Is Diversity All You Need for Scalable Robotic Manipulation?

2025-07-08

AI Summary

The role of data diversity in robotic manipulation remains poorly understood, and the empirical intuition that “more diversity is always better” lacks rigorous validation. This work systematically disentangles three orthogonal dimensions of diversity—task, embodiment, and expert—and investigates their effects via multi-platform experiments within a pretraining-finetuning framework. We find that task diversity significantly improves cross-task transfer performance; single-embodiment pretraining enhances cross-platform generalization; and expert diversity introduces velocity-modality ambiguity that degrades policy learning. To address this, we propose a distribution debiasing method that mitigates velocity ambiguity, yielding a 15% performance gain on the GO-1-Pro model—equivalent to scaling pretraining data by 2.5×. Our findings establish a high-quality data curation paradigm: prioritizing high task diversity while restricting embodiment to a single platform. This provides a reproducible, empirically grounded methodology for designing large-scale robotic manipulation datasets.

Technology Category

• Intelligent Robots: Manipulation
• Natural Language Processing: Safety and Robustness
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Data scaling has driven remarkable success in foundation models for Natural Language Processing (NLP) and Computer Vision (CV), yet the principles of effective data scaling in robotic manipulation remain insufficiently understood. In this work, we investigate the nuanced role of data diversity in robot learning by examining three critical dimensions-task (what to do), embodiment (which robot to use), and expert (who demonstrates)-challenging the conventional intuition of "more diverse is better". Throughout extensive experiments on various robot platforms, we reveal that (1) task diversity proves more critical than per-task demonstration quantity, benefiting transfer from diverse pre-training tasks to novel downstream scenarios; (2) multi-embodiment pre-training data is optional for cross-embodiment transfer-models trained on high-quality single-embodiment data can efficiently transfer to different platforms, showing more desirable scaling property during fine-tuning than multi-embodiment pre-trained models; and (3) expert diversity, arising from individual operational preferences and stochastic variations in human demonstrations, can be confounding to policy learning, with velocity multimodality emerging as a key contributing factor. Based on this insight, we propose a distribution debiasing method to mitigate velocity ambiguity, the yielding GO-1-Pro achieves substantial performance gains of 15%, equivalent to using 2.5 times pre-training data. Collectively, these findings provide new perspectives and offer practical guidance on how to scale robotic manipulation datasets effectively.

Problem

• Role of data diversity in robotic manipulation learning
• Impact of task, embodiment, expert diversity on robot performance
• Debiasing method to mitigate velocity ambiguity in policies

Innovation

• Task diversity enhances transfer to novel scenarios
• Single-embodiment data enables efficient cross-platform transfer
• Distribution debiasing mitigates velocity ambiguity in policies