Comvla improves robot cloud communication with language guided data cuts
ComVLA: Communication-Aware Split Inference for VLA Models in 6G-Connected Robotics
Robotics
Summary
Robots that understand images and language need powerful computers, often in the cloud, to decide what to do. Sending all robot vision data over wireless networks can be too slow and costly. The authors created ComVLA, which uses the language instructions robots hear to decide exactly which parts of the images to send. This smart selection reduces the amount of data sent, saving time and computing power while keeping performance nearly the same. Their approach shows that linking how robots think with how they communicate can make robot control more efficient over 6G networks.
connected robotics6G wireless networksvision-language-action modelscloud inferencetoken pruningsemantic communicationchannel capacityRayleigh fadingRician fadingedge computing
Authors
Boliang Liu, Wint Yi Poe, Jingyun Di, Riccardo Trivisonno, Giuseppe Caire
Abstract
Connected robotics is an emerging 6G application where mobile robots follow natural-language instructions to manipulate physical objects. The Vision-Language-Action (VLA) models that enable this are too large to run on the robot; a common trend is to offload inference to the cloud. The wireless link, however, limits how much sensing data the edge can transmit per control step. Two recent lines address this constraint: semantic communication codecs compress sensor data but require channel-specific retraining, and VLA token pruners select tokens from image but ignore the channel. Our insight is that the dense semantic information contained in the language already indicates which visual tokens matter. We propose ComVLA, a framework that uses this language guidance to adapt the VLA token budget to the channel capacity. Transmitting 32 tokens instead of 512 on the LIBERO benchmark, ComVLA cuts inference compute by 74% and inference latency by 22% versus the original OpenVLA-OFT baseline, at a cost of 1.5 pp in average task success (95.4% vs. 96.9%), and it stays within the capacity budget under Rayleigh and Rician fading. These results demonstrate that co-designing VLA inference and wireless communication is a practical direction for 6G-connected robotics.