Physical AI フロンティア・レーダー · 2026-10-11

言語:ZH · EN · 日本語

フィジカルAIの最前線を毎日更新:8つの軸、三言語の要約、追跡可能な根拠。

本日の注目

1. RESETTLE: Robotic Recovery through Disagreement-Triggered Retrieval and Efficient Corrective Control

Reliable robotic manipulation requires timely intervention to correct emerging deviations and restore progress after execution errors.

2. Being-M0.7: A Latent World-Action Model for Humanoid Robots

Humanoid loco-manipulation requires coordinated locomotion and manipulation informed by future scene evolution and whole-body motion, yet learning these capabilities is constrained by scarce robot demonstrations.

3. Residual Modeling Closes the Regression and Generative Policy Gap in Robot Learning

Learning from demonstration has enabled impressive robot behaviors. A common choice for policy learning is to use diffusion or flow matching (Flow-Policies), which often outperforms direct action regression trained with mean squared error…

4. ARC: A Reasoning Recipe for Robot Foundation Models

The prevailing approach to improving robot foundation models (RFMs) relies on larger models, more robot demonstrations, and costly training at scale.

5. SkillWeave: Weaving Heterogeneous Demonstrations into Long-Horizon Manipulation Skills

Dexterous manipulation requires both large-scale task progression and precise contact-rich interaction, making it challenging to collect demonstrations that effectively support both regimes.

6. FAITH: Feasibility-Aware Safety-Filtered RL for High-Dimensional Systems

Safe reinforcement learning commonly places safety and task performance in the same policy objective, where they can introduce competing updates.

7. Rewiring Semantics, Dynamics, and Control: A Simple yet Effective Action-Centric Tri-Stream Transformer

Vision-Language-Action (VLA) models have emerged as a prominent framework for complex robotic manipulation, building on the strong semantic understanding of pretrained Vision-Language Models (VLMs).

8. Why humanoid robot demos still fail the generalization test

How better data and human judgment can help robots move beyond demos toward reliable real-world performance.

フロンティア基準(キュレーション)

1. PL-CBF: finite-horizon safety at runtime via parallel fallback rollouts

2. RD-VLA: implicit test-time compute scaling via latent iterative reasoning

3. ISO 25785-1 (under development): safety for dynamically stable industrial mobile robots

4. GR00T N1.7: open reasoning VLA model for humanoid robots

5. RoboWorld: fast and reliable neural simulators for generalist robot policy evaluation

6. NVIDIA Jetson Thor: up to 2070 FP4 TFLOPS and 128 GB at 40-130 W

軸ごとの分布 (直近 7 日の採択)

軸 #
基盤モデル(VLA / WAM) 12
データエンジン(収集 / 精査 / スケーリング) 5
学習と自己改善(RL / 事後学習) 4
シミュレーションと評価(sim2real / 閉ループ) 14
エッジとリアルタイム(オンデバイス推論 / 制御周期) 6
安全・権限・コンプライアンス 2
システムとオーケストレーション(長期タスク / 複数方策) 3
ハードウェアとサプライチェーン(機体 / コスト) 10

直近の推移

標準・規制ウォッチ


本ページの項目別分析は radar-analyst(claude-opus-5.5)が下書きし、決定的な検査とエージェント査読を経て自動的に取り込まれました。人間の確認は経ていません。タイトルと数値は改変していません。

本リポジトリは技術観察であり、投資助言・製品保証・安全認証の結論ではありません。論文の主張は著者による自己報告です。

方法論 · 週間まとめ · アーカイブ