Table of Contents
Fetching ...

Exceeding the Maximum Speed Limit of the Joint Angle for the Redundant Tendon-driven Structures of Musculoskeletal Humanoids

Kento Kawaharazuka, Yuya Koga, Kei Tsuzuki, Moritaka Onitsuka, Yuki Asano, Kei Okada, Koji Kawasaki, Masayuki Inaba

TL;DR

This work tackles the limitation that the maximum joint velocity in redundant tendon-driven musculoskeletal humanoids is constrained by the slowest muscle among the redundant set. It introduces two hardware-aware strategies: (i) antagonist inhibition, which zeros the current to large-$q$ antagonists to exploit backdrivability, and (ii) antagonist elongation, which pre-tensions selected antagonists via a mask-based quadratic program and a forward simulation to reduce restraint. Experimental validation on the Musashi robot shows that antagonist inhibition can substantially increase velocity in favorable backdrivability conditions (up to $4.4$ rad/s in simulation and about $3.7$ rad/s in reality), while antagonist elongation offers a backdrivability-agnostic alternative that still achieves notable speedups (up to or around $3.4$ rad/s). The study demonstrates that carefully leveraging redundancy and gravity/body inertia can surpass actuator-imposed limits, with clear trade-offs in required pre-tension and sensitivity to hardware backdrivability. These methods could enable higher-speed, biomimetic movements in complex musculoskeletal robots.

Abstract

The musculoskeletal humanoid has various biomimetic benefits, and the redundant muscle arrangement is one of its most important characteristics. This redundancy can achieve fail-safe redundant actuation and variable stiffness control. However, there is a problem that the maximum joint angle velocity is limited by the slowest muscle among the redundant muscles. In this study, we propose two methods that can exceed the limited maximum joint angle velocity, and verify the effectiveness with actual robot experiments.

Exceeding the Maximum Speed Limit of the Joint Angle for the Redundant Tendon-driven Structures of Musculoskeletal Humanoids

TL;DR

This work tackles the limitation that the maximum joint velocity in redundant tendon-driven musculoskeletal humanoids is constrained by the slowest muscle among the redundant set. It introduces two hardware-aware strategies: (i) antagonist inhibition, which zeros the current to large- antagonists to exploit backdrivability, and (ii) antagonist elongation, which pre-tensions selected antagonists via a mask-based quadratic program and a forward simulation to reduce restraint. Experimental validation on the Musashi robot shows that antagonist inhibition can substantially increase velocity in favorable backdrivability conditions (up to rad/s in simulation and about rad/s in reality), while antagonist elongation offers a backdrivability-agnostic alternative that still achieves notable speedups (up to or around rad/s). The study demonstrates that carefully leveraging redundancy and gravity/body inertia can surpass actuator-imposed limits, with clear trade-offs in required pre-tension and sensitivity to hardware backdrivability. These methods could enable higher-speed, biomimetic movements in complex musculoskeletal robots.

Abstract

The musculoskeletal humanoid has various biomimetic benefits, and the redundant muscle arrangement is one of its most important characteristics. This redundancy can achieve fail-safe redundant actuation and variable stiffness control. However, there is a problem that the maximum joint angle velocity is limited by the slowest muscle among the redundant muscles. In this study, we propose two methods that can exceed the limited maximum joint angle velocity, and verify the effectiveness with actual robot experiments.

Paper Structure

This paper contains 14 sections, 7 equations, 11 figures, 1 table.

Figures (11)

  • Figure 1: The basic musculoskeletal structure.
  • Figure 2: Characteristics of movements with high joint velocity. These movements use the effect of gravity and body inertia.
  • Figure 3: Muscle arrangement of the left arm of the musculoskeletal humanoid Musashi kawaharazuka2019musashi.
  • Figure 4: Characteristics of muscle length velocity transition.
  • Figure 5: Experimental motion of simulation and actual robot.
  • ...and 6 more figures