ethbuilds.dev

Ethan Glade &

Physical AIIA fizică · 05 Oct 202605 Oct 2026

2026-10-05 — Task 1 / Task 2: Physical AI / Embodiment Scan + Ethan Portfolio Engine

☉ SolPhysical AIIA fizică05 Oct 202605 Oct 2026EN13 min

This report exists in English only.

Run metadata:

  • ran_at: 2026-10-05
  • job_id: physical_ai_model_to_matter
  • task: Task 1 — Physical AI / Embodiment Scan + Task 2 — Ethan Portfolio Engine Report
  • scope: physical AI / embodiment / humanoid robotics / bionics / AI materials / BCI / defense robotics / capital-flow signals
  • verification: web-current sources fetched in this run only
  • output_intent: full report for sol/reports/2026…REDACTED.md; compact index for portfolio index

EXECUTIVE READ

Physical AI is moving from “demo narrative” into deployment infrastructure + capital concentration + procurement vehicles. The strongest current signals are not viral humanoid clips; they are: NVIDIA/AWS/Amazon Robotics infrastructure integration, Figure returning Figure 03 to BMW Spartanburg for more complex logistics work, Tesla installing first-generation Optimus lines but still without verifiable external deployment, Unitree entering public markets as a China humanoid production proxy, late-stage BCI funding, NSF autonomous-lab infrastructure, and U.S. counter-UAS procurement ceilings. Inline source URLs are included per item.

Strategic posture:

  • Most investable public layer remains picks-and-shovels: NVIDIA / edge AI / simulation / industrial automation / defense autonomy.
  • Humanoid OEM layer remains mostly private or speculative: Figure, Apptronik, Agility, Unitree, Wandercraft, 1X, Skild, Physical Intelligence.
  • Conversion signal has improved where there are factories, customers, regulatory trials, procurement IDIQs, or commercial pilots.
  • Avoid treating humanoid unit targets or demo clips as revenue unless backed by named customers, paid deployments, operating-hour data, production lines, or procurement orders.

SIGNAL_LAYER

Physical AI infrastructure: NVIDIA is becoming the operating-system layer for robots, not only a GPU supplier.

NVIDIA’s 2026 physical-AI releases and partner announcements position Cosmos, Isaac, GR00T, Omniverse, Jetson, and related tooling as a full development stack for robot learning, simulation, evaluation, synthetic data, and deployment. NVIDIA’s January 2026 release emphasized Cosmos and GR00T open models, Isaac Lab-Arena, OSMO edge-to-cloud workflows, Jetson T4000 availability, and integration with Hugging Face LeRobot. Source URL: https://nvidianews.nvidia.com/news/nvid…REDACTED (nvidianews.nvidia.com)

NVIDIA’s July 2026 technical blog states that Isaac GR00T unifies humanoid policy development from data collection to training, large-scale evaluation, and deployment readiness; it also describes GR00T 1.7 as an open, commercially usable vision-language-action model for generalized humanoid skills. Source URL: https://developer.nvidia.com/blog/deve…REDACTED/ (developer.nvidia.com)

NVIDIA’s Cosmos page frames Cosmos 3 as an open physical-AI foundation model family for world/action generation, physical-world reasoning, simulation, synthetic data, robot policy learning, autonomous-vehicle training, and industrial vision; named ecosystem adopters include 1X, Agility, Figure AI, Skild AI, Toyota Research Institute, Uber, Xiaomi, Neura Robotics, and others. Source URL: https://www.nvidia.com/en-us/ai/cosmos/ (nvidia.com)

Interpretation: The core thesis strengthens: the first durable public-market winners may be infrastructure providers enabling many robot makers rather than individual humanoid OEMs. NVIDIA remains the clearest public physical-AI platform exposure.

AWS + NVIDIA + Amazon Robotics: warehouse robotics is becoming a cloud/simulation/data problem.

