Learning Force‐Relevant Skills from Human Demonstration

Complexity - Tập 2019 Số 1 - 2019
Xiao Gao1, Jie Ling1, Xiaohui Xiao2,1, Miao Li1,3
1School of Power and Mechanical Engineering, Wuhan University, China
2National Key Laboratory of Human Factors Engineering, China Astronauts Research and Trainning Center, Beijing, China
3The Institute of technological Sciences, Wuhan University, China

Tóm tắt

Many human manipulation skills are force relevant, such as opening a bottle cap and assembling furniture. However, it is still a difficult task to endow a robot with these skills, which largely is due to the complexity of the representation and planning of these skills. This paper presents a learning‐based approach of transferring force‐relevant skills from human demonstration to a robot. First, the force‐relevant skill is encapsulated as a statistical model where the key parameters are learned from the demonstrated data (motion, force). Second, based on the learned skill model, a task planner is devised which specifies the motion and/or the force profile for a given manipulation task. Finally, the learned skill model is further integrated with an adaptive controller that offers task‐consistent force adaptation during online executions. The effectiveness of the proposed approach is validated with two experiments, i.e., an object polishing task and a peg‐in‐hole assembly.

Từ khóa


Tài liệu tham khảo

10.1007/978-3-319-32552-1

10.1115/1.3149634

10.1109/TSMC.1981.4308708

10.1115/1.3140713

GuS. HollyE. LillicrapT. andLevineS. Deep reinforcement learning for robotic manipulation with asynchronous off-policy updates Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) 2017 IEEE 3389–3396 2-s2.0-85027967014.

Levine S., 2016, International Symposium on Experimental Robotics, 173

10.1038/nature14236

KalakrishnanM. RighettiL. PastorP. andSchaalS. Learning force control policies for compliant manipulation Proceedings of the IEEE International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics (IROS) 2011 IEEE 4639–4644 2-s2.0-84455188451.

LeeA. X. LuH. GuptaA. LevineS. andAbbeelP. Learning force-based manipulation of deformable objects from multiple demonstrations Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) 2015 IEEE 177–184 2-s2.0-84938252581.

10.1109/TMECH.2015.2510165

10.1109/TRO.2011.2159412

10.1007/978-3-540-30301-5_60

10.1162/NECO_a_00393

LiM. YinH. TaharaK. andBillardA. Learning object-level impedance control for robust grasping and dexterous manipulation Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) 2014 IEEE 6784–6791 https://doi.org/10.1109/icra.2014.6907861 2-s2.0-84929208215.

LiM. BekirogluY. KragicD. andBillardA. Learning of grasp adaptation through experience and tactile sensing Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2014 IEEE 3339–3346 2-s2.0-84911463955.

KronanderK.andBillardA. Online learning of varying stiffness through physical human-robot interaction Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) 2012 IEEE 1842–1849 2-s2.0-84864429731.

10.1007/s10514-017-9643-z

10.1109/TASE.2017.2743000

10.1109/TMECH.2018.2817589

Siciliano B., 2012, Robot Force Control

LinY. RenS. ClevengerM. andSunY. Learning grasping force from demonstration Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) IEEE 1526–1531 2-s2.0-84864483234.

10.1163/016918611X558261

10.1007/s11370-012-0128-9

10.1007/s10514-015-9501-9

VogtD. StepputtisS. GrehlS. JungB. andBen AmorH. A system for learning continuous human-robot interactions from human-human demonstrations Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) 2017 IEEE 2882–2889 2-s2.0-85028018545.

10.1177/0278364916648389

Kronander K., 2014, Workshop on Human-Robot Interaction for Industrial Manufacturing, Robotics, Science and Systems

10.1109/TCST.2016.2514609

10.7551/mitpress/3676.001.0001

Calinon S., 2009, Robot programming by demonstration

Andersen T. T., 2015, Optimizing the universal robots ros driver

Rasmussen C. E., 2010, Gaussian processes for machine learning (GPML) toolbox, Journal of Machine Learning Research, 11, 3011