FASTER: Rethinking Real-Time Flow VLAs

TL;DR

This paper introduces FASTER, a method that significantly reduces reaction latency in real-time vision-language-action models by adaptively prioritizing immediate actions, enabling more responsive robot behaviors.

Contribution

FASTER proposes a horizon-aware scheduling approach that compresses immediate reaction sampling, improving real-time responsiveness without sacrificing long-term trajectory quality.

Findings

FASTER reduces reaction latency tenfold in real-world tasks.

It enables rapid, accurate, and smooth trajectories in dynamic environments.

Experimental results show improved responsiveness on consumer-grade GPUs.

Abstract

Real-time execution is crucial for deploying Vision-Language-Action (VLA) models in the physical world. Existing asynchronous inference methods primarily optimize trajectory smoothness, but neglect the critical latency in reacting to environmental changes. By rethinking the notion of reaction in action chunking policies, this paper presents a systematic analysis of the factors governing reaction time. We show that reaction time follows a uniform distribution determined jointly by the Time to First Action (TTFA) and the execution horizon. Moreover, we reveal that the standard practice of applying a constant schedule in flow-based VLAs can be inefficient and forces the system to complete all sampling steps before any movement can start, forming the bottleneck in reaction latency. To overcome this issue, we propose Fast Action Sampling for ImmediaTE Reaction (FASTER). By introducing a Horizon-Aware Schedule, FASTER adaptively prioritizes near-term actions during flow sampling, compressing the denoising of the immediate reaction by tenfold (e.g., in $\pi_{0.5}$ and X-VLA) into a single step, while preserving the quality of long-horizon trajectory. Coupled with a streaming client-server pipeline, FASTER substantially reduces the effective reaction latency on real robots, especially when deployed on consumer-grade GPUs. Real-world experiments, including a highly dynamic table tennis task, prove that FASTER unlocks substantially improved real-time responsiveness for generalist policies, enabling rapid generation of accurate and smooth trajectories.