Preprint Open access
VLA-ULAP: Interleaving Cloud VLA Calls with Ultra-Lightweight Local Action Prediction at the Edge
Billion-parameter vision-language-action (VLA) policies run either onboard, consuming substantial power, or on remote servers, adding communication latency. To address these drawbacks and better balance latency and onboard energy consumption, we propose VLA-ULAP. It partitions inference across decision times, interleav …