Robotics · 2026-11-05
PyRUA-REAL: Robot Exploration and Adaptive Learning through RSI in the Real World
Qiang Wang, Jianxin Bi, Shulong Jiang, Ruiyang Si, Wenbo Huang, Kaiqi Chen, Jiaxin Li, Liyun Li, Xuanyi Li, Daquan Zhou
AI-assisted summary based on the paper. Consult the linked source for methods and limitations. Benchmark numbers appear after editor verification.
TL;DR
PyRUA-REAL is an embodied recursive self-improvement framework for real-world robot manipulation. Coding agents iteratively explore, execute, diagnose, and revise robot programs, storing interaction experience as persistent memory and reusable executable skills. The resulting programs can be frozen and executed without runtime language-model calls, while still using live observations during execution. Across eight real-world tasks, PyRUA-REAL reports 91.1% average success.
Why it matters
The authors study how physical execution experience can be turned into reusable robot skills that support continued improvement and robust, efficient deployment. They evaluate both development-stage performance and “agent-free” frozen-program execution, and they analyze how multi-agent coordination, execution feedback, explicit task requirements, and prior skills affect revision efficiency, robustness, and transfer—using real-world manipulation and a RoboDojo benchmark comparison.
Method
- Researcher–engineer collaboration develops task programs: researcher selects technical approach and stopping criteria; engineer implements, executes trials, diagnoses failures, and revises code with iterative experiments.
- Persistent experience repository stores facts, problem/outcome/failure causes, and implementation-level reusable task programs/primitives; later exploration retrieves relevant historical tasks and skills to adapt code.
- Deployment uses frozen programs that run without runtime language-model calls, while still responding to live observations through perception, planning, control, and recovery routines.
Limitation
The paper notes that retrospective selected-version summaries (in the execution-feedback protocol) “do not establish that an agent could identify the selected version or stop at it online.”
Abstract (from arXiv)
Coding agents powered by models such as GPT-6-Astra connect spatial reasoning with executable robot behavior, raising the question of how such agents can produce lasting improvements in robot capabilities from physical interaction. In this work, we study embodied recursive self-improvement (RSI) through real-world exploration, code refinement, and deployment using PyRUA-REAL, a framework in which coding agents retain interaction experience as persistent memory and reusable executable skills that inform subsequent exploration and improvement. Our empirical study places physical manipulation at the center of the improvement process, examining how agents collaborate, what kinds of execution feedback support effective code refinement, and how task specifications and previously acquired skills influence robustness and task transfer. We further investigate the efficiency and capability of the resulting programs when deployed without involving an online agent for decision-making. Across eight real-world tasks, PyRUA-REAL achieves an average success rate of 91.1%. In complementary evaluation on 42 tasks from the RoboDojo benchmark, it achieves an average success rate of 84.4%. This study provides empirical guidance for turning physical exploration into reusable robot skills that support continued improvement and robust, efficient deployment.
Related papers
- Toward Recursive Self-Improvement in Physical AI
- EmbodiedEvo: Self-improving Coding Agents for Physical Intelligence
- Recursive Self-Improvement in Physical AI: Mechanisms, Evidence, and Evaluation
- AdaRSI: On-Demand Recursive Self-Improvement for Embodied Agents
- EmbodiedRSI: Active Continual Robot Learning Through Hypothesis-Guided Co-Evolution
Publication record
Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.
First public release: 2026-11-04 · Indexed by Pub-AI: 2026-11-05 UTC
PyRUA-REAL: Robot Exploration and Adaptive Learning through RSI in the Real World
Preprint posted: 2026-11-04 · Original record ↗
- arxiv
- 2611.06768
Explicit source versions (1)
- v1 · 2026-11-04
- Preprint available
Source date: 2026-11-04 · First detected: 2026-11-05 UTC · Historical baseline
Source evidence (2)
- Official evidence ↗ · RSS guid=oai:arXiv.org:2611.06768v1; arxiv:announce_type=new; item pubDate (official metadata)
- Official evidence ↗ · atom:published (official metadata)
Abstract · arxiv
Coding agents powered by models such as GPT-6-Astra connect spatial reasoning with executable robot behavior, raising the question of how such agents can produce lasting improvements in robot capabilities from physical interaction. In this work, we study embodied recursive self-improvement (RSI) through real-world exploration, code refinement, and deployment using PyRUA-REAL, a framework in which coding agents retain interaction experience as persistent memory and reusable executable skills that inform subsequent exploration and improvement. Our empirical study places physical manipulation at the center of the improvement process, examining how agents collaborate, what kinds of execution feedback support effective code refinement, and how task specifications and previously acquired skills influence robustness and task transfer. We further investigate the efficiency and capability of the resulting programs when deployed without involving an online agent for decision-making. Across eight real-world tasks, PyRUA-REAL achieves an average success rate of 91.1%. In complementary evaluation on 42 tasks from the RoboDojo benchmark, it achieves an average success rate of 84.4%. This study provides empirical guidance for turning physical exploration into reusable robot skills that support continued improvement and robust, efficient deployment.
