ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learning

TL;DR

ABot-M0 introduces a unified framework for robotic manipulation that leverages large-scale data, a low-dimensional action manifold, and modular perception to enhance generalization, efficiency, and stability in embodied agents.

Contribution

The paper presents a comprehensive data pipeline, the Action Manifold Hypothesis, and modular perception mechanisms, advancing general-purpose robotic manipulation with large-scale datasets and novel learning strategies.

Findings

Constructed UniACT-dataset with over 6 million trajectories from six datasets.

Unified pre-training improves transferability across robots and tasks.

Action Manifold Learning enhances action prediction speed and stability.

Abstract

Building general-purpose embodied agents across diverse hardware remains a central challenge in robotics, often framed as the ''one-brain, many-forms'' paradigm. Progress is hindered by fragmented data, inconsistent representations, and misaligned training objectives. We present ABot-M0, a framework that builds a systematic data curation pipeline while jointly optimizing model architecture and training strategies, enabling end-to-end transformation of heterogeneous raw data into unified, efficient representations. From six public datasets, we clean, standardize, and balance samples to construct UniACT-dataset, a large-scale dataset with over 6 million trajectories and 9,500 hours of data, covering diverse robot morphologies and task scenarios. Unified pre-training improves knowledge transfer and generalization across platforms and tasks, supporting general-purpose embodied intelligence. To improve action prediction efficiency and stability, we propose the Action Manifold Hypothesis: effective robot actions lie not in the full high-dimensional space but on a low-dimensional, smooth manifold governed by physical laws and task constraints. Based on this, we introduce Action Manifold Learning (AML), which uses a DiT backbone to predict clean, continuous action sequences directly. This shifts learning from denoising to projection onto feasible manifolds, improving decoding speed and policy stability. ABot-M0 supports modular perception via a dual-stream mechanism that integrates VLM semantics with geometric priors and multi-view inputs from plug-and-play 3D modules such as VGGT and Qwen-Image-Edit, enhancing spatial understanding without modifying the backbone and mitigating standard VLM limitations in 3D reasoning. Experiments show components operate independently with additive benefits. We will release all code and pipelines for reproducibility and future research.