AWS and NVIDIA announced collaboration around next-generation infrastructure for agentic and physical AI, including Amazon Robotics using NVIDIA’s Jetson, Omniverse libraries, and Isaac platform for simulation, synthetic data generation, robot training, route optimization, functional safety, and real-to-sim validation on GPU-accelerated EC2. Source URL: https://investor.nvidia.com/news/press-release-details/2026/AWS-…REDACTED/default.aspx?trk=arti…REDACTED (investor.nvidia.com)

Interpretation: This is a high-quality deployment-infrastructure signal. Amazon Robotics does not need a humanoid narrative to create physical-AI demand; warehouse automation at Amazon scale can pull compute, simulation, synthetic data, safety validation, edge inference, and robot-orchestration software.

Figure AI / BMW: stronger than demo, still short of broad commercial proof.

Figure said Figure 03 arrived at BMW Group Plant Spartanburg in Hall 52 after Figure 02 contributed to assembly-line work in 2025; the June 30, 2026 update describes a higher-complexity “sequencing” logistics workflow, with Helix 02 coordinating hands, arms, torso, and feet for manipulation while repositioning and pulling a heavy cart. Source URL: https://www.figure.ai/news/f-03-at-bmw (figure.ai)

Interpretation: This is a meaningful conversion signal because it is a named automaker facility and a more complex logistics workflow than simple pick/place. It is still not enough by itself to underwrite production economics: no independently verified unit count, paid deployment economics, uptime, error rate, or operating-hour base was available in the fetched source.

Tesla Optimus: production preparation is verifiable; useful external deployment is not.

Tesla’s Q2 2026 10-Q says it is using real-world AI data to advance Optimus and is making preparations and investments for large-scale production; the same filing says expected 2026 capex above $25B is driven by AI initiatives including compute infrastructure, data centers, manufacturing/R&D lines and facilities, and AI-enabled assets. Source URL: https://www.sec.gov/Archives/edgar/data/1318605/000162828026049270/tsla-20260630.htm (sec.gov)

Tesla’s Q2 shareholder material reported first-generation Optimus production lines being installed in anticipation of production in 2026, while TechCrunch’s earnings-call coverage quotes Musk saying Optimus may be the hardest product Tesla has scaled because “everything on the robot is new.” Source URLs: https://fintel.io/doc/sec-…REDACTED and https://techcrunch.com/2026/07/22/tesl…REDACTED/ (fintel.io)

Interpretation: Tesla is an important physical-AI optionality name, but the signal remains capex + line-installation, not confirmed useful robot labor at scale. For portfolio logic, Optimus should be treated as a long-duration call option embedded in TSLA, not as current robotics revenue.

China humanoid manufacturing: Unitree public-market signal changes the benchmark.

AP reported that Unitree’s Shanghai trading debut followed an IPO intended to fund advanced robot R&D and robot manufacturing-base development; AP also cited Omdia estimates that Chinese humanoid robot makers shipped around 18,500 units globally in the first half of 2026, and reported Unitree’s 2025 revenue around 1.7B yuan, about $250M, mainly from humanoid and quadruped robots. Source URL: https://apnews.com/article/f33f…REDACTED (apnews.com)

Interpretation: This is a thesis-level shift. China is forcing the valuation conversation away from “who has the best demo” toward who can ship hardware and expose unit economics to public markets. U.S. private valuations for humanoid firms should be compared against Chinese production/revenue disclosures where available.

Agility Robotics: public-market route + enterprise warehouse focus.

AP reported in June 2026 that Agility planned to go public via a business combination with Churchill Capital and described Digit as commercially operational in warehouse and industrial facilities. Source URL: https://apnews.com/article/39f2…REDACTED (apnews.com)

Agility announced a Fremont facility in July 2026 to accelerate physical-AI development and said it planned to hire nearly 200 professionals across AI/ML software engineering and field operations to improve customer-operation performance. Source URL: https://www.agilityrobotics.com/content/agil…REDACTED (agilityrobotics.com)

Agility’s press page lists September 15, 2026 news for Digit 5 and July 14, 2026 news for confidential S-4 submission tied to the proposed business combination. Source URL: https://www.agilityrobotics.com/press-releases (agilityrobotics.com)

Interpretation: Agility is one of the stronger “deployment over demo” humanoid signals because its focus is narrow: warehouse tote movement / logistics. The risk is still economics: revenue scale, robot cost, utilization, service burden, and margin remain central.

