Run metadata:

• ran_at: 2026-07-28T08:01:57+03:00
• job_id: physical_ai_model_to_matter
• task: Task 1 — Physical AI / Embodiment Scan + Task 2 — Ethan Portfolio Engine
• scope: physical AI / embodiment / humanoids / robotics integration / bionics / prosthetics / AI materials / BCI / defense robotics / capital-flow signals
• research freshness: web-current, fetched in this run
• filter applied: signal over hype; deployment over demos; capital flow over PR; thesis changes over generic news; portfolio relevance over curiosity

---

EXECUTIVE TAKE

Physical AI is no longer a single “humanoid hype” lane. The investable stack is separating into five layers:

1. Robot brains / foundation models — NVIDIA, Google DeepMind, Skild, Physical Intelligence, Figure Helix, Shield AI Hivemind.
2. Embodied hardware / humanoid platforms — Figure, Apptronik, NEURA, Humanoid, Boston Dynamics/Hyundai, Tesla Optimus, 1X, Unitree.
3. Deployment channels — BMW, Mercedes-Benz, GXO, Catalyst Brands, Hyundai, Schaeffler/Bosch/Siemens/SAP ecosystems.
4. Defense autonomy — Anduril, Shield AI, AeroVironment, Palladyne AI, Air Force CCA, counter-UAS, border surveillance, autonomous resupply.
5. Model → Matter / autonomous labs — Microsoft Discovery, MatterGen/MatterSim, NSF cloud-programmable autonomous labs, CuspAI, Atinary.

The strongest current signal is capital + deployment convergence: large rounds are not only funding demos; they are attaching to named industrial, logistics, defense, and lab-automation channels. The bottleneck remains useful manipulation, reliability, safety validation, and manufacturing economics.

---

SIGNAL_LAYER

Humanoid deployment is moving from staged demos to named industrial sites

• Figure AI reported that Figure 03 arrived at BMW Group Plant Spartanburg Hall 52 on June 30, 2026, following an earlier Figure 02 assembly-line deployment in 2025. Figure frames the new deployment as Figure 03 + Helix 02 applied to cognitive and dexterous manufacturing logistics tasks. Source URL: https://www.figure.ai/news/f-03-at-bmw (figure.ai)
• Figure also signed an agreement with Catalyst Brands to integrate humanoids into a distribution facility, extending the signal from automotive manufacturing into logistics/retail operations. Source URL: https://www.figure.ai/news/figu…REDACTED (figure.ai)
• Figure claimed its BotQ production ramp delivered 350+ Figure 03 robots and improved production cadence from 1 robot/day to 1 robot/hour. Treat as company-stated and unverified externally, but important because it shifts Figure’s story from model/demo to fleet/data-scale claim. Source URL: https://www.figure.ai/news/ramp…REDACTED (figure.ai)

Interpretation: Figure is currently one of the clearest “deployment proof” names because the signal includes named sites, robot generation transition, production cadence, and a data-fleet thesis. Risk: still company-stated; no independent unit economics or uptime disclosure.

---

Apptronik is becoming an industrial-channel humanoid bet, not only a robot OEM

• Apptronik announced a $520M Series A-X extension, taking its Series A financing to over $935M, with participation from B Capital, Google, Mercedes-Benz, PEAK6, AT&T Ventures, John Deere, and QIA, aimed at scaling production and deployment of Apollo. Source URL: https://www.globenewswire.com/news-release/2026/02/11/3236352/0/en/Appt…REDACTED.html (globenewswire.com)
• TechCrunch reported Apptronik’s raise at a $5B+ valuation, with partnerships including Google DeepMind, GXO, and Mercedes-Benz. Source URL: https://techcrunch.com/2026/02/11/huma…REDACTED/ (techcrunch.com)
• GXO announced a multi-phase R&D initiative with Apptronik around AI-enabled humanoid robots, making GXO one of the important public-company logistics channels for humanoid trials. Source URL: https://investors.gxo.com/node/9391/pdf (investors.gxo.com)

Interpretation: Apptronik is strong on partner quality. Portfolio signal is indirect through Google/Alphabet, Mercedes-Benz, GXO, John Deere, and possibly supply-chain names, because Apptronik itself is private.

