ActionFlow: A Pipelined Action Acceleration for Vision Language Models on Edge
cs.AI, cs.RO
Submitted: 2025-12-23
Updated: 2026-11-01
Comments: 7 pages, 5 figures. Accepted by DAC 2026
License: http://creativecommons.org/licenses/by-nc-nd/4.0/
The gist: Vision-Language-Action (VLA) models have emerged as a unified paradigm for robotic perception and control, enabling emergent generalization and long-horizon task execution.
Terminology
Abstract
Vision-Language-Action (VLA) models have emerged as a unified paradigm for robotic perception and control, enabling emergent generalization and long-horizon task execution. However, their deployment in dynamic, real-world environments is severely hindered by high inference latency. While smooth robotic interaction requires control frequencies of 20-30 Hz, current VLA models typically operate at only 3-5 Hz on edge devices due to the memory-bound nature of autoregressive decoding. Existing optimizations often require extensive retraining or compromise model accuracy. To bridge this gap, we introduce ActionFlow, a system-level inference framework tailored for resource-constrained edge platforms. At the core of ActionFlow is a Cross-Request Pipelining strategy, a novel scheduler that redefines VLA inference as a macro-pipeline of micro-requests. The strategy intelligently batches memory-bound Decode phases with compute-bound Prefill phases across continuous time steps to maximize hardware utilization. Furthermore, to support this scheduling, we propose a Cross-Request State Packed Forward operator and a Unified KV Ring Buffer, which fuse fragmented memory operations into efficient dense computations. Experimental results demonstrate that ActionFlow achieves a 2.55x improvement in FPS on the OpenVLA-7B model without retraining, enabling real-time dynamic manipulation on edge hardware. Our work is available at https://github.com/Frankd35/Action-Flow.
Sources
- $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
- Real-Time Execution of Action Chunking Flow Policies
- Fast-in-Slow: A Dual-System Foundation Model Unifying Fast Manipulation within Slow Reasoning
- OpenVLA: An Open-Source Vision-Language-Action Model
- Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
- CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation
- RoboNurse-VLA: Robotic Scrub Nurse System based on Vision-Language-Action Model
- Vision-Language-Action (VLA) Models: Concepts, Progress, Applications and Challenges
- Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism
- PD-VLA: Accelerating Vision-Language-Action Model Integrated with Action Chunking via Parallel Decoding
- BitVLA: 1-bit Vision-Language-Action Models for Robotics Manipulation
- VLA-Cache: Efficient Vision-Language-Action Manipulation via Adaptive Token Caching
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents