Physical AI Radar · 2026-10-10

Language: ZH · EN · JA

A daily radar on the Physical AI frontier: eight lanes, trilingual digests, traceable evidence.

• Generated: 2026-10-10 07:15 UTC
• Window: papers 2026-10-06 → 2026-10-10 · posts 2026-09-10 → 2026-10-10 · no repeats within 30 days (UTC)
• Evidence tags: O = first-party official · R = paper/preprint · M = media/secondary

Today’s picks

1. RoboRender: Robot-Oriented Video Generation for Visual Sim-to-Real Transfer

• Lane: Simulation & evaluation (sim2real / closed-loop) ｜ Evidence: [R] ｜ Source: arXiv 2610.09254 ｜ 2026-10-07
• Key numbers: 71% (“RoboRender-generated data achieve a 71% average success rate”) · 7.1x (“domain randomization by approximately 7.1x and 3.6x”) · 3.6x (“by approximately 7.1x and 3.6x”)
• Signals: quantitative results · closed-loop success reported · real-robot experiments
• Why it matters: Preprint, not yet peer-reviewed; the text gives real-robot results, closed-loop results and figures. Watch how well the simulated result predicts real-robot performance.

Simulation enables large-scale, low-cost robot data generation, but policies trained in simulation often fail to transfer to the real world due to the sim-to-real visual discrepancies.

2. USDCraft: Geometrically Grounded Programmatic Modeling of Articulated 3D Assets for Simulation

• Lane: Simulation & evaluation (sim2real / closed-loop) ｜ Evidence: [R] ｜ Source: arXiv 2610.11322 ｜ 2026-10-08
• Signals: real-robot experiments
• Why it matters: Preprint, not yet peer-reviewed; the text gives no closed-loop results or figures. Watch how well the simulated result predicts real-robot performance.

Geometrically faithful and functional articulated 3D assets are essential for real-to-sim robot manipulation, where policies trained in simulation must transfer to physical objects.

3. Humanoid World Action Model With Joint State–Action Generation

• Lane: Foundation models (VLA / WAM) ｜ Evidence: [R] ｜ Source: arXiv 2610.12026 ｜ 2026-10-08
• Key numbers: 70.6% (“HWAM achieves a 70.6% success rate”) · 43.3% (“compared with 43.3% for Fast-WAM”)
• Signals: quantitative results · closed-loop success reported
• Why it matters: Preprint, not yet peer-reviewed; the text gives no real-robot results. Watch whether it cuts the data a new task or embodiment needs.

Humanoid robots are a promising platform for general-purpose manipulation. Recent Vision-Language-Action (VLA) policies learn actions directly from multimodal observations, while World Action Models (WAMs) further incorporate future…

4. ManiUnit: A Manipulation Skill Dataset and Benchmark for Long-Horizon Tasks

• Lane: Systems & orchestration (long-horizon / multi-policy) ｜ Evidence: [R] ｜ Source: arXiv 2610.12089 ｜ 2026-10-08
• Key numbers: 56% (“success rates by approximately 56% relative to those”) · 78.7% (“on ManiUnit segments achieves 78.7% local manipulation success”) · 49.3% (“compared with 49.3% for a task”) · 4.0% (“raises full-task success from 4.0% to 18.0%”)
• Signals: quantitative results · closed-loop success reported
• Why it matters: Preprint, not yet peer-reviewed; the text gives no real-robot results. Watch whether it holds up over long tasks without human resets.

Long-horizon mobile manipulation requires a robot to navigate multi-room environments and execute a sequence of manipulation skills under a single natural language instruction.

5. Video Friday: Robot Decommissioning Takes a Fun Turn

• Lane: Embodiment & supply chain (hardware / cost) ｜ Evidence: [M] ｜ Source: IEEE Spectrum Robotics ｜ 2026-10-09
• Why it matters: Secondary report, so check the primary source before citing; the text gives no figures. Watch price, availability and who is shipping it in volume.

It’s possible that I wasn’t the only one who thought that Figure’s recent stunt was somewhat ridiculous. Devanthro is in fact selling these plastic balls for $1000 each, and they’ll arrive in time for Christmas.

6. 让触觉像视觉一样可规模化，哥大李昀烛团队 FlexiTac 的开源实践 | IROS 2026

• Lane: Embodiment & supply chain (hardware / cost) ｜ Evidence: [M] ｜ Source: 雷峰网 Leiphone ｜ 2026-10-10
• Why it matters: Secondary report, so check the primary source before citing; the text gives no figures. Watch price, availability and who is shipping it in volume.

FlexiTac压阻式传感器可兼顾柔性、低成本与一致性。机器人的视觉能力正快速增强，但 触觉仍然是具身智能系统中的薄弱环节 。

7. 这届 IROS 2026 上，具身学术与产业边界正在消融

• Lane: Data engine (collection / curation / scaling) ｜ Evidence: [M] ｜ Source: 雷峰网 Leiphone ｜ 2026-10-10
• Why it matters: Secondary report, so check the primary source before citing; the text gives no figures. Watch whether the data is reusable beyond the authors’ own robot.

