Whole-body Multi-contact Motion Control for Humanoid Robots Based on Distributed Tactile Sensors

TL;DR

This work tackles robust whole-body multi-contact motion for humanoid robots in confined spaces by integrating deformable distributed tactile sensors on limb surfaces to enable real-time contact sensing and feedback. The authors extend a prior multi-contact framework with tactile feedback, formulating centroidal control via a 6D resultant wrench $m{ar{w}}$ and an MPC-based planner, complemented by stabilization, estimation, and limb-wrench distribution components. They implement damping control and online contact-region updates using tactile data, solving a QP for wrench distribution and applying axis-angle based pose updates to maintain balance during intermediate-area contacts. Validation includes dynamics simulations and real-world demonstrations on a life-sized robot (RHP Kaleido) with e-skin patches on forearms and thighs, achieving stable transitions such as stepping with a forearm contact and sitting balance with thigh contacts. The results show that tactile feedback extends stability and robustness against environmental disturbances, enabling more versatile, contact-rich humanoid behaviors in constrained environments; future work envisions autonomous contact planning and learning-based enhancement to exploit richer tactile data.

Abstract

To enable humanoid robots to work robustly in confined environments, multi-contact motion that makes contacts not only at extremities, such as hands and feet, but also at intermediate areas of the limbs, such as knees and elbows, is essential. We develop a method to realize such whole-body multi-contact motion involving contacts at intermediate areas by a humanoid robot. Deformable sheet-shaped distributed tactile sensors are mounted on the surface of the robot's limbs to measure the contact force without significantly changing the robot body shape. The multi-contact motion controller developed earlier, which is dedicated to contact at extremities, is extended to handle contact at intermediate areas, and the robot motion is stabilized by feedback control using not only force/torque sensors but also distributed tactile sensors. Through verification on dynamics simulations, we show that the developed tactile feedback improves the stability of whole-body multi-contact motion against disturbances and environmental errors. Furthermore, the life-sized humanoid RHP Kaleido demonstrates whole-body multi-contact motions, such as stepping forward while supporting the body with forearm contact and balancing in a sitting posture with thigh contacts.

Figures (13)

• Figure 1: Control system for whole-body multi-contact motion in a humanoid robot. The control system consists of centroidal motion control and limb motion control. Newly added extensions for tactile sensing compared to the previously developed control system Motion6DoF:Murooka:RAL2022 are indicated in red in the figure. For definitions of the symbols in the figure, see Sections \ref{['sec:centroidal']} and \ref{['sec:limb']}.
• Figure 2: Contact wrench representation.
• Figure 3: Conversion of distributed tactile intensity to contact wrench.
• Figure 4: RHP Kaleido with distributed tactile sensors mounted on one forearm and both thighs.
• Figure 5: Simulation of whole-body multi-contact motions. The top row shows snapshots of the simulation, and the bottom row shows the distributed tactile sensors (in each cell, green represents no contact and red represents contact). (A) Assuming that there is an obstacle on the left side, the robot walks with the right elbow on the wall, leaning the body to the right. (B) The robot stands with the right foot on the ground and the left knee on the block, balancing itself. (C) The robot sits on the block with both thighs, keeping its balance.