---

New capital inflows confirm Physical AI is a funding priority

• Skild AI announced close to $1.4B in funding led by SoftBank, with participation from NVIDIA’s NVentures, Macquarie Capital, Bezos Expeditions, Disruptive, and 1789 Capital, for a robotics foundation model / “Skild Brain” strategy. Source URL: https://www.morningstar.com/news/business-wire/20260114335623/skil…REDACTED (morningstar.com)
• Physical Intelligence was reported by TechCrunch/Bloomberg to be in talks to raise about $1B at a valuation above $11B. This remains reported / not confirmed by company release in the fetched source. Source URL: https://techcrunch.com/2026/03/27/phys…REDACTED/ (techcrunch.com)
• NEURA Robotics announced a Series C of up to $1.4B to build a Physical AI platform, with investors/partners including Tether, Qualcomm, Amazon, NVIDIA, Bosch, Schaeffler, EIB, and others. Source URL: https://neura-robotics.com/record-series-c/ (neura-robotics.com)
• UK-based Humanoid raised $152M Series A at a $1.35B post-money valuation, with beta robot deployments planned for Q4 2026 and partnerships including SAP, NVIDIA, Bosch, Siemens, and Schaeffler. Source URL: https://live.euronext.com/en/financial-news/robo…REDACTED (live.euronext.com)
• Walden Robotics launched from stealth with $300M to build and deploy general-purpose robots for “real work.” Source URL: https://www.waldenrobotics.com/news/wald…REDACTED (waldenrobotics.com)

Interpretation: Funding is concentrating around “foundation model + deployment channel + manufacturing partner,” not isolated academic robotics. This supports the thesis that the investable layer may favor compute, simulation, data, industrial partners, and defense primes before pure humanoid OEMs are public.

---

NVIDIA remains the core horizontal Physical AI infrastructure name

• NVIDIA released new Physical AI models including Cosmos Transfer 2.5, Cosmos Predict 2.5, and Isaac GR00T N1.6, with partners including Boston Dynamics, Caterpillar, Franka Robotics, Humanoid, LG Electronics, and NEURA Robotics. Source URL: https://investor.nvidia.com/news/press-release-details/2026/NVID…REDACTED/default.aspx (investor.nvidia.com)
• NVIDIA also positioned Jetson Thor and Isaac GR00T N models as deployment infrastructure for AGIBOT, Humanoid, LG, NEURA, Noble Machines, and other robotics players. Source URL: https://investor.nvidia.com/news/press-release-details/2026/NVID…REDACTED/ (investor.nvidia.com)
• The June 2026 Cosmos 3 paper says NVIDIA released code, checkpoints, curated synthetic datasets, and benchmarks for physical AI under the Linux Foundation OpenMDW license. Source URL: https://arxiv.org/abs/2606.02800 (arxiv.org)

Interpretation: NVIDIA is the highest-conviction public horizontal exposure in this scan: robotics simulation, synthetic data, edge compute, VLA / world models, and partner ecosystem all converge. Risk is valuation, not signal quality.

---

Google DeepMind is moving robotics reasoning into developer-accessible infrastructure

• Google DeepMind’s Gemini Robotics-ER 1.6 model card describes it as a vision-language model for spatial and physical reasoning; Google says it is available to developers through Gemini API and Google AI Studio. Source URLs: https://deepmind.google/models/model-cards/gemi…REDACTED/ and https://blog.google/innovation-and-ai/models-and-research/google-deepmind/gemi…REDACTED/ (deepmind.google)
• Google DeepMind’s robotics page frames Gemini Robotics-ER as an embodied reasoning model for planning complex tasks and making decisions in the physical world. Source URL: https://deepmind.google/models/gemini-robotics/ (deepmind.google)

Interpretation: Alphabet exposure is less pure than NVIDIA but strategically important: Google is one of the few public companies with frontier models, robotics history, Apptronik connection, and developer distribution.

---

Defense autonomy has moved from “future thesis” into procurement and capital acceleration

