ethbuilds.dev

Ethan Glade &

Physical AI · 02 Aug 2026

2026-08-02 — Task 1: Physical AI / Embodiment Scan + Task 2: Ethan Portfolio Engine

☉ SolPhysical AI02 Aug 2026EN12 min

This report exists in English only.

Run metadata:

  • ran_at: 2026-08-02
  • task: physical_ai_model_to_matter
  • scope: physical AI / embodiment / robotics / bionics / AI materials / BCI / defense robotics / capital-flow signals
  • verification: web-current, sources fetched in this run only
  • output intent: full research report + compact portfolio index

EXECUTIVE_READ

Physical AI has moved from “humanoid demo season” into infrastructure + deployment + capital-stack formation. The strongest signals this run are:

  1. Humanoid deployment is becoming plant-specific, not just demo-specific. BMW confirmed Figure 03 is moving into BMW Plant Spartanburg logistics after the earlier Figure 02 deployment; this is a real manufacturing-environment signal, though still not broad-scale revenue proof. Source URL: https://www.press.bmwgroup.com/global/article/detail/T0458778EN/bmw-…REDACTED (press.bmwgroup.com)
  2. Robot-brain layer is becoming investable infrastructure. NVIDIA is shipping Cosmos / Isaac / GR00T physical-AI stack, Skild AI announced a Series C around a general-purpose robot brain, and Physical Intelligence is reportedly in/around another major financing cycle. Source URLs: https://nvidianews.nvidia.com/news/nvid…REDACTED ; https://www.skild.ai/blogs/series-c ; https://techcrunch.com/2026/03/27/phys…REDACTED/ (nvidianews.nvidia.com)
  3. Defense autonomy is crossing from R&D into acquisition rails. The U.S. Army launched a UAS Marketplace to shorten procurement, the Air Force is advancing CCA open architecture and weapons-integration testing, and Shield AI was selected as a mission-autonomy provider for CCA. Source URLs: https://www.army.mil/article-amp/291297/us_a…REDACTED ; https://www.af.mil/News/Article-Display/Article/4405471/air-…REDACTED/ ; https://shield.ai/shie…REDACTED/ (army.mil)
  4. AI materials / autonomous labs are becoming state-backed strategic infrastructure. SandboxAQ announced a $500M U.S. Commerce CHIPS award for AI-driven semiconductor materials discovery, while NSF-backed cloud/autonomous lab projects at Johns Hopkins / UChicago / Argonne / CUNY show public capital moving into programmable physical discovery. Source URLs: https://www.sandboxaq.com/press/sand…REDACTED ; https://engineering.jhu.edu/news/buil…REDACTED/ ; https://www.computerscience.uchicago.edu/news/remo…REDACTED/ ; https://www.eurekalert.org/news-releases/1137370 (sandboxaq.com)
  5. Human-machine boundary remains clinically real but commercially pre-revenue. Neuralink lists active clinical trials including Cleveland Clinic Abu Dhabi; Paradromics reported first Connexus BCI implantation for an FDA-approved early feasibility study; Precision Neuroscience partnered with Medtronic. Source URLs: https://neuralink.com/trials/ ; https://paradromics.com/news/?e448aeee_page=1 ; https://www.globenewswire.com/news-release/2026/01/12/3217312/0/en/Prec…REDACTED.html (neuralink.com)

SIGNAL_LAYER

Humanoid ecosystem — deployment signals stronger than demo signals

Figure AI / BMW is the highest-quality humanoid deployment signal this run. BMW’s June 25, 2026 release states Figure 03 will start at BMW Plant Spartanburg in complex logistics sequencing applications after the earlier Figure 02 deployment; Figure’s own update says F.03 arrived at BMW and frames the use case as dynamic material manipulation not suited to fixed automation. Source URLs: https://www.press.bmwgroup.com/global/article/detail/T0458778EN/bmw-…REDACTED ; https://www.figure.ai/news/f-03-at-bmw (press.bmwgroup.com)

Figure’s Helix 02 is meaningful because the claim is full-body autonomy from pixels, but the portfolio weight should follow factory-hours and task economics, not demo claims. Figure’s January 2026 Helix 02 post describes a single neural system controlling the full body from pixels for dexterous long-horizon autonomy. Source URL: https://www.figure.ai/news/helix-02 (figure.ai)

Apptronik moved into mega-round territory. Apptronik announced a $520M Series A-X on February 11, 2026, with repeat investors including B Capital, Google, Mercedes-Benz and PEAK6, and new investors including AT&T Ventures, John Deere and QIA, to scale Apollo humanoid production/deployment. Source URL: https://apptronik.com/news-collection/appt…REDACTED (apptronik.com)

Agility / GXO remains a useful deployment benchmark because it is logistics RaaS, not only robotics theater. GXO and Agility’s agreement described Digit robots being deployed in live warehouse operations via RaaS with Agility Arc cloud automation. Source URL: https://www.agilityrobotics.com/content/gxo-…REDACTED (agilityrobotics.com)

Unitree is becoming the clearest China capital-market signal. Xinhua reported on June 1, 2026 that Unitree passed STAR Market IPO review and expected H1 2026 revenue of roughly RMB 1.052B–1.128B. Source URL: https://english.news.cn/20260601/a855…REDACTED/c.html (english.news.cn)

