques for commanding our torque-controlled humanoid robot TORO. We provide the implementation details regarding the trajectory generation and control that help preventing discontinuities in the commanded joint torques, which facilitates precise trajectory tracking and robust locomotion. We present extensive experimental results of TORO walking over rough terrain, grass, and, to the best of our knowledge, the first report of a humanoid robot walking over a soft gym mattress.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132975729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9035057
Hiroki Mineshita, T. Otani, K. Hashimoto, M. Sakaguchi, Y. Kawakami, Hun-ok Lim, A. Takanishi
When humanoid robots perform dynamic operations such as jumping and running, large outputs are required at each joint. It is known that humans save energy by using muscles and tendons effectively during dynamic motion. Therefore, we consider that energy saving and dynamic motion can be realized in robots by adding elements that replace such muscles and tendons. Based on this, we previously developed a robot with elasticity in the leg joints. However, its ankle joint mechanism did not have sufficient power to kick like a human while running. In addition, although the joint quasi-stiffness of the human leg changed according to the running speed, it could not handle high speeds nor simulate the required stiffness at low speeds. Therefore, we developed an ankle mechanism that is capable of kicking while jumping and running and adaptable to changes in running speed. By placing leaf springs in series, the mechanism achieved a joint stiffness of 250 to 350 Nm/rad, which is the ankle joint quasi-stiffness required for running at speeds of 2.0 to 5.0 m/s. By using a double motor, moreover, the mechanism succeeded at active kicking with a load torque of 110 Nm, equivalent to the value of active kicking while jumping.
{"title":"Robotic Ankle Mechanism Capable of Kicking While Jumping and Running and Adaptable to Change in Running Speed","authors":"Hiroki Mineshita, T. Otani, K. Hashimoto, M. Sakaguchi, Y. Kawakami, Hun-ok Lim, A. Takanishi","doi":"10.1109/Humanoids43949.2019.9035057","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035057","url":null,"abstract":"When humanoid robots perform dynamic operations such as jumping and running, large outputs are required at each joint. It is known that humans save energy by using muscles and tendons effectively during dynamic motion. Therefore, we consider that energy saving and dynamic motion can be realized in robots by adding elements that replace such muscles and tendons. Based on this, we previously developed a robot with elasticity in the leg joints. However, its ankle joint mechanism did not have sufficient power to kick like a human while running. In addition, although the joint quasi-stiffness of the human leg changed according to the running speed, it could not handle high speeds nor simulate the required stiffness at low speeds. Therefore, we developed an ankle mechanism that is capable of kicking while jumping and running and adaptable to changes in running speed. By placing leaf springs in series, the mechanism achieved a joint stiffness of 250 to 350 Nm/rad, which is the ankle joint quasi-stiffness required for running at speeds of 2.0 to 5.0 m/s. By using a double motor, moreover, the mechanism succeeded at active kicking with a load torque of 110 Nm, equivalent to the value of active kicking while jumping.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127886436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9035043
Yan Wang, K. Harada, Weiwei Wan
Complex and skillful motions in actual assembly process are challenging for the robot to generate with existing motion planning approaches, because some key poses during the human assembly can be too skillful for the robot to realize automatically. In order to deal with this problem, this paper develops a motion planning method using skillful motions from demonstration, which can be applied to complete robotic assembly process including complex and skillful motions. In order to demonstrate conveniently without redundant third-party devices, we attach augmented reality (AR) markers to the manipulated object to track and capture poses of the object during the human assembly process, which are employed as key poses to execute motion planning by the planner. Derivative of every key pose serves as criterion to determine the priority of use of key poses in order to accelerate the motion planning. The effectiveness of the presented method is verified through some numerical examples and actual robot experiments.