• Anduril was selected by the U.S. Air Force for the next phase of the Collaborative Combat Aircraft mission-autonomy program, with Lattice for Mission Autonomy intended to move from concept toward production-ready capability. Source URL: https://www.anduril.com/news/u-s-…REDACTED (anduril.com)
• The U.S. Air Force separately stated that production options were awarded to Anduril, RTX Collins Aerospace, and Shield AI to accelerate mission-autonomy software delivery for CCA. Source URL: https://www.hill.af.mil/News/Article-Display/Article/4521153/air-…REDACTED/ (hill.af.mil)
• Anduril raised $5B at a $61B valuation in May 2026, and Axios reported on July 24, 2026 that Anduril was seeking a new round that could value it around $100B. Source URLs: https://techcrunch.com/2026/05/13/andu…REDACTED/ and https://www.axios.com/2026/07/24/andu…REDACTED (techcrunch.com)
• Shield AI announced it is raising $1.5B Series G at a $12.7B post-money valuation plus $500M fixed-return preferred equity financing, and said the capital supports Hivemind Foundation Model for Defense trained in simulation and refined through real-world operations. Source URL: https://shield.ai/shie…REDACTED/ (shield.ai)
• AeroVironment’s Titan MS counter-UAS system was selected for an $80.5M JIATF-401 contract award supporting Domestic Shield; AV describes Titan MS as fusing AI/ML algorithms and sensors to counter RF-controlled and autonomous drone threats. Source URL: https://www.avinc.com/?avinc_solution_tax=titan-ms-c-uas (avinc.com)

Interpretation: Defense robotics/autonomy is currently more investable than consumer humanoids because contracts, budget channels, and urgent demand are clearer. Public exposure: AVAV, RTX, PLTR-adjacent defense AI ecosystem, LMT/NOC/GD/BA indirectly, but the purest high-growth names Anduril and Shield AI remain private.

---

BCI / bionics are progressing, but commercial investability remains more limited than robotics/defense

• Synchron’s site states its BCI system is investigational, clinical benefit has not been validated, and it is not commercially approved in any geography. Source URL: https://synchron.com/ (synchron.com)
• Synchron’s BCI site says its trial was conducted under FDA IDE oversight and completed 12 months of safety follow-up. Source URL: https://synchronbci.com/ (synchronbci.com)
• A July 2026 landscape analysis of China BCI clinical translation reported 134 ChiCTR clinical trials, 26 investigator-initiated trials, and five BCI-related NMPA-approved products as of June 2026, including semi-invasive, invasive closed-loop DBS, and non-invasive EEG rehabilitation systems. Source URL: https://arxiv.org/abs/2607.07185 (arxiv.org)
• BrainCo demonstrated a BCI-to-robot platform at WAIC 2026 and showed products including Revo 3 Dexterous Hand, Intelligent Bionic Hand, and Intelligent Bionic Leg; this is company/PR-sourced and should be treated as demo signal, not commercial proof. Source URL: https://www.prnewswire.com/news-releases/brai…REDACTED.html (prnewswire.com)
• Open Bionics’ 2026 profile says it took a £600,000 NatWest loan secured against IP to fund six additional U.S. clinics, a much smaller capital signal than humanoids or defense autonomy. Source URL: https://www.bridgeheadcommunications.com/medtech-index/open-bionics (bridgeheadcommunications.com)

Interpretation: BCI/bionics are strategically relevant to human-machine boundary and prosthetic dexterity, but public-market investability is still weak. This remains watchlist unless tied to larger medtech / robotics / neurotech platforms.

---

AI materials / autonomous labs are becoming part of Model → Matter infrastructure

• Microsoft said Microsoft Discovery is now generally available and supports autonomous lab orchestration, AI models for science, continuous learning through the scientific loop, and agentic memory. Source URL: https://blogs.microsoft.com/blog/2026/07/22/powe…REDACTED/ (blogs.microsoft.com)
• Microsoft explicitly tied MatterGen and MatterSim to AI-enabled autonomous laboratories for structural materials and superconductors. Source URL: https://blogs.microsoft.com/blog/2026/07/22/powe…REDACTED/ (blogs.microsoft.com)
• NSF-backed autonomous lab infrastructure accelerated in July 2026: Johns Hopkins received $20M for AI-driven materials discovery networks, CUNY received $18.1M for AI-BIOME cloud-programmable bio-inspired materials labs, and NC State received $20M for SPEED self-driving chemistry/materials platforms. Source URLs: https://engineering.jhu.edu/news/buil…REDACTED/ ; https://asrc.gc.cuny.edu/headlines/2026/07/cuny…REDACTED/ ; https://research.ncsu.edu/nc-s…REDACTED/ (engineering.jhu.edu)
• CuspAI reportedly raised $450M for AI materials discovery, supporting expansion of an AI materials platform for industrial applications. Source URL: https://www.eweek.com/news/cuspai-ai-materials-funding/ (eweek.com)
• Atinary opened a Boston self-driving lab integrating AI with robotics/lab instrumentation from ABB, Agilent, Bruker, Chemspeed, and Mettler-Toledo. Source URL: https://atinary.com/news/open…REDACTED/ (atinary.com)

Interpretation: Model → Matter is not only robots moving boxes. Autonomous lab systems are becoming the “physical execution layer” for science, materials, chemistry, pharma, and national labs. Public exposure: MSFT, ABB, Agilent, Bruker, Mettler-Toledo, NVIDIA, cloud/compute ecosystem.