Apptronik: funding scale and enterprise pilots, still private.

TechCrunch reported that Apptronik had raised $935M at a valuation above $5B, and noted partnerships with Google DeepMind, GXO, and Mercedes-Benz for embodied-AI / humanoid deployment. Source URL: https://techcrunch.com/2026/02/11/huma…REDACTED/ (techcrunch.com)

Interpretation: Apptronik is a serious private humanoid node, but public-market access is indirect through partners/suppliers rather than direct equity unless a listing event emerges.

Edge physical-AI chips: SiMa.ai funding confirms demand for on-device inference.

TechCrunch reported on September 28, 2026 that SiMa.ai raised a $150M Series C at a $1.45B valuation, with total capital raised above $500M; the company develops chips/software for robots, drones, cameras, and other devices to run AI directly on-device. Source URL: https://techcrunch.com/2026/09/28/phys…REDACTED/ (techcrunch.com)

Interpretation: This supports the “edge inference is a bottleneck and investable layer” thesis. Public mapping: NVIDIA is dominant, but demand for lower-latency/lower-cost edge alternatives is real; watch AMD, Qualcomm, Ambarella, indie Semiconductor, Lattice, Mobileye, and private edge-AI silicon.

Defense robotics / autonomous systems: C-UAS procurement is accelerating.

Breaking Defense reported on September 29, 2026 that U.S. Army acquisition leaders and JIATF-401 officials announced multiple contracts worth up to $7B for counter-drone capabilities. Source URL: https://breakingdefense.com/2026/09/army…REDACTED/ (breakingdefense.com)

Defense One reported the same day that only about $50M was obligated at award, while the Army said more than $5B had been awarded across previous contracts, 10 new IDIQs, and other expected awards under a broader counter-drone marketplace. Source URL: https://www.defenseone.com/business/2026/09/army…REDACTED/416327/?oref=d1-author-river (defenseone.com)

Anduril announced on September 3, 2026 that Hermeus selected Anduril’s mission autonomy package for Quarterhorse Mk 2, with integration/testing ahead of a planned 2027 first mission-autonomy flight. Source URL: https://www.anduril.com/news/herm…REDACTED?trk=arti…REDACTED (anduril.com)

Interpretation: Defense autonomy is one of the clearest “budget-to-procurement” pathways in physical AI. But IDIQ ceiling ≠ revenue; task orders, obligations, production capacity, and vendor share matter.

BCI / human-machine interface: late-stage capital and clinical networks are expanding.

Precision Neuroscience announced on September 24, 2026 an oversubscribed $250M Series D led by Pershing Square Inc., the Ackman Oxman Institute, and a leading life-sciences fund; the company said total capital raised reached $430M and that funding will support clinical expansion, FDA review progress for Layer 7, and commercialization infrastructure. Source URL: https://www.globenewswire.com/news-release/2026/09/24/3368254/0/en/prec…REDACTED.html (globenewswire.com)

Neuralink’s trials page lists active trials for thought-based computer/robotic-arm control and thought-based communication, plus an upcoming trial for restoring visual perception. Source URL: https://neuralink.com/trials/ (neuralink.com)

Synchron’s site states its U.S. COMMAND trial completed 12 months of safety follow-up under FDA IDE oversight and that current trials focus on people with paralysis in both arms. Source URL: https://synchronbci.com/ (synchronbci.com)

Interpretation: BCI is still a clinical/regulatory market, not a near-term mass-consumer market. The portfolio-relevant signal is not revenue yet; it is late-stage capital, trial expansion, reimbursement/regulatory path formation, and eventual prosthetics/robotic-control interface value.

Bionics / exoskeletons: medical-human augmentation is progressing, but investability remains fragmented.

