Musculoskeletal AutoEncoder: A Unified Online Acquisition Method of Intersensory Networks for State Estimation, Control, and Simulation of Musculoskeletal Humanoids

While the musculoskeletal humanoid has various biomimetic benefits, the modeling of its complex structure is difficult, and many learning-based systems have been developed so far. There are various methods, such as control methods using acquired relationships between joints and muscles represented by a data table or neural network, and state estimation methods using Extended Kalman Filter or table search. In this letter, we construct a Musculoskeletal AutoEncoder representing the relationship among joint angles, muscle tensions, and muscle lengths, and propose a unified method of state estimation, control, and simulation of musculoskeletal humanoids using it. By updating the Musculoskeletal AutoEncoder online using the actual robot sensor information, we can continuously conduct more accurate state estimation, control, and simulation than before the online learning. We conducted several experiments using the musculoskeletal humanoid Musashi, and verified the effectiveness of this study.

[1]  Hiroaki Kobayashi,et al.  On Tendon-Driven Robotic Mechanisms with Redundant Tendons , 1998, Int. J. Robotics Res..

[2]  Masayuki Inaba,et al.  Online Learning of Joint-Muscle Mapping Using Vision in Tendon-Driven Musculoskeletal Humanoids , 2018, IEEE Robotics and Automation Letters.

[3]  Masayuki Inaba,et al.  Learning nonlinear muscle-joint state mapping toward geometric model-free tendon driven musculoskeletal robots , 2015, 2015 IEEE-RAS 15th International Conference on Humanoid Robots (Humanoids).

[4]  Alois Knoll,et al.  Toward Anthropomimetic Robotics: Development, Simulation, and Control of a Musculoskeletal Torso , 2013, Artificial Life.

[5]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[6]  Yasuo Kuniyoshi,et al.  Athlete Robot with applied human muscle activation patterns for bipedal running , 2010, 2010 10th IEEE-RAS International Conference on Humanoid Robots.

[7]  Masayuki Inaba,et al.  Design Approach of Biologically-Inspired Musculoskeletal Humanoids , 2013 .

[8]  Masayuki Inaba,et al.  Motion control based on modification of the Jacobian map between the muscle space and work space with musculoskeletal humanoid , 2012, 2012 12th IEEE-RAS International Conference on Humanoid Robots (Humanoids 2012).

[9]  Ying Tan,et al.  CASPR: A comprehensive cable-robot analysis and simulation platform for the research of cable-driven parallel robots , 2016, 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[10]  Masayuki Inaba,et al.  Long-Time Self-Body Image Acquisition and Its Application to the Control of Musculoskeletal Structures , 2019, IEEE Robotics and Automation Letters.

[11]  Alois Knoll,et al.  CALIPER: A universal robot simulation framework for tendon-driven robots , 2011, 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[12]  Masayuki Inaba,et al.  Body Information Acquisition System of Redundant Musculo-Skeletal Humanoid , 2004, ISER.

[13]  Alois Knoll,et al.  Anthrob - A printed anthropomimetic robot , 2013, 2013 13th IEEE-RAS International Conference on Humanoid Robots (Humanoids).

[14]  Masayuki Inaba,et al.  Online Self-body Image Acquisition Considering Changes in Muscle Routes Caused by Softness of Body Tissue for Tendon-driven Musculoskeletal Humanoids , 2018, 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[15]  Masayuki Inaba,et al.  A joint-space controller based on redundant muscle tension for multiple DOF joints in musculoskeletal humanoids , 2016, 2016 IEEE-RAS 16th International Conference on Humanoid Robots (Humanoids).

[16]  A Diamond,et al.  Reaching control of a full-torso, modelled musculoskeletal robot using muscle synergies emergent under reinforcement learning , 2014, Bioinspiration & biomimetics.

[17]  Masayuki Inaba,et al.  Human mimetic musculoskeletal humanoid Kengoro toward real world physically interactive actions , 2016, 2016 IEEE-RAS 16th International Conference on Humanoid Robots (Humanoids).

[18]  Alois Knoll,et al.  CARDSFlow: An End-to-End Open-Source Physics Environment for the Design, Simulation and Control of Musculoskeletal Robots , 2018, 2018 IEEE-RAS 18th International Conference on Humanoid Robots (Humanoids).

[19]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[20]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[21]  Kei Okada,et al.  Component Modularized Design of Musculoskeletal Humanoid Platform Musashi to Investigate Learning Control Systems , 2019, 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[22]  Masayuki Inaba,et al.  Whole body adapting behavior with muscle level stiffness control of tendon-driven multijoint robot , 2011, 2011 IEEE International Conference on Robotics and Biomimetics.

[23]  Masayuki Inaba,et al.  Joint proprioception acquisition strategy based on joints-muscles topological maps for musculoskeletal humanoids , 2010, 2010 IEEE International Conference on Robotics and Automation.