RoboMIND 2.0: A Multimodal, Bimanual Mobile Manipulation Dataset for Generalizable Embodied Intelligence

TL;DR

RoboMIND 2.0 introduces a large-scale, multimodal dataset of bimanual, mobile manipulation tasks in real-world and simulated environments, enabling advances in generalizable embodied robotic intelligence.

Contribution

The paper provides a comprehensive dataset with diverse manipulation trajectories and a hierarchical system combining semantic planning and low-level control for improved robotic learning.

Findings

Over 310K real-world manipulation trajectories collected

High-fidelity digital twins facilitate sim-to-real transfer

Hierarchical MIND-2 system improves task decomposition and execution

Abstract

While data-driven imitation learning has revolutionized robotic manipulation, current approaches remain constrained by the scarcity of large-scale, diverse real-world demonstrations. Consequently, the ability of existing models to generalize across long-horizon bimanual tasks and mobile manipulation in unstructured environments remains limited. To bridge this gap, we present RoboMIND 2.0, a comprehensive real-world dataset comprising over 310K dual-arm manipulation trajectories collected across six distinct robot embodiments and 739 complex tasks. Crucially, to support research in contact-rich and spatially extended tasks, the dataset incorporates 12K tactile-enhanced episodes and 20K mobile manipulation trajectories. Complementing this physical data, we construct high-fidelity digital twins of our real-world environments, releasing an additional 20K-trajectory simulated dataset to facilitate robust sim-to-real transfer. To fully exploit the potential of RoboMIND 2.0, we propose MIND-2 system, a hierarchical dual-system frame-work optimized via offline reinforcement learning. MIND-2 integrates a high-level semantic planner (MIND-2-VLM) to decompose abstract natural language instructions into grounded subgoals, coupled with a low-level Vision-Language-Action executor (MIND-2-VLA), which generates precise, proprioception-aware motor actions.