Barrier-Constrained Guidance For Real-time Asynchronous VLA Execution

Wenbo Yu, Jifan Li, Wenke Xia, Yuchen Li, Yizhuo Zhang, Pei Ren, Zhengping Che, Di hu · N/A · 2026

Summary

Abstract Summary

While asynchronous execution enables VLA models to operate in real-time environments by predicting the next chunk during execution, it inevitably introduces boundary misalignment with the committed action stream. Existing delay-aware solutions address this at the prohibitive cost of extensive retraining, architectural modifications, or fixed-prefix inpainting that inflates computation when low latency is vital. To address these problems, we propose FreeBar, a training-free real-time inference plugin that enables robust asynchronous execution with only marginal additional inference cost by casting chunk alignment in flow-based VLAs as barrier-guided action-space constraint. Specifically, FreeBar introduces a lightweight Control Barrier Function (CBF)-inspired consistency constraint between consecutive action chunks and solves the resulting correction in closed form, avoiding costly backpropagation through the VLA action head. By progressively tightening the consistency bound during generation, FreeBar aligns generated chunks with the committed execution stream while preserving early-stage generative and within-chunk future action flexibility,. Across Kinetix, RoboCasa, and diverse real-world robotic platforms with multiple VLA backbones, FreeBar improves success rates over naive asynchronous execution by 10%--15% on average, with larger gains in highly dynamic scenarios. It also preserves real-time computational efficiency, decreasing inference latency per action chunk over 45% than RTC. Additionally, as a universal plugin, FreeBar remains robust across distinct latency environments, base policies, and hyperparameter settings.

Key Points

While asynchronous execution enables VLA models to operate in real-time environments by predicting the next chunk during execution, it inevitably introduces boundary misalignment with the committed action stream.
Existing delay-aware solutions address this at the prohibitive cost of extensive retraining, architectural modifications, or fixed-prefix inpainting that inflates computation when low latency is vital.
To address these problems, we propose FreeBar, a training-free real-time inference plugin that enables robust asynchronous execution with only marginal additional inference cost by casting chunk alignment in flow-based VLAs as barrier-guided action-space constraint.
Specifically, FreeBar introduces a lightweight Control Barrier Function (CBF)-inspired consistency constraint between consecutive action chunks and solves the resulting correction in closed form, avoiding costly backpropagation through the VLA action head.
By progressively tightening the consistency bound during generation, FreeBar aligns generated chunks with the committed execution stream while preserving early-stage generative and within-chunk future action flexibility,.

Method

We propose FreeBar, a training-free inference plugin for real-time asynchronous VLA execution. Instead of using VJP-based gradient guidance, FreeBar applies a closed-form action-space projection that aligns the generated chunk with the ongoing execution stream.

The projection imposes a feasible correction set that contracts over denoising time and across the prediction horizon, preserving short-term continuity while ensuring the generation fidelity of the entire action chunk.

Method — Barrier-Constrained Guidance For Real-time Asynchronous VLA Execution — Figure 1: FreeBar enables smooth, efficient, and robust asynchronous VLA execution. (Top) Swift and fluent robotic responses to dynamic targets (e.g., a moving button). (Bottom) FreeBar ensures smoother kinematics than RTC, efficient inference via lightweight constraint over costly VJP, and robust stability across diverse latencies, frameworks, and hyperparameters. ( ∗ GR00T on RoboCasa, 4 denoising steps.)

Experiments

To comprehensively evaluate our method, we design our experiments to investigate four key aspects: (1) the improvement of task success rates and execution stability under asynchronous inference in both simulated and real-world environments; (2) the reduction in inference overhead compared to VJP-based baselines like RTC; (3) the necessity of core components, specifically the horizon-wise graded constraints; and (4) the method’s robustness across diverse delay conditions, execution horizons, hyperparameters, and base VLA policies.

Sources and Demonstrations

Figures are reproduced from Yu et al. See the full paper for experimental details and the project page, when the paper links one, for demonstrations and videos.