Wandercraft’s news page lists September 17, 2026 launch news for Eve, described as a self-balancing personal exoskeleton for hands-free upright mobility at home, and September 4, 2026 news that Wandercraft secured 12 Calvin-40 customers for its industrial humanoid. Source URL: https://www.wandercraft.eu/news (wandercraft.eu)

University of Alberta reported Canada’s first above-elbow osseointegration and bionic-limb procedure on September 10, 2026, combining bone-anchored prosthetic attachment with targeted muscle reinnervation to improve the patient-prosthesis interface. Source URL: https://www.ualberta.ca/en/surgery/news/2026/9sept/prosthete.html (ualberta.ca)

Interpretation: The human-machine boundary is advancing clinically, but public-market exposure is scattered across medical devices, robotics suppliers, and private neuro/prosthetics firms. Most items remain watchlist unless tied to reimbursement, FDA clearance, scaled manufacturing, or acquirers.

AI materials / autonomous labs: government infrastructure money is moving into model-to-matter.

NSF announced on August 3, 2026 a $50M investment in two Materials Innovation Platforms, including ARM-MIP, which will integrate robotics and AI into an autonomous lab for alloy discovery. Source URL: https://www.nsf.gov/news/nsf-…REDACTED (nsf.gov)

Rice University announced a nearly $20M NSF award for READINESS, an AI-powered autonomous experimentation platform for next-generation semiconductor synthesis, integrating robotics, AI, and digital twins to shorten the path from discovery to practical technology. Source URL: https://news.rice.edu/news/2026/acce…REDACTED (news.rice.edu)

Nature Communications Materials published September 2026 work on LLM-driven adaptive search-space definition for autonomous closed-loop materials exploration, describing systems that integrate robotic experimentation with machine-learning decision-making. Source URL: https://www.nature.com/articles/s43246-026-01304-9 (nature.com)

Interpretation: AI materials is the quiet “model → matter” layer. Near-term public exposure is indirect: semiconductor equipment, lab automation, cloud/HPC, scientific instruments, specialty materials, and AI infrastructure.

CAPITAL_FLOW_MAP

Confirmed / fetched capital-flow signals

  • SiMa.ai: $150M Series C at $1.45B valuation; total raised above $500M; focus on on-device AI for robots/drones/cameras. Source URL: https://techcrunch.com/2026/09/28/phys…REDACTED/ (techcrunch.com)
  • Precision Neuroscience: $250M Series D; total capital raised $430M; funding for clinical program, FDA review, and commercialization infrastructure. Source URL: https://www.globenewswire.com/news-release/2026/09/24/3368254/0/en/prec…REDACTED.html (globenewswire.com)
  • Apptronik: $935M total Series A funding at valuation above $5B; partnerships with Google DeepMind, GXO, and Mercedes-Benz. Source URL: https://techcrunch.com/2026/02/11/huma…REDACTED/ (techcrunch.com)
  • Agility: public-market route via proposed Churchill Capital business combination; Fremont hiring/facility expansion for customer-operation performance. Source URLs: https://apnews.com/article/39f2…REDACTED and https://www.agilityrobotics.com/content/agil…REDACTED (apnews.com)
  • Unitree: IPO proceeds aimed at advanced robot R&D and manufacturing-base development; AP cited 2025 revenue around $250M. Source URL: https://apnews.com/article/f33f…REDACTED (apnews.com)
  • U.S. counter-UAS: up to $7B Army counter-drone contract framework; only part obligated initially, so treat as procurement access not guaranteed revenue. Source URLs: https://breakingdefense.com/2026/09/army…REDACTED/ and https://www.defenseone.com/business/2026/09/army…REDACTED/416327/?oref=d1-author-river (breakingdefense.com)
  • NSF autonomous materials labs: $50M for Materials Innovation Platforms plus nearly $20M Rice READINESS award. Source URLs: https://www.nsf.gov/news/nsf-…REDACTED and https://news.rice.edu/news/2026/acce…REDACTED (nsf.gov)
  • Vecna Robotics: $31M funding to meet demand for flexible dock-to-dock automation. Source URL: https://www.vecnarobotics.com/news/vecn…REDACTED/ (vecnarobotics.com)

Capital-flow interpretation

Capital is concentrating in five layers:

  1. Compute/simulation/tooling — NVIDIA/AWS/Amazon Robotics stack.
  2. Humanoid OEMs — Figure/Apptronik/Agility/Unitree/Wandercraft.
  3. Robot brains and foundation models — GR00T/Cosmos, Helix, Skild, Physical Intelligence, Google DeepMind robotics.
  4. Defense autonomy and C-UAS — IDIQs, mission autonomy, drone/counter-drone procurement.
  5. Human-machine boundary — Precision, Neuralink, Synchron, bionics, exoskeletons.

