Search papers, labs, and topics across Lattice.
This paper introduces FlashVLA, a novel streaming action decoding framework designed to enhance the efficiency of Vision-Language-Action (VLA) models for robotic manipulation. By utilizing a streaming action buffer with multiple noise-level chunks and chunk-wise causal attention, FlashVLA achieves low-latency inference and maintains action continuity, enabling smooth asynchronous execution. Experimental results demonstrate that FlashVLA can achieve control frequencies of 30 Hz or more on a single GPU while preserving strong task performance in both simulated and real-world environments.
FlashVLA achieves over 30 Hz control frequency with smooth asynchronous execution, revolutionizing real-time robotic manipulation.
Vision-Language-Action (VLA) models are increasingly promising for robotic manipulation, yet their real-world deployment remains bottlenecked by high inference latency and unstable asynchronous execution. This challenge is particularly pronounced in flow-matching-based VLA models, where action decoding requires multiple iterative steps conditioned on the VLM context. While efficient inference methods improve control frequency and asynchronous methods reduce execution idle time, existing approaches often fail to jointly achieve low-latency inference and accurate, temporally consistent asynchronous execution. We introduce \textbf{FlashVLA}, a streaming action decoding framework that addresses both challenges in a unified formulation. FlashVLA maintains a streaming action buffer with multiple chunks at different noise levels and decodes them using chunk-wise causal attention. This design allows FlashVLA to produce one executable action chunk per inference step. Moreover, its chunk-wise autoregressive formulation implicitly preserves action continuity, enabling smooth asynchronous execution without extra future-state conditioning. Across extensive simulated and real-world experiments, FlashVLA substantially improves inference speed while maintaining strong task performance. It can achieve $\geq$30\,Hz control frequency on a single GPU with smooth asynchronous inference in real-world deployment.
