Search papers, labs, and topics across Lattice.
TemporalFlow-VLA introduces a novel approach to robot manipulation by learning a compact execution history that is grounded in physical interactions, addressing the limitations of traditional vision-language-action models in multi-stage tasks. By utilizing recorded robot states and temporal flow as a training target, the model effectively captures the nuances of action selection based on prior executions, achieving a remarkable 97.63% average success rate on the LIBERO benchmark. This method demonstrates significant advantages in long-horizon manipulation tasks, revealing that both historical content and its temporal order are critical for accurate action prediction.
Achieving over 97% success in long-horizon robot manipulation, TemporalFlow-VLA reveals the critical role of execution history in action prediction.
Vision-language-action (VLA) models leverage pretrained vision-language representations for robot control, yet simply adding historical frames does not reliably capture recent physical change. This is especially problematic in multi-stage manipulation, where visually similar states may require different actions depending on prior execution. To address this challenge, we present TemporalFlow-VLA, which learns compact execution history through physically grounded temporal supervision. Using recorded robot states, robot geometry, and calibrated cameras, we construct robot-surface temporal flow as a training-only target and supervise two execution-aligned temporal queries that provide structured history to the action expert. The geometric supervision path is not evaluated at deployment. TemporalFlow-VLA achieves 97.63 +/- 0.26% average success on LIBERO, including 96.60 +/- 0.87% on LIBERO Long, and 85.5%/84.2% Clean/Randomized success across 12 RoboTwin tasks. It shows its clearest advantage over prior methods on longer-horizon, multi-stage manipulation. Controlled history interventions show that action prediction depends on both historical content and temporal order. With asynchronous feature caching, temporal conditioning maintains single-frame-level server-side sampling latency without additional historical-encoding overhead. Overall, TemporalFlow-VLA provides a compact, physically grounded interface for exploiting ordered execution history without explicit motion estimation or geometric processing at deployment.
