roto 2.0: The Robot Tactile Olympiad

TL;DR

Roto 2.0 introduces a standardized, tactile-based RL benchmark for robotic manipulation across multiple morphologies, emphasizing end-to-end blind control without state info, and demonstrates significant performance improvements.

Contribution

The paper presents Roto 2.0, a new benchmark for tactile RL that enables standardized evaluation across diverse robots and promotes research on blind manipulation tasks.

Findings

Blind agents achieved 13 Baoding ball rotations in 10 seconds.

Roto 2.0 outperforms existing benchmarks in manipulation speed.

Open-sourced environments facilitate research and reduce tuning barriers.

Abstract

Tactile-based reinforcement learning (RL) is currently hindered by fragmented research and a focus on over-saturated orientation tasks. We introduce v2 of the Robot Tactile Olympiad (\texttt{roto 2.0}), a GPU-parallelised benchmark designed to standardise tactile-based RL across four distinct robotic morphologies (16-DOF to 24-DOF). Unlike prior benchmarks, roto focuses on end-to-end "blind" manipulation, utilising only proprioception and tactile sensing without state information or distillation. We demonstrate a significant performance leap, with our blind agents achieving 13 Baoding ball rotations in 10 seconds, an order of magnitude faster than current state-of-the-art speeds. By open-sourcing our environments and robustly tuned baselines, we reduce the barrier to entry and enable researchers to prioritise fundamental algorithmic challenges over tedious RL tuning. Website: https://elle-miller.github.io/roto/