ASPIRE：面向机器人的自主技能發現

摘要

傳統機器人程式設計充滿挑戰：需要組織多模態感知、管理物理接觸動態，並處理多樣化的配置與執行失敗。我們提出 ASPIRE（透過迭代機器人探索的智能技能程式設計），這是一套持續學習系統，能以「程式碼即策略」的範式自主編寫並完善機器人控制程式，同時將經驗累積成可重複使用的技能庫。ASPIRE 能發現跨任務、跨模擬與真實環境、以及跨機器人形態皆能持久存在的技能。系統在開放式循環中運作，包含三個組成部分：(1) 閉環機器人執行引擎，能提供細粒度的多模態軌跡，實現自主故障診斷、修復合成與驗證；(2) 持續擴展的技能庫，將經驗證的修復提煉為可重複使用、可遷移的知識；(3) 演化搜索，生成多樣化的任務序列與控制程式，以探索超越單一軌跡的改進。ASPIRE 在 LIBERO-Pro 擾動操作任務中以高達 77% 的優勢超越先前方法，在 Robosuite 雙臂交接任務中以 72% 領先，在 BEHAVIOR-1K 長時域家務任務中則提升 32%。其累積的技能庫亦能對未見過的長時域任務進行零樣本泛化：在 LIBERO-Pro Long 任務中，儘管先前方法使用了測試時推理與重試，ASPIRE 仍以 31% 的成功率遠勝其 4% 的表現。最後，模擬中發現的技能提供了模擬到真實遷移的初步證據，大幅減少了不同機器人形態與應用程式介面上之真實機器人程式設計的工作量。

English

Traditional robot programming is challenging: it requires orchestrating multimodal perception, managing physical contact dynamics, and handling diverse configurations and execution failures. We introduce ASPIRE (Agentic Skill Programming through Iterative Robot Exploration), a continual learning system that autonomously writes and refines robot control programs in a code-as-policy paradigm while compounding experience into a reusable skill library. ASPIRE discovers skills that persist across tasks, simulation and real-world settings, and embodiments. It operates in an open-ended loop with three components: (1) a closed-loop robot execution engine that exposes fine-grained multimodal traces, enabling autonomous failure diagnosis, repair synthesis, and validation; (2) a continually expanding skill library that distills validated fixes into reusable, transferable knowledge; and (3) evolutionary search that generates diverse task sequences and control programs to explore beyond single-trajectory refinement. ASPIRE surpasses prior methods by up to 77% on LIBERO-Pro manipulation under perturbation, 72% on Robosuite bimanual handover, and 32% on BEHAVIOR-1K long-horizon household tasks. Its accumulated library also enables zero-shot generalization to unseen long-horizon tasks: on LIBERO-Pro Long, ASPIRE achieves 31% success versus 4% for prior methods despite their use of test-time reasoning and retries. Finally, simulation-discovered skills provide initial evidence of sim-to-real transfer, substantially reducing real-robot programming effort across different embodiments and robot APIs.