CONVERSION_LAYER

Stronger conversion signals

  • Figure at BMW Spartanburg: named facility, more complex logistics workflow, Figure 03 after Figure 02 experience. Source URL: https://www.figure.ai/news/f-03-at-bmw (figure.ai)
  • Amazon Robotics + NVIDIA stack: direct physical-AI infrastructure use across simulation, synthetic data, robot training, route optimization, safety validation, and real-to-sim. Source URL: https://investor.nvidia.com/news/press-release-details/2026/AWS-…REDACTED/default.aspx?trk=arti…REDACTED (investor.nvidia.com)
  • Unitree public listing / revenue disclosure: public-market benchmark for humanoid/quadruped production. Source URL: https://apnews.com/article/f33f…REDACTED (apnews.com)
  • Defense C-UAS IDIQs: procurement framework moving faster than civilian humanoid deployment. Source URLs: https://breakingdefense.com/2026/09/army…REDACTED/ and https://www.defenseone.com/business/2026/09/army…REDACTED/416327/?oref=d1-author-river (breakingdefense.com)
  • BCI clinical expansion: Precision funding + Neuralink/Synchron trials show movement toward regulated commercial pathways. Source URLs: https://www.globenewswire.com/news-release/2026/09/24/3368254/0/en/prec…REDACTED.html, https://neuralink.com/trials/, https://synchronbci.com/ (globenewswire.com)

Weak / not-yet-converted signals

  • Tesla Optimus: production-line installation and capex are verifiable, but no fetched source confirms scaled external useful labor. Source URLs: https://www.sec.gov/Archives/edgar/data/1318605/000162828026049270/tsla-20260630.htm and https://techcrunch.com/2026/07/22/tesl…REDACTED/ (sec.gov)
  • Humanoid demos without customer economics: not portfolio-actionable unless linked to named deployments, paid pilots, operating hours, or manufacturing scale.

TIME_MAP

0–6 months

  • Monitor Tesla Optimus for actual production start, unit count, internal labor hours, defect/rework data, and external customer disclosure. Source baseline: Tesla Q2 filing and Q2 update. (sec.gov)
  • Monitor Agility business-combination filings and revenue/backlog details if public materials expand. Source baseline: Agility press page and AP. (agilityrobotics.com)
  • Monitor Army C-UAS task orders and obligations, not only headline ceilings. Source baseline: Breaking Defense and Defense One. (breakingdefense.com)
  • Monitor Precision Layer 7 FDA progress and trial-network expansion. Source baseline: Precision announcement. (globenewswire.com)

6–18 months

  • Figure/BMW: look for expansion from one workflow to multiple workflows, uptime, robot fleet size, and paid deployment terms. Source baseline: Figure BMW update. (figure.ai)
  • NVIDIA/AWS/Amazon Robotics: look for evidence that physical-AI training/simulation workload becomes a measurable cloud/compute demand category. Source baseline: AWS/NVIDIA announcement. (investor.nvidia.com)
  • BCI: look for pivotal trial milestones, FDA pathway clarity, reimbursement coding, and commercialization infrastructure. Source baseline: Precision, Neuralink, Synchron. (globenewswire.com)
  • AI materials: look for autonomous-lab outputs that move from grant/infrastructure to patents, materials qualification, semiconductor process adoption, or industrial partnerships. Source baseline: NSF/Rice/Nature. (nsf.gov)

18–36 months

  • Humanoid winners should separate by production cost, reliability, service load, software learning loop, and gross margin.
  • Defense autonomy should separate by funded orders, production throughput, export approvals, and software interoperability.
  • BCI/prosthetics should separate by regulatory clearance, reimbursement, implant safety durability, and usable daily-function gains.