---

CAPITAL_FLOW_MAP

Mega-round / high-signal private capital

Sources: Skild (morningstar.com); Physical Intelligence (techcrunch.com); NEURA (neura-robotics.com); Apptronik (globenewswire.com); Shield AI (shield.ai); CuspAI (eweek.com); Walden (waldenrobotics.com); Humanoid (live.euronext.com); Anduril (techcrunch.com)

Public-company capital / strategic exposure

• NVIDIA / NVDA: strongest horizontal Physical AI infrastructure exposure via Cosmos, Isaac, GR00T, Jetson Thor, synthetic data, simulation, robot compute, and NVentures participation in Skild. (investor.nvidia.com)
• Alphabet / GOOGL: Gemini Robotics-ER, Google AI Studio/API distribution, Apptronik / Google DeepMind relationship. (deepmind.google)
• Microsoft / MSFT: autonomous lab orchestration, MatterGen/MatterSim, Microsoft Discovery for scientific workflows. (blogs.microsoft.com)
• AeroVironment / AVAV: direct public defense robotics / counter-UAS exposure, including Titan MS $80.5M award. (avinc.com)
• RTX / RTX: named in Air Force CCA autonomy software production options via Collins Aerospace. (hill.af.mil)
• GXO Logistics / GXO: public logistics deployment channel for Apptronik and humanoid automation. (investors.gxo.com)
• Tesla / TSLA: Optimus remains a major option-value narrative, but current fetched evidence is stronger for vehicle/storage operating scale than robot commercial deployment; Q2 2026 Tesla produced 451,758 vehicles, delivered 480,126, and deployed 13.5 GWh of energy storage. Source URL: https://ir.tesla.com/press-release/tesl…REDACTED (ir.tesla.com)
• Hyundai / 005380.KS / HYMTF: pursuing SoftBank’s stake in Boston Dynamics and planning Atlas deployment at HMGMA beginning in 2028, initially for parts sequencing. (hyundai.com)
• Schaeffler / SHA.DE, Siemens / SIE.DE, SAP / SAP, Bosch private: deployment/manufacturing/software ecosystem around Humanoid. (live.euronext.com)
• Qualcomm / QCOM, Amazon / AMZN: named in NEURA’s funding/partner ecosystem and AWS collaboration. (neura-robotics.com)

---

CONVERSION_LAYER

What is converting from hype to deployment?

1. Industrial humanoids - Figure at BMW + Catalyst distribution facility. - Apptronik with GXO / Mercedes-Benz / Google DeepMind. - Humanoid planning beta deployments in Q4 2026 with Schaeffler/Bosch/SAP/Siemens ecosystem. - Hyundai planning Atlas deployment in 2028.

2. Defense autonomy - Air Force CCA autonomy selections. - Anduril Lattice mission autonomy. - Shield AI Hivemind. - AeroVironment Titan MS counter-UAS contract.

3. Robot foundation models - NVIDIA Cosmos/GR00T. - Google Gemini Robotics-ER. - Skild Brain. - Physical Intelligence π0.7 paper. - Shield AI Hivemind Foundation Model for Defense.

4. Autonomous labs / AI materials - Microsoft Discovery general availability. - NSF programmable cloud lab grants. - MatterGen/MatterSim being connected to autonomous labs. - CuspAI funding. - Atinary Boston lab.

What is not yet converted?

• Consumer home humanoids remain weakly validated relative to factory/logistics/defense. 1X offers NEO early access with priority 2026 delivery, but this is still consumer rollout risk, not proven mass-market adoption. Source URL: https://www.1x.tech/discover/neo-home-robot (1x.tech)
• BCI remains clinically meaningful but commercially early; Synchron explicitly says the device is investigational and not commercially approved. (synchron.com)
• Tesla Optimus remains high-upside but under-verified as a near-term external revenue line in fetched sources; Tesla’s verified Q2 data here is vehicles/storage, not robot sales. (ir.tesla.com)

---

TIME_MAP

0–6 months: watch for confirmation

• Figure: BMW Hall 52 operating metrics, Catalyst deployment proof, external verification of BotQ production cadence.
• Apptronik: GXO/Mercedes-Benz multi-site movement, Apollo uptime, safety certification, revenue disclosure.
• Humanoid: Q4 2026 beta deployments at customer sites.
• Anduril / Shield AI / RTX: CCA software test milestones and award-fee / operator-feedback signals.
• AVAV: Titan MS delivery and follow-on counter-UAS awards.
• Microsoft / NSF labs: first evidence of autonomous labs producing validated materials or industrial workflows.

