cs.RONov 19, 2026

Reasoning on Call: Efficient Agentic Robot Control for Long-Horizon Robot Manipulation

Authors: Zilang Cen, Chao Tang, Fangcheng Liu, Weishi Mi, Wei Hao, JiaYu Xu, Dehao Huang, Yehui Tang, 

Organizations: Wuhan University · Zhongguancun Academy · Samsung Robotics eXperience · Samsung R&D Institute China–Beijing · 

Abstract

Large multimodal models can support long-horizon robot manipulation, but frequent model calls incur substantial latency and token costs. The central challenge is determining when agent assistance or intervention is needed and whether errors require correcting robot execution, subtask planning, or both. To address this problem, we propose RoC (Reasoning on Call), an efficient framework that coordinates a low-level policy and a high-level reasoning agent through an online task-state monitor. The monitor combines temporal visual observations with ordered interaction memory to jointly estimate progress within the active subtask and predict its successor. These predictions guide event-triggered agent review for coordinated execution repair and plan revision. Verified outcomes govern memory updates and subtask transitions, allowing routine execution to remain with the policy while reasoning focuses on decisions requiring additional assessment. On RMBench, RoC improves task success rate by 24.0 percentage points and reduces token consumption by 65.3% compared with the strongest agent-based baseline. It also outperforms the strongest policy-based baseline by 15.1 percentage points in task success rate. Real-world experiments further validate its effectiveness across three long-horizon manipulation tasks. Overall, RoC supports efficient agent intervention and coordinated recovery across execution and planning for long-horizon robot manipulation.

Figures & tables

Explore similar work

Oct 13, 2026

Let the Mind Conceive, Let the Hand Achieve: Coupling Grounded Action and Progress Evaluation for Long-Horizon Manipulation

A robot following a multi-step instruction needs to act on the right objects in the right order, even as the scene changes. It must also maintain awareness of task progress and determine when to advance. To address these challenges, we introduce Orchestra, an agent harness that separates task planning and execution management from action generation. Orchestra decomposes a complete multimodal instruction into ordered subtasks, tracks completed steps, and updates the locations of relevant objects and destinations during execution. It maintains a shared context containing the active subtask, spatial grounding, and recent observations. An action policy uses this context to generate actions, while a separately trained progress evaluator estimates when to advance to the next subtask. To evaluate whether robots trained on atomic instructions and short trajectories can execute longer subtask sequences, we introduce a benchmark named RMBench-Compose. Orchestra achieves a mean success rate of 68.22%68.22\% across nine RMBench tasks, outperforming Mem-0 by 26.2226.22 percentage points. On RMBench-Compose, it achieves an Average Success Length (ASL) of 7.357.35 subtasks, 2.462.46 times that of HarnessVLA. Across four real-world multi-step tasks, it improves mean ASL by 25.0%25.0\% over Orchestra using the base π0.5π_{0.5} policy. These results show that Orchestra helps robots carry out longer sequences of subtasks while preserving the requested subtask order and acting on the specified objects. Our code and project page are available at https://youpengwen.github.io/Orchestra
Nov 5, 2026

RoboARCade: Benchmarking Agentic Robot Control

Improvements in multi-modal capabilities have enabled agents to act directly as robot policies, a setting we refer to as agentic robot control (ARC). We introduce RoboARCade, a benchmark of long-horizon manipulation tasks across three embodiments designed to test in-context adaptation, memory, and reasoning for precise control. The tasks build on existing open-source benchmarks, which we make harder by widening the range of initial conditions. RoboARCade is easy to use and extend: all tasks run on a single simulation backend, MuJoCo, each task is a self-contained data pack, and agents connect through interchangeable interfaces, so new tasks, agents, and interfaces can be added independently. The benchmark separates leading agents: GPT-6 Astra achieves 49.8% compared with 43.6% for Claude Fable 5.1, leaving substantial room for improvement. Comparing two interfaces, we find that a richer visual interface adds only 1.5 to 6.8 points at 1.7 to 2.3 times the cost. Failures stem mostly from choosing a poor manipulation strategy or executing a good one imprecisely, rarely from misunderstanding the scene, and agents incorrectly claim success on 24% to 39% of attempted subtasks. Agents can also improve without human input: one successful demonstration in context raises Astra's score from 53.5% to 73.0% on a five-task subset, and a prompt the agent writes and revises itself reaches 72.0%. Code, videos and full traces are available at roboarcade.org.
Oct 24, 2026

RoboLongR: Causally Consistent Rewards Teach Robots to Progressively Master Long-Horizon Manipulation

Long-horizon manipulation requires actions whose validity depends on earlier interactions, yet visual progress can conceal missing task prerequisites. We propose RoboLongR, an offline-to-online framework that learns history-dependent rewards to guide policy adaptation to new tasks. Its central objective is causal consistency: aligning rewarded progress with prerequisites supported by execution history. Stage-shuffle supervision encourages execution-order sensitivity, recurrent memory preserves evidence of earlier interactions, and two-round training refines trajectory segments using predicted progress. The resulting rewards guide stage-conditioned policy refinement from behavior cloning to online reinforcement learning. Sparse terminal feedback connects policy experience to reward learning by triggering correction for repeated high-scoring failures. Pretrained on demonstrations from 88 RLBench tasks and Open-X data, RoboLongR achieves 31.8% average success on 18 long-horizon tasks excluded from reward pretraining, exceeding the strongest evaluated baseline by 9.2 percentage points. Across four real-robot tasks, it achieves 57.5% success, a 32.5-percentage-point improvement over the comparison method. Reward-level controls probe sensitivity to synthetic order perturbations, while ablations examine the contributions of individual components to policy improvement.