Exploiting Interaction Dynamics for Learning Collaborative Robot Behaviors∗

The new generation of smarter and safer robots aimed at assisting humans in industries and homes demands to innovate the ways in which these machines are designed and controlled. In this context, one of the biggest challenges is to empower these collaborative robots with a wide range of learning and adaptation capabilities so that they can easily assist humans in a vast variety of scenarios, ranging from assembly lines to health-care facilities. In this paper we propose to teach a collaborative robot reactive and proactive behaviors that exploit the interaction dynamics between the robot and the user. We call the proposed approach Adaptive Duration Hidden-Semi Markov Model (ADHSMM) that enables the robot to both react to the user actions and lead the task when needed. ADHSMM is used to retrieve a sequence of states governing a trajectory optimization technique that provides the reference and gain matrices to the robot controller. The proposed framework is tested in handover and transportation tasks using a 7 DOF backdrivable manipulator.

[1]  Henk Nijmeijer,et al.  Robot Programming by Demonstration , 2010, SIMPAR.

[2]  Peter Secretan Learning , 1965, Mental Health.

[3]  Sandra Hirche,et al.  An experience-driven robotic assistant acquiring human knowledge to improve haptic cooperation , 2011, 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[4]  Lawrence R. Rabiner,et al.  A tutorial on hidden Markov models and selected applications in speech recognition , 1989, Proc. IEEE.

[5]  Heiga Zen,et al.  Hidden Semi-Markov Model Based Speech Synthesis System , 2006 .

[6]  Keiichi Tokuda,et al.  An algorithm for speech parameter generation from continuous mixture HMMs with dynamic features , 1995, EUROSPEECH.

[7]  Heiga Zen,et al.  Reformulating the HMM as a trajectory model by imposing explicit relationships between static and dynamic feature vector sequences , 2007, Comput. Speech Lang..

[8]  Keng Peng Tee,et al.  Continuous Role Adaptation for Human–Robot Shared Control , 2015, IEEE Transactions on Robotics.

[9]  Darwin G. Caldwell,et al.  Learning bimanual end-effector poses from demonstrations using task-parameterized dynamical systems , 2015, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[10]  Mary P. Harper,et al.  On the complexity of explicit duration HMM's , 1995, IEEE Trans. Speech Audio Process..

[11]  Minija Tamosiunaite,et al.  Interaction learning for dynamic movement primitives used in cooperative robotic tasks , 2013, Robotics Auton. Syst..

[12]  Paul Evrard,et al.  Teaching physical collaborative tasks: object-lifting case study with a humanoid , 2009, 2009 9th IEEE-RAS International Conference on Humanoid Robots.

[13]  Satoshi Nakamura,et al.  Learning, Generation and Recognition of Motions by Reference-Point-Dependent Probabilistic Models , 2011, Adv. Robotics.

[14]  Shunzheng Yu,et al.  Hidden semi-Markov models , 2010, Artif. Intell..

[15]  Jan Peters,et al.  Probabilistic Movement Primitives , 2013, NIPS.

[16]  Darwin G. Caldwell,et al.  Learning optimal controllers in human-robot cooperative transportation tasks with position and force constraints , 2015, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[17]  Stefanos Nikolaidis,et al.  Optimization of Temporal Dynamics for Adaptive Human-Robot Interaction in Assembly Manufacturing , 2012, Robotics: Science and Systems.

[18]  Carme Torras,et al.  Learning Physical Collaborative Robot Behaviors From Human Demonstrations , 2016, IEEE Transactions on Robotics.

[19]  Sadaoki Furui,et al.  Speaker-independent isolated word recognition using dynamic features of speech spectrum , 1986, IEEE Trans. Acoust. Speech Signal Process..

[20]  Shunzheng Yu,et al.  Practical implementation of an efficient forward-backward algorithm for an explicit-duration hidden Markov model , 2006, IEEE Transactions on Signal Processing.

[21]  Jan Peters,et al.  A Probabilistic Framework for Semi-autonomous Robots Based on Interaction Primitives with Phase Estimation , 2015, ISRR.

[22]  Philip C. Woodland,et al.  Maximum likelihood linear regression for speaker adaptation of continuous density hidden Markov models , 1995, Comput. Speech Lang..

[23]  Takashi Nose,et al.  Style estimation of speech based on multiple regression hidden semi-Markov model , 2007, INTERSPEECH.

[24]  Jan Peters,et al.  Learning responsive robot behavior by imitation , 2013, 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems.