BiCoord: A Bimanual Manipulation Benchmark towards Long-Horizon Spatial-Temporal Coordination

2026-04-07

AI Summary

Existing benchmarks for dual-arm manipulation primarily focus on short-duration tasks with loose coordination, failing to capture the strong spatiotemporal coupling inherent in real-world collaborative scenarios. To address this gap, this work introduces the first benchmark specifically designed for long-horizon, tightly coupled dual-arm manipulation, featuring diverse tasks that require sustained inter-arm dependency and dynamic role switching. We further propose a multi-scale coordination evaluation framework encompassing temporal, spatial, and joint dimensions. Through systematic evaluation in simulation of state-of-the-art policies—including DP, RDT, Pi0, and OpenVLA-OFT—we demonstrate their significant limitations in high-coupling, long-horizon settings, thereby revealing core challenges in achieving robust spatiotemporal coordination and establishing a new foundation for future research.

Abstract

Bimanual manipulation, i.e., the coordinated use of two robotic arms to complete tasks, is essential for achieving human-level dexterity in robotics. Recent simulation benchmarks, e.g., RoboTwin and RLBench2, have advanced data-driven learning for bimanual manipulation. However, existing tasks are short-horizon and only loosely coordinated, failing to capture the spatial-temporal coupling inherent in real-world bimanual behaviors. To address this gap, we introduce BiCoord, a benchmark for long-horizon and tightly coordinated bimanual manipulation. Specifically, BiCoord comprises diverse tasks that require continuous inter-arm dependency and dynamic role exchange across multiple sub-goals. Also, we propose a suite of quantitative metrics that evaluate coordination from temporal, spatial, and spatial-temporal perspectives, enabling systematic measurement of bimanual cooperation. Experimental results show that representative manipulation policies, e.g., DP, RDT, Pi0, and OpenVLA-OFT, struggle with long-duration and highly coupled tasks, revealing fundamental challenges in achieving long-horizon and tight coordination tasks. We hope BiCoord can serve as a foundation for studying long-horizon cooperative manipulation and inspire future research on coordination-aware robotic learning. All datasets, codes and supplements could be found at https://buaa-colalab.github.io/BiCoord/.

Problem

• bimanual manipulation
• long-horizon coordination
• spatial-temporal coupling
• robotic dexterity
• cooperative manipulation

Innovation

• bimanual manipulation
• long-horizon coordination
• spatial-temporal coupling
• benchmark
• quantitative metrics