{"title":"Motion Planning through Demonstration to Deal with Complex Motions in Assembly Process","authors":"Yan Wang, K. Harada, Weiwei Wan","doi":"10.1109/Humanoids43949.2019.9035043","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035043","url":null,"abstract":"Complex and skillful motions in actual assembly process are challenging for the robot to generate with existing motion planning approaches, because some key poses during the human assembly can be too skillful for the robot to realize automatically. In order to deal with this problem, this paper develops a motion planning method using skillful motions from demonstration, which can be applied to complete robotic assembly process including complex and skillful motions. In order to demonstrate conveniently without redundant third-party devices, we attach augmented reality (AR) markers to the manipulated object to track and capture poses of the object during the human assembly process, which are employed as key poses to execute motion planning by the planner. Derivative of every key pose serves as criterion to determine the priority of use of key poses in order to accelerate the motion planning. The effectiveness of the presented method is verified through some numerical examples and actual robot experiments.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125809071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper an efficient 2D push recovery and balancing controller for the EVER3 humanoid robotic platform with differentially steered wheel-base is presented. Real world utility of humanoid robots requires a balance algorithm that minimizes motion due to small disturbances while simultaneously being quick enough to recover from large pushes. The proposed work uses a Model Predictive Control scheme along with a whole body controller to achieve desired performance. Various experimental runs were performed on the robot with the presented approach to do the performance analysis. The experimental results presented here prove the efficacy of the proposed control scheme.
{"title":"2D Push Recovery and Balancing of the EVER3 - a Humanoid Robot with Wheel-Base, using Model Predictive Control and Gain Scheduling","authors":"Nikhil Gupta, Jesper Smith, Brandon Shrewsbury, Bernt Børnich","doi":"10.1109/Humanoids43949.2019.9035044","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035044","url":null,"abstract":"In this paper an efficient 2D push recovery and balancing controller for the EVER3 humanoid robotic platform with differentially steered wheel-base is presented. Real world utility of humanoid robots requires a balance algorithm that minimizes motion due to small disturbances while simultaneously being quick enough to recover from large pushes. The proposed work uses a Model Predictive Control scheme along with a whole body controller to achieve desired performance. Various experimental runs were performed on the robot with the presented approach to do the performance analysis. The experimental results presented here prove the efficacy of the proposed control scheme.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125363838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-10-01DOI: 10.1109/Humanoids43949.2019.9035045
Takayuki Murooka, K. Okada, M. Inaba
In this paper, we propose methods for tightening screws of self-body using a driver, which enable self-repair and self-extension. There are two difficulties for tightening screws of self-body. First, the precise calculation of the screw pose is needed. When calculation with visual images using a camera, the observation error is so high. The merit of the robot is that the robot has CAD data of self-body. There we calculate the precise screw pose with self CAD data. Second, because of the small closed links when tightening screws of self-body, that the robot cannot move the driver for rotating around the screw sometimes happens because inverse kinematics cannot be solved. To solve this problem, we propose a method of tightening motion generation with regrasping a driver if inverse kinematics cannot be solved. With these methods, humanoid robots PR2 and HIRO realized self-repair and self-extension by tightening screws of self-body.
{"title":"Self-Repair and Self-Extension by Tightening Screws based on Precise Calculation of Screw Pose of Self-Body with CAD Data and Graph Search with Regrasping a Driver","authors":"Takayuki Murooka, K. Okada, M. Inaba","doi":"10.1109/Humanoids43949.2019.9035045","DOIUrl":"https://doi.org/10.1109/Humanoids43949.2019.9035045","url":null,"abstract":"In this paper, we propose methods for tightening screws of self-body using a driver, which enable self-repair and self-extension. There are two difficulties for tightening screws of self-body. First, the precise calculation of the screw pose is needed. When calculation with visual images using a camera, the observation error is so high. The merit of the robot is that the robot has CAD data of self-body. There we calculate the precise screw pose with self CAD data. Second, because of the small closed links when tightening screws of self-body, that the robot cannot move the driver for rotating around the screw sometimes happens because inverse kinematics cannot be solved. To solve this problem, we propose a method of tightening motion generation with regrasping a driver if inverse kinematics cannot be solved. With these methods, humanoid robots PR2 and HIRO realized self-repair and self-extension by tightening screws of self-body.","PeriodicalId":404758,"journal":{"name":"2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids)","volume":"190 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121433690","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：