6–24 months: likely inflection window

• Hyundai says Atlas deployment at HMGMA is currently planned to begin in 2028, initially for parts sequencing and later possibly component assembly by 2030. (hyundai.com)
• Humanoid plans beta customer deployments in Q4 2026. (live.euronext.com)
• NVIDIA and Google ecosystem adoption should show through developer uptake, partner deployment, Jetson/robotics compute demand, and simulation-to-real progress.

24+ months: thesis validation

• Humanoids must prove full-shift reliability, safe human collaboration, maintenance economics, and ROI vs fixed automation.
• Defense autonomy must prove procurement scale, field reliability, export policy compliance, and human-in/on-the-loop doctrine acceptance.
• Model → Matter must prove AI-generated materials can move from candidate discovery into validated, manufacturable materials.

---

BOTTLENECK_INDEX

---

THESIS_STATE

Current thesis rating: Strengthened / broadened

The Physical AI thesis is stronger than a humanoid-only narrative. The most durable version is:

AI is moving from language/software into physical execution through robots, autonomous defense systems, industrial automation, lab automation, and materials discovery. The first investable public winners are likely infrastructure, compute, simulation, autonomy software, deployment channels, and defense/lab-tool ecosystems — not necessarily standalone humanoid OEMs.

What changed this cycle

• Deployment evidence improved: Figure at BMW/Catalyst, Apptronik with GXO/Mercedes, Humanoid planning Q4 customer beta deployments, Hyundai’s 2028 Atlas deployment plan.
• Capital intensity increased: Skild, NEURA, Apptronik, Shield AI, Anduril, CuspAI, and Humanoid all show significant funding momentum.
• Defense autonomy became a priority lane: CCA, counter-UAS, border systems, and autonomous mission software are now central to the thesis.
• Model → Matter expanded: Microsoft Discovery and NSF autonomous labs connect AI models to real experiments and manufacturable materials.

What would weaken the thesis

• Industrial humanoid deployments fail to show uptime/ROI.
• Robot foundation models remain demo-good but field-fragile.
• Defense autonomy procurement slows because of policy, safety, or budget reversals.
• AI materials discovery remains candidate-generation without lab validation/manufacturing conversion.
• Public-market proxies overprice the theme before revenue conversion.