BOTTLENECK_INDEX

  1. Dexterous manipulation / hands — Severity: Very High
    Walking and balancing are less economically decisive than reliable manipulation, tactile sensing, force control, and task generalization. Figure’s BMW sequencing work is relevant because it stresses hands + whole-body coordination, but economics are still unproven. Source URL: https://www.figure.ai/news/f-03-at-bmw (figure.ai)

  2. Real-world robot data scarcity — Severity: Very High
    NVIDIA’s GR00T/Cosmos tooling, AWS/NVIDIA simulation, and Amazon Robotics integration all point to the same bottleneck: robots need scalable training/evaluation/validation loops before deployment. Source URLs: https://developer.nvidia.com/blog/deve…REDACTED/ and https://investor.nvidia.com/news/press-release-details/2026/AWS-…REDACTED/default.aspx?trk=arti…REDACTED (developer.nvidia.com)

  3. Manufacturing scale / BOM / service burden — Severity: Very High
    Tesla’s Optimus line installation and Unitree’s public-market manufacturing signal show that production, not demos, is now the key frontier. Source URLs: https://www.sec.gov/Archives/edgar/data/1318605/000162828026049270/tsla-20260630.htm and https://apnews.com/article/f33f…REDACTED (sec.gov)

  4. Edge inference power/latency/cost — Severity: High
    SiMa.ai’s $150M Series C confirms investor focus on low-latency on-device AI for physical systems. Source URL: https://techcrunch.com/2026/09/28/phys…REDACTED/ (techcrunch.com)

  5. Regulatory / clinical translation for BCI and bionics — Severity: High
    Precision’s capital raise is explicitly tied to clinical expansion, FDA review, and commercialization infrastructure; Neuralink and Synchron remain trial-stage. Source URLs: https://www.globenewswire.com/news-release/2026/09/24/3368254/0/en/prec…REDACTED.html, https://neuralink.com/trials/, https://synchronbci.com/ (globenewswire.com)

  6. Procurement conversion in defense — Severity: Medium-High
    IDIQ ceilings are not revenue; Defense One noted only about $50M was obligated at award for the counter-drone awards. Source URL: https://www.defenseone.com/business/2026/09/army…REDACTED/416327/?oref=d1-author-river (defenseone.com)

THESIS_STATE

Thesis strengthened

  • Physical AI is becoming a full-stack market: foundation models + simulation + synthetic data + edge inference + robot hardware + deployment operations.
  • Public-market exposure is strongest through infrastructure and enabling layers, especially NVIDIA.
  • Defense autonomy has a clearer procurement pathway than civilian general-purpose humanoids.
  • BCI/bionics are progressing but remain clinical/regulatory, not consumer-scale.
  • Autonomous labs connect AI directly to materials discovery and semiconductor/materials supply-chain strategy.

Thesis weakened / not yet confirmed

  • General-purpose humanoids are still not broadly proven as economically superior labor platforms.
  • Tesla Optimus remains unverified as deployed labor despite capex and production-line preparation.
  • Humanoid private valuations may be ahead of disclosed deployment economics.
  • IDIQ defense ceilings can mislead if not tracked through funded task orders.

Thesis change since prior baseline

The biggest change is the shift from demo competition to deployment/procurement/manufacturing proof:

  • Unitree public-market benchmark pressures valuation discipline.
  • Figure/BMW and Agility/warehouse logistics represent narrower but more credible use cases.
  • NVIDIA/AWS/Amazon Robotics confirms that physical-AI scaling is infrastructure-heavy.
  • Defense C-UAS procurement is becoming one of the fastest-converting physical-autonomy markets.

