Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Walking, Rolling, and Beyond: First-Principles and RL Locomotion on a TARS-Inspired Robot

Aditya Sripada, Abhishek Warrier

TL;DR

The work addresses expanding locomotion beyond anthropomorphic forms by leveraging fiction-inspired morphology. It couples analytic reduced-order models for bipedal walking and rolling with a PPO-based RL pipeline, validated on hardware as well as in simulation. Key contributions include the TARS3D platform, closed-form gait syntheses for two modes, and prior-guided RL that recovers known gaits while uncovering novel low-speed behaviors, illustrating a design-space expansion for multimodal locomotion. The findings demonstrate that a seed-and-learn workflow can effectively explore and exploit non-traditional morphologies to realize efficient, multimodal locomotion with practical implications for robotics design and control.

Abstract

Robotic locomotion research typically draws from biologically inspired leg designs, yet many human-engineered settings can benefit from non-anthropomorphic forms. TARS3D translates the block-shaped 'TARS' robot from Interstellar into a 0.25 m, 0.99 kg research platform with seven actuated degrees of freedom. The film shows two primary gaits: a bipedal-like walk and a high-speed rolling mode. For TARS3D, we build reduced-order models for each, derive closed-form limit-cycle conditions, and validate the predictions on hardware. Experiments confirm that the robot respects its +/-150 degree hip limits, alternates left-right contacts without interference, and maintains an eight-step hybrid limit cycle in rolling mode. Because each telescopic leg provides four contact corners, the rolling gait is modeled as an eight-spoke double rimless wheel. The robot's telescopic leg redundancy implies a far richer gait repertoire than the two limit cycles treated analytically. So, we used deep reinforcement learning (DRL) in simulation to search the unexplored space. We observed that the learned policy can recover the analytic gaits under the right priors and discover novel behaviors as well. Our findings show that TARS3D's fiction-inspired bio-transcending morphology can realize multiple previously unexplored locomotion modes and that further learning-driven search is likely to reveal more. This combination of analytic synthesis and reinforcement learning opens a promising pathway for multimodal robotics.

Walking, Rolling, and Beyond: First-Principles and RL Locomotion on a TARS-Inspired Robot

TL;DR

The work addresses expanding locomotion beyond anthropomorphic forms by leveraging fiction-inspired morphology. It couples analytic reduced-order models for bipedal walking and rolling with a PPO-based RL pipeline, validated on hardware as well as in simulation. Key contributions include the TARS3D platform, closed-form gait syntheses for two modes, and prior-guided RL that recovers known gaits while uncovering novel low-speed behaviors, illustrating a design-space expansion for multimodal locomotion. The findings demonstrate that a seed-and-learn workflow can effectively explore and exploit non-traditional morphologies to realize efficient, multimodal locomotion with practical implications for robotics design and control.

Abstract

Robotic locomotion research typically draws from biologically inspired leg designs, yet many human-engineered settings can benefit from non-anthropomorphic forms. TARS3D translates the block-shaped 'TARS' robot from Interstellar into a 0.25 m, 0.99 kg research platform with seven actuated degrees of freedom. The film shows two primary gaits: a bipedal-like walk and a high-speed rolling mode. For TARS3D, we build reduced-order models for each, derive closed-form limit-cycle conditions, and validate the predictions on hardware. Experiments confirm that the robot respects its +/-150 degree hip limits, alternates left-right contacts without interference, and maintains an eight-step hybrid limit cycle in rolling mode. Because each telescopic leg provides four contact corners, the rolling gait is modeled as an eight-spoke double rimless wheel. The robot's telescopic leg redundancy implies a far richer gait repertoire than the two limit cycles treated analytically. So, we used deep reinforcement learning (DRL) in simulation to search the unexplored space. We observed that the learned policy can recover the analytic gaits under the right priors and discover novel behaviors as well. Our findings show that TARS3D's fiction-inspired bio-transcending morphology can realize multiple previously unexplored locomotion modes and that further learning-driven search is likely to reveal more. This combination of analytic synthesis and reinforcement learning opens a promising pathway for multimodal robotics.

Paper Structure

This paper contains 20 sections, 9 equations, 10 figures, 1 table.

Table of Contents

  1. Introduction
  2. Background and Motivation
  3. Science Fiction as a Driver of Robotics Innovation
  4. Exploring the Design Space Beyond Anthropocentrism
  5. Contributions
  6. System Design
  7. Analytical Locomotion Modes: Bipedal Walking and Rolling
  8. Bipedal Gait Model
  9. Rolling Gait Model (Rimless-Wheel Formulation)
  10. Exploring Alternative Gaits via Reinforcement Learning
  11. Setup and Training
  12. Learning Results
  13. Terrain Adaptation
  14. Discussion
  15. Multiple gaits from a non-anthropomorphic morphology
  16. ...and 5 more sections

Figures (10)

  • Figure 1: Primary locomotion modes emulated by TARS: (left) bipedal-like gait, (right) rolling gait.
  • Figure 2: TARS3D mechanical design: (a) overall CAD model of the 4-leg (${L_1, ..., L_4})$ assembly, (b) leg section showing the rotary hip joint ($R_{n}$) and prismatic leg extension ($T_{n}$)
  • Figure 3: Phases of bipedal gait of TARS3D
  • Figure 4: Four phases of TARS3D bipedal gait
  • Figure 5: Double rimless-wheel morphology of TARS3D
  • ...and 5 more figures