---

SOURCES

1. NVIDIA — “NVIDIA Releases New Physical AI Models as Global Partners Unveil Next-Generation Robots” — https://investor.nvidia.com/news/press-release-details/2026/NVID…REDACTED/default.aspx (investor.nvidia.com)
2. NVIDIA — “NVIDIA and Global Robotics Leaders Take Physical AI to the Real World” — https://investor.nvidia.com/news/press-release-details/2026/NVID…REDACTED/ (investor.nvidia.com)
3. NVIDIA Cosmos 3 paper — https://arxiv.org/abs/2606.02800 (arxiv.org)
4. Figure AI — “F.03 Arrives at BMW” — https://www.figure.ai/news/f-03-at-bmw (figure.ai)
5. Figure AI — Catalyst Brands agreement — https://www.figure.ai/news/figu…REDACTED (figure.ai)
6. Figure AI — “Ramping Figure 03 Production” — https://www.figure.ai/news/ramp…REDACTED (figure.ai)
7. Apptronik financing release — https://www.globenewswire.com/news-release/2026/02/11/3236352/0/en/Appt…REDACTED.html (globenewswire.com)
8. TechCrunch — Apptronik valuation/funding — https://techcrunch.com/2026/02/11/huma…REDACTED/ (techcrunch.com)
9. GXO / Apptronik R&D initiative — https://investors.gxo.com/node/9391/pdf (investors.gxo.com)
10. Skild AI funding — https://www.morningstar.com/news/business-wire/20260114335623/skil…REDACTED (morningstar.com)
11. Physical Intelligence funding report — https://techcrunch.com/2026/03/27/phys…REDACTED/ (techcrunch.com)
12. Physical Intelligence π0.7 paper — https://arxiv.org/abs/2604.15483 (arxiv.org)
13. Skild AI collaborations — https://www.globenewswire.com/news-release/2026/03/17/3256839/0/en/skil…REDACTED.html (globenewswire.com)
14. Google Gemini Robotics-ER model card — https://deepmind.google/models/model-cards/gemi…REDACTED/ (deepmind.google)
15. Google Gemini Robotics-ER announcement — https://blog.google/innovation-and-ai/models-and-research/google-deepmind/gemi…REDACTED/ (blog.google)
16. Google DeepMind Gemini Robotics page — https://deepmind.google/models/gemini-robotics/ (deepmind.google)
17. NEURA Robotics Series C — https://neura-robotics.com/record-series-c/ (neura-robotics.com)
18. NEURA / AWS collaboration — https://neura-robotics.com/neur…REDACTED/ (neura-robotics.com)
19. Humanoid Reuters/Euronext funding report — https://live.euronext.com/en/financial-news/robo…REDACTED (live.euronext.com)
20. Walden Robotics launch — https://www.waldenrobotics.com/news/wald…REDACTED (waldenrobotics.com)
21. Hyundai statement on Boston Dynamics — https://www.hyundai.com/worldwide/en/newsroom/detail/0000001225 (hyundai.com)
22. Boston Dynamics Atlas product announcement — https://bostondynamics.com/blog/bost…REDACTED/ (bostondynamics.com)
23. 1X NEO home robot — https://www.1x.tech/discover/neo-home-robot (1x.tech)
24. Tesla Q2 2026 production/deliveries/deployments — https://ir.tesla.com/press-release/tesl…REDACTED (ir.tesla.com)
25. Anduril CCA mission autonomy — https://www.anduril.com/news/u-s-…REDACTED (anduril.com)
26. U.S. Air Force CCA contracts — https://www.hill.af.mil/News/Article-Display/Article/4521153/air-…REDACTED/ (hill.af.mil)
27. TechCrunch — Anduril $5B raise — https://techcrunch.com/2026/05/13/andu…REDACTED/ (techcrunch.com)
28. Axios — Anduril possible $100B valuation round — https://www.axios.com/2026/07/24/andu…REDACTED (axios.com)
29. Shield AI raise / Aechelon acquisition — https://shield.ai/shie…REDACTED/ (shield.ai)
30. AeroVironment Titan MS award — https://www.avinc.com/?avinc_solution_tax=titan-ms-c-uas (avinc.com)
31. AeroVironment 10-K — https://www.sec.gov/Archives/edgar/data/1368622/000110465926078906/avav-20260430x10k.htm (sec.gov)
32. Synchron — investigational BCI statement — https://synchron.com/ (synchron.com)
33. Synchron BCI — FDA IDE / follow-up — https://synchronbci.com/ (synchronbci.com)
34. China BCI landscape analysis — https://arxiv.org/abs/2607.07185 (arxiv.org)
35. BrainCo WAIC 2026 BCI-to-robot demo — https://www.prnewswire.com/news-releases/brai…REDACTED.html (prnewswire.com)
36. Open Bionics profile — https://www.bridgeheadcommunications.com/medtech-index/open-bionics (bridgeheadcommunications.com)
37. Microsoft Discovery / Genesis Mission — https://blogs.microsoft.com/blog/2026/07/22/powe…REDACTED/ (blogs.microsoft.com)
38. Johns Hopkins AI materials grant — https://engineering.jhu.edu/news/buil…REDACTED/ (engineering.jhu.edu)
39. CUNY AI-BIOME grant — https://asrc.gc.cuny.edu/headlines/2026/07/cuny…REDACTED/ (asrc.gc.cuny.edu)
40. NC State SPEED autonomous labs grant — https://research.ncsu.edu/nc-s…REDACTED/ (research.ncsu.edu)
41. CuspAI funding report — https://www.eweek.com/news/cuspai-ai-materials-funding/ (eweek.com)
42. Atinary Boston self-driving lab — https://atinary.com/news/open…REDACTED/ (atinary.com)
43. AI materials validation caution paper — https://arxiv.org/abs/2606.30967 (arxiv.org)
44. Humanoid grasping / deployment pipeline paper — https://arxiv.org/abs/2604.17258 (arxiv.org)