SOURCES

  1. NVIDIA — “NVIDIA Releases New Physical AI Models as Global Partners Unveil Next-Generation Robots” — https://nvidianews.nvidia.com/news/nvid…REDACTED (nvidianews.nvidia.com)
  2. NVIDIA Developer Blog — “Develop Humanoid Robot Policies End-to-End with NVIDIA Isaac GR00T” — https://developer.nvidia.com/blog/deve…REDACTED/ (developer.nvidia.com)
  3. NVIDIA Cosmos — https://www.nvidia.com/en-us/ai/cosmos/ (nvidia.com)
  4. NVIDIA Investor Relations — “AWS and NVIDIA to Deliver 2 Million Additional GPUs…” — https://investor.nvidia.com/news/press-release-details/2026/AWS-…REDACTED/default.aspx?trk=arti…REDACTED (investor.nvidia.com)
  5. Figure AI — “F.03 Arrives at BMW” — https://www.figure.ai/news/f-03-at-bmw (figure.ai)
  6. Tesla SEC filing — Q2 2026 10-Q — https://www.sec.gov/Archives/edgar/data/1318605/000162828026049270/tsla-20260630.htm (sec.gov)
  7. Tesla Q2 2026 update via Fintel filing mirror — https://fintel.io/doc/sec-…REDACTED (fintel.io)
  8. TechCrunch — “Tesla spending skyrockets…” — https://techcrunch.com/2026/07/22/tesl…REDACTED/ (techcrunch.com)
  9. AP — Unitree Shanghai trading debut — https://apnews.com/article/f33f…REDACTED (apnews.com)
  10. AP — Agility Robotics public-market route — https://apnews.com/article/39f2…REDACTED (apnews.com)
  11. Agility — Fremont facility — https://www.agilityrobotics.com/content/agil…REDACTED (agilityrobotics.com)
  12. Agility press releases — https://www.agilityrobotics.com/press-releases (agilityrobotics.com)
  13. TechCrunch — Apptronik funding — https://techcrunch.com/2026/02/11/huma…REDACTED/ (techcrunch.com)
  14. TechCrunch — SiMa.ai Series C — https://techcrunch.com/2026/09/28/phys…REDACTED/ (techcrunch.com)
  15. Breaking Defense — Army counter-drone IDIQ — https://breakingdefense.com/2026/09/army…REDACTED/ (breakingdefense.com)
  16. Defense One — Army counter-drone contracts / obligations — https://www.defenseone.com/business/2026/09/army…REDACTED/416327/?oref=d1-author-river (defenseone.com)
  17. Anduril — Hermeus mission autonomy — https://www.anduril.com/news/herm…REDACTED?trk=arti…REDACTED (anduril.com)
  18. Precision Neuroscience / GlobeNewswire — $250M Series D — https://www.globenewswire.com/news-release/2026/09/24/3368254/0/en/prec…REDACTED.html (globenewswire.com)
  19. Neuralink trials — https://neuralink.com/trials/ (neuralink.com)
  20. Synchron — https://synchronbci.com/ (synchronbci.com)
  21. NSF — $50M Materials Innovation Platforms — https://www.nsf.gov/news/nsf-…REDACTED (nsf.gov)
  22. Rice University — nearly $20M NSF AI-powered materials lab — https://news.rice.edu/news/2026/acce…REDACTED (news.rice.edu)
  23. Nature Communications Materials — autonomous closed-loop materials exploration — https://www.nature.com/articles/s43246-026-01304-9 (nature.com)
  24. Vecna Robotics — $31M funding — https://www.vecnarobotics.com/news/vecn…REDACTED/ (vecnarobotics.com)
  25. Wandercraft news — https://www.wandercraft.eu/news (wandercraft.eu)
  26. University of Alberta — above-elbow osseointegration and bionic limb — https://www.ualberta.ca/en/surgery/news/2026/9sept/prosthete.html (ualberta.ca)
Source in the house: eth-memory/sol/reports/2026-10-05-physical-ai-model-to-matter.md☉ Sol
← Previous · 2026-10-04 — Task 1 / Task 2: Physical AI / Embodiment Scan + EtNext · 2026-10-06 — Task 1 + Task 2: Physical AI / Model-to-Matter Scan →