Probabilistic decomposition of sequential force interaction tasks into Movement Primitives

Learning sequential force interaction tasks from kinesthetic demonstrations is a promising approach to transfer human manipulation abilities to a robot. In this paper we propose a novel concept to decompose such demonstrations into a set of Movement Primitives (MPs). The decomposition is based on a probability distribution we call Directional Normal Distribution (DND). To capture the sequential properties of the manipulation task, we model the demonstrations with a Hidden Markov Model (HMM). Here, we employ mixtures of DNDs as the HMM's output emissions. The combination of HMMs and mixtures of DNDs allows to infer the MP's composition, i.e., its coordinate frames, control variables and target coordinates from the demonstration data. In addition, it permits to determine an appropriate number of MPs that explains the demonstrations best. We evaluate the approach on kinesthetic demonstrations of a light bulb unscrewing task. Decomposing the task leads to intuitive and meaningful MPs that reflect the natural structure of the task.

[1]  Sven Behnke,et al.  Incremental action recognition and generalizing motion generation based on goal-directed features , 2012, 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[2]  Dana Kulic,et al.  Incremental learning of full body motion primitives and their sequencing through human motion observation , 2012, Int. J. Robotics Res..

[3]  Gregory D. Hager,et al.  Transition state clustering: Unsupervised surgical trajectory segmentation for robot learning , 2017, ISRR.

[4]  Scott Kuindersma,et al.  Robot learning from demonstration by constructing skill trees , 2012, Int. J. Robotics Res..

[5]  Jonathan P. How,et al.  Bayesian Nonparametric Reward Learning From Demonstration , 2015, IEEE Transactions on Robotics.

[6]  Pravesh Ranchod,et al.  Nonparametric Bayesian reward segmentation for skill discovery using inverse reinforcement learning , 2015, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[7]  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).

[8]  Jan Peters,et al.  Probabilistic segmentation applied to an assembly task , 2015, 2015 IEEE-RAS 15th International Conference on Humanoid Robots (Humanoids).

[9]  Andrew McCallum,et al.  Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data , 2001, ICML.

[10]  N. Higham MATRIX NEARNESS PROBLEMS AND APPLICATIONS , 1989 .

[11]  Jimmy A. Jørgensen,et al.  Adaptation of manipulation skills in physical contact with the environment to reference force profiles , 2015, Auton. Robots.

[12]  Aude Billard,et al.  Task Parameterization Using Continuous Constraints Extracted From Human Demonstrations , 2015, IEEE Transactions on Robotics.

[13]  Jan Peters,et al.  Learning movement primitive attractor goals and sequential skills from kinesthetic demonstrations , 2015, Robotics Auton. Syst..

[14]  Alberto Montebelli,et al.  Simultaneous kinesthetic teaching of positional and force requirements for sequential in-contact tasks , 2015, 2015 IEEE-RAS 15th International Conference on Humanoid Robots (Humanoids).

[15]  Miles C. Bowman,et al.  Control strategies in object manipulation tasks , 2006, Current Opinion in Neurobiology.

[16]  Lars Omlor,et al.  Emulating human observers with bayesian binning: Segmentation of action streams , 2011, TAP.

[17]  Scott Niekum,et al.  Learning grounded finite-state representations from unstructured demonstrations , 2015, Int. J. Robotics Res..

[18]  Stefan Schaal,et al.  Towards Associative Skill Memories , 2012, 2012 12th IEEE-RAS International Conference on Humanoid Robots (Humanoids 2012).

[19]  Jan Peters,et al.  Probabilistic progress prediction and sequencing of concurrent movement primitives , 2015, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[20]  Jochen J. Steil,et al.  Learning movement primitives for force interaction tasks , 2015, 2015 IEEE International Conference on Robotics and Automation (ICRA).

[21]  Tobias Luksch,et al.  Adaptive movement sequences and predictive decisions based on hierarchical dynamical systems , 2012, 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[22]  Manuel Lopes,et al.  Temporal segmentation of pair-wise interaction phases in sequential manipulation demonstrations , 2015, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).