Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Robust Anthropomorphic Robotic Manipulation through Biomimetic Distributed Compliance

Kai Junge, Josie Hughes

TL;DR

The paper addresses the challenge of achieving robust open-loop manipulation in anthropomorphic hands. It introduces the ADAPT Hand, a biomimetic platform with distributed compliance across skin, fingers, and a compliant wrist, enabling self-organizing and human-like grasp behaviors under minimal planning. Through extensive experiments, the authors show that skin and finger compliance improve contact stability and robustness, while the wrist enables a spectrum of emergent grasps across 24 objects with a 93% success rate, and over 800–845 grasps with high reliability. The work demonstrates that physical intelligence distributed across the hand can yield near-optimal geometric performance, substantial robustness in open-loop tasks, and meaningful alignment with human grasp strategies, suggesting a practical pathway toward robust, low-control robotic manipulation.

Abstract

The impressive capabilities of humans to robustly perform manipulation relies on compliant interactions, enabled through the structure and materials spatially distributed in our hands. We propose by mimicking this distributed compliance in an anthropomorphic robotic hand, the open-loop manipulation robustness increases and observe the emergence of human-like behaviours. To achieve this, we introduce the ADAPT Hand equipped with tunable compliance throughout the skin, fingers, and the wrist. Through extensive automated pick-and-place tests, we show the grasping robustness closely mirrors an estimated geometric theoretical limit, while `stress-testing' the robot hand to perform 800+ grasps. Finally, 24 items with largely varying geometries are grasped in a constrained environment with a success rate of 93%. We demonstrate the hand-object self-organization behavior underlines this extreme robustness, where the hand automatically exhibits different grasp types depending on object geometries. Furthermore, the robot grasp type mimics a natural human grasp with a direct similarity of 68%.

Robust Anthropomorphic Robotic Manipulation through Biomimetic Distributed Compliance

TL;DR

The paper addresses the challenge of achieving robust open-loop manipulation in anthropomorphic hands. It introduces the ADAPT Hand, a biomimetic platform with distributed compliance across skin, fingers, and a compliant wrist, enabling self-organizing and human-like grasp behaviors under minimal planning. Through extensive experiments, the authors show that skin and finger compliance improve contact stability and robustness, while the wrist enables a spectrum of emergent grasps across 24 objects with a 93% success rate, and over 800–845 grasps with high reliability. The work demonstrates that physical intelligence distributed across the hand can yield near-optimal geometric performance, substantial robustness in open-loop tasks, and meaningful alignment with human grasp strategies, suggesting a practical pathway toward robust, low-control robotic manipulation.

Abstract

The impressive capabilities of humans to robustly perform manipulation relies on compliant interactions, enabled through the structure and materials spatially distributed in our hands. We propose by mimicking this distributed compliance in an anthropomorphic robotic hand, the open-loop manipulation robustness increases and observe the emergence of human-like behaviours. To achieve this, we introduce the ADAPT Hand equipped with tunable compliance throughout the skin, fingers, and the wrist. Through extensive automated pick-and-place tests, we show the grasping robustness closely mirrors an estimated geometric theoretical limit, while `stress-testing' the robot hand to perform 800+ grasps. Finally, 24 items with largely varying geometries are grasped in a constrained environment with a success rate of 93%. We demonstrate the hand-object self-organization behavior underlines this extreme robustness, where the hand automatically exhibits different grasp types depending on object geometries. Furthermore, the robot grasp type mimics a natural human grasp with a direct similarity of 68%.
Paper Structure (29 sections, 15 figures)

This paper contains 29 sections, 15 figures.

Table of Contents

  1. Introduction
  2. Results
  3. ADAPT Hand
  4. Skin: Contact stability
  5. Skin compliance on task performance
  6. Finger: Pose adaptation
  7. Human motion comparison
  8. Measuring robustness through finger compliance
  9. In-hand cube re-orientation
  10. Systematic pick-and-place robustness assessment of the ADAPT Hand
  11. Measuring open-loop robustness
  12. Extended period of operation
  13. Self-organizing grasps
  14. Emergence of discrete grasp types
  15. Discussion
  16. ...and 14 more sections

Figures (15)

  • Figure 1: An anthropomorphic robot hand designed with biomimetic distribution of compliance leads to a emergence of robust and self-organizing behavior. A) The ADAPT Hand is compliant in its skin, finger, and wrist, which operate in the 0.1cm, 1cm, and 10cm displacement ranges. B) Through identical open-loop waypoints (four waypoints describing the full hand-arm motion for this grasp), the hand can grasp objects (a large apple vs three small items) successfully. The distributed compliance allows for the hand to be robust upon unknown environmental interactions, and a self-organization to take place between the hand and objects, resulting in different grasps.
  • Figure 2: A) The ADAPT hand. B) Finger design with two independent actuation and a series spring on the MCP flexor. C) The force-displacement measurement curve for the skin, finger, and wrist for the robot and a human. For the skin and finger, the rigid configuration of the ADAPT Hand is also shown for comparison.
  • Figure 3: A) Results from the finger sliding experiment. Schematic (top-left), box plot and raw values for the shear force of rigid and soft skins measured at the midpoint of the interaction (top-right), example time series of the shear force of rigid and soft skins for the two sliding motions (bottom). B) Results from the knob turning experiment. Schematic (top) and turn angles for the soft and rigid skins as the diameter $d$ and resisting torque $\tau$ is varied (bottom). C) Results from the finger gaiting experiment. Schematic (top) and completed gaits for soft and rigid skins as the held block width $w$ is varied (bottom).
  • Figure 4: A) Key frames from two finger sliding motions(left). By overdriving the MCP joint, pseudo force control is possible with good repeatability (right). B) Trajectories of relative changes in the three joint angles for the sliding motions executed by a human, soft robot finger, and a rigid robot finger. C) Schematic for the finger sliding experiment (top). Maximum forces recorded as the sliding plate is displaced in the $z$ and $\theta$ directions for the two motions/soft and rigid fingers (bottom). D) Schematic for the knob turning experiment (top). Tendon waypoints required for the motion(mid) and turn angles as the environment is varied (bottom) for the soft and rigid fingers E) Schematic for the finger gaiting experiment (top). Completed gaits (mid) and average holding forces (bottom) for soft and rigid fingers as the held block width $w$ is varied. F) Success rates for the three cubes rotated in-hand (top). Pictorial sequence of the in-hand manipulation sequence (bottom).
  • Figure 5: A) Robotic setup to conduct large quantities of automatic pick-and-place experiments while controlling the displacement of the object. B) Measured limits on object displacement (orange axis), estimated geometric limits based on object size and hand closure motion. C) Success and failed grasps throughout the two experiments: robustness assessment and uninterrupted pick-and-place totalling 845 grasps.
  • ...and 10 more figures