On-Device Robotic Planning: Eliminating Inference Redundancy for Efficient Decision-Making
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916066521776128 |
|---|---|
| author | Lee, Joonhee Shin, Hyunseung Kim, Hyunmi Zhang, Pei Ko, Jeonggil |
| author_facet | Lee, Joonhee Shin, Hyunseung Kim, Hyunmi Zhang, Pei Ko, Jeonggil |
| contents | Reasoning-based robotic policies using large language and vision-language models achieve strong semantic planning capabilities but mostly suffer from a high inference latency that limits practical real-time deployment. In this work, we observe that robotic reasoning workloads contain substantial temporal redundancy, where consecutive observations frequently produce identical actions and subgoals. Based on this insight, we present REIS, a human cognition inspired robotic decision-making framework that minimizes unnecessary reasoning while preserving semantic adaptability. REIS combines lightweight scene gating, KV-steered affordance routing, and deliberative reasoning to accelerate robotic control under embodied constraints. Experiments on ALFRED, and real-world robotic tasks demonstrate that REIS significantly suppresses reasoning overhead while maintaining competitive task performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_31460 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | On-Device Robotic Planning: Eliminating Inference Redundancy for Efficient Decision-Making Lee, Joonhee Shin, Hyunseung Kim, Hyunmi Zhang, Pei Ko, Jeonggil Robotics Systems and Control Reasoning-based robotic policies using large language and vision-language models achieve strong semantic planning capabilities but mostly suffer from a high inference latency that limits practical real-time deployment. In this work, we observe that robotic reasoning workloads contain substantial temporal redundancy, where consecutive observations frequently produce identical actions and subgoals. Based on this insight, we present REIS, a human cognition inspired robotic decision-making framework that minimizes unnecessary reasoning while preserving semantic adaptability. REIS combines lightweight scene gating, KV-steered affordance routing, and deliberative reasoning to accelerate robotic control under embodied constraints. Experiments on ALFRED, and real-world robotic tasks demonstrate that REIS significantly suppresses reasoning overhead while maintaining competitive task performance. |
| title | On-Device Robotic Planning: Eliminating Inference Redundancy for Efficient Decision-Making |
| topic | Robotics Systems and Control |
| url | https://arxiv.org/abs/2605.31460 |