模型、数据、灵巧操作都在工程化，顶会上具身传递出的信号变了9 月 27 日到 10 月 1 日，全球机器人领域最顶级的国际会议之一 IROS 2026 将在美国匹兹堡开幕。 如果你前两年也在关注这个会，你会发现那时候大家都在 “卷概念”。

8. 再裁员5000人！保时捷明确回归燃油车，高端车型计划涨价20%；腾讯拟发行50亿美元离岸债券，或筹钱加码AI；OpenAI全面上线GPT-6

• Lane: Foundation models (VLA / WAM) ｜ Evidence: [M] ｜ Source: 雷峰网 Leiphone ｜ 2026-10-09
• Why it matters: Secondary report, so check the primary source before citing; the text gives no figures. Watch whether it cuts the data a new task or embodiment needs.

要闻提示 再裁员5000人！保时捷明确回归燃油车，高端车型计划涨价20% 阿里巴巴蔡崇信：开源是欧洲发展 AI 的最佳选择 腾讯拟发行50亿美元离岸债券，或筹钱加码AI 超33亿元！

Frontier baseline (curated)

1. π*0.6 / Recap: a VLA that learns from experience via advantage-conditioned policies

• Lane: Training & self-improvement (RL / post-training) ｜ Evidence: [R] ｜ Source: arXiv 2511.14759 ｜ 2026
• Key numbers: RL from autonomous rollouts + human corrections · value function fine-tuned on online data
• Why it matters: The methodological basis for ‘deployment is training’: to use it, operations must first make intervention logs and success judgements first-class data.

2. Post-train NVIDIA Cosmos 3 Edge for on-device robot control

• Lane: Edge & real-time (on-device inference / control rate) ｜ Evidence: [O] ｜ Source: NVIDIA Technical Blog ｜ 2026-08
• Key numbers: 4B omni-model · 2B Nemotron reasoner · runs on Jetson Thor · receding-horizon control loop
• Signals: closed-loop success reported · open code or weights
• Why it matters: The first reproducible path for a world model on-device: post-train, serve on Thor, then evaluate closed-loop, with code and checkpoint released.

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

• Lane: Edge & real-time (on-device inference / control rate) ｜ Evidence: [O] ｜ Source: NVIDIA ｜ 2026
• Key numbers: 7.5x AI compute vs Orin · 3.5x energy efficiency vs AGX Orin · Jetson T2000: 400 FP4 TFLOPS / 16 GB
• Signals: quantitative results · latency or control-rate numbers
• Why it matters: A generational jump in on-device compute creates room for world models and test-time compute, yet the real-time budget still has to be measured item by item.

4. Jetson-PI: onboard real-time control via foresight-aligned asynchronous inference

• Lane: Edge & real-time (on-device inference / control rate) ｜ Evidence: [R] ｜ Source: arXiv 2607.12659 ｜ 2026
• Key numbers: 8.66x control frequency vs naive PyTorch on Orin · +14.8% average success vs VLASH on LIBERO
• Signals: latency or control-rate numbers
• Why it matters: Nearly an order of magnitude in control frequency from systems engineering on the same hardware - deployment optimisation often has more headroom than swapping models.

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

• Lane: Simulation & evaluation (sim2real / closed-loop) ｜ Evidence: [R] ｜ Source: arXiv 2607.01060 ｜ 2026
• Key numbers: closed-loop rollouts in a video world model · VLM judge · rankings strongly correlated with real-world eval
• Why it matters: If rankings inside a world model align with real-robot rankings, the evaluation budget shifts from human hours to GPU hours - real-robot spot checks still cannot be dropped.

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

• Lane: Safety, permissions & compliance ｜ Evidence: [R] ｜ Source: arXiv 2605.16588 ｜ 2026
• Key numbers: library of fallback policies · least-invasive safe mode · QP minimally modifies the nominal policy
• Why it matters: ‘Least-invasive fallback’ can go straight into customer acceptance criteria - far more concrete than ‘we are safe’.

Lane distribution (picks, last 7 days)

Recent cadence

• Source mix (picks, last 7 days): [O] 0 · [R] 40 · [M] 16
• Picks per recent run: 10-04 8 · 10-05 8 · 10-06 8 · 10-07 8 · 10-08 8 · 10-09 8 · 10-10 8

Standards & compliance watch

• ISO 25785-1 — Safety requirements for dynamically stable industrial mobile robots — Under development; committee-stage as of 2026-09.
• Regulation (EU) 2023/1230 (Machinery Regulation) — Fully applicable 2027-01-20.
• EU AI Act official timeline (Council of the EU) — High-risk deadlines 2027-12-02 / 2028-08-02 per Digital Omnibus.

This repository is technical observation. It is not investment advice, a product commitment, or a safety certification conclusion. Paper claims are author-reported.

Methodology · Weekly roundup · Archive