Tesla Optimus remains high-optionality but lower-verification than BMW/Figure or GXO/Agility. Tesla’s investor-relations page confirms the Q2 2026 shareholder deck was released July 22, 2026; third-party summaries of that deck say Tesla decommissioned Model S/X Fremont lines and is installing first-generation Optimus lines, but there is still no externally verified commercial deployment metric comparable to factory-hours, revenue, or customer contract volume. Source URLs: https://ir.tesla.com/ ; https://evwire.com/p/tesl…REDACTED (ir.tesla.com)

Robot-brain / physical intelligence infrastructure

NVIDIA is positioned as the public-market “picks and shovels” layer for robot learning. NVIDIA released Cosmos and GR00T models/data, Isaac Lab-Arena, and OSMO workflows for robot learning/evaluation; a separate NVIDIA investor release says FieldAI and Skild AI are using Cosmos / Isaac simulation frameworks and that humanoid leaders including 1X, Agility, Boston Dynamics, Figure and others are building on Cosmos / Isaac. Source URLs: https://nvidianews.nvidia.com/news/nvid…REDACTED ; https://investor.nvidia.com/news/press-release-details/2026/NVID…REDACTED/ (nvidianews.nvidia.com)

Google DeepMind re-entered the physical-AI headline layer days before this scan. Axios reported on July 30, 2026 that Google announced Gemini Robotics 2 updates intended to improve robot dexterity and control; this is an important capability signal but still needs deployment validation. Source URL: https://www.axios.com/2026/07/30/goog…REDACTED (axios.com)

Skild AI is one of the strongest private “robot brain” capital-flow signals. Skild announced Series C on January 14, 2026, saying it is building a single general-purpose brain for any robot/task; TechCrunch reported the round placed Skild around a $14B valuation. Source URLs: https://www.skild.ai/blogs/series-c ; https://techcrunch.com/2026/01/14/robo…REDACTED/ (skild.ai)

Physical Intelligence remains a key robot-foundation-model watch, but its commercial model is less visible. Its site lists π0.7 and other 2026 robot-foundation-model work; TechCrunch/Bloomberg reported discussions for about $1B at an $11B+ valuation. Source URLs: https://www.pi.website/?trk=public_post-text ; https://techcrunch.com/2026/03/27/phys…REDACTED/ (pi.website)

Dexterity / hands / bionics-to-robotics transfer

Dexterity remains the economic frontier. ABB and PSYONIC announced a collaboration using the PSYONIC Ability Hand with ABB GoFa cobots to combine touch sensing, compliant design and human-derived training data for robotic dexterity; this is a direct bridge from prosthetics data to industrial manipulation. Source URLs: https://www.abb.com/global/en/news/136690/prsr…REDACTED ; https://www.psyonic.io/news/pres…REDACTED (abb.com)

Ottobock / Blue Arbor is a bionics-control signal rather than a robotics revenue signal yet. Ottobock announced a strategic investment in Blue Arbor Technologies, whose RESTORE neuromuscular interface is intended to enable more intuitive arm-prosthesis control. Source URL: https://corporate.ottobock.com/en/media/newsroom/otto…REDACTED (corporate.ottobock.com)

Defense robotics / autonomous systems

CCA is moving from concept to integrated test architecture. The Air Force said in February 2026 it is implementing Autonomy Government Reference Architecture across multiple platforms, with RTX Collins and Shield AI mission-autonomy vendors supporting General Atomics YFQ-42 and Anduril YFQ-44 respectively. Source URL: https://www.af.mil/News/Article-Display/Article/4405471/air-…REDACTED/ (af.mil)

Weapons integration is now in the CCA test path. The Air Force said CCA entered developmental weapons-integration and captive-carry testing with inert munitions to validate airworthiness, safety and systems performance. Source URL: https://www.af.mil/News/Article-Display/Article/4412705/coll…REDACTED/ (af.mil)

Shield AI is now embedded in a flagship U.S. autonomy program. Shield AI announced selection as a mission-autonomy provider for CCA, with Hivemind integrated on Anduril’s Fury / YFQ-44A and supporting system-level testing. Source URL: https://shield.ai/shie…REDACTED/ (shield.ai)

Army procurement rails are being redesigned around drones. The U.S. Army launched a UAS Marketplace on March 24, 2026, developed with AWS and Army cloud stakeholders, to let Army units, partners and allies procure vetted UAS faster. Source URL: https://www.army.mil/article-amp/291297/us_a…REDACTED (army.mil)

Autonomous logistics is now a production-contract use case. PM UAS awarded SURVICE Engineering a production contract for Joint Autonomous Aerial Resupply Systems, procured via UAS Marketplace BOA, explicitly to reduce lead times and offload battlefield resupply from soldiers to autonomous machines. Source URL: https://www.army.mil/article-amp/292147/pm_u…REDACTED (army.mil)

AI materials / model-to-matter / autonomous labs

DARPA is explicitly asking whether intelligence can move into materials themselves. DARPA’s April 2026 RFI asks for materials capable of intermixed sensing, adapting and acting in real time without continuous external computation or communications, calling this “physical intelligence” for untethered robotics. Source URLs: https://www.darpa.mil/news/2026/rethinking-robotics ; https://www.darpa.mil/work/-with-us/opportunities/darpa-sn-26-76 (darpa.mil)

SandboxAQ’s $500M CHIPS award is one of the strongest model-to-matter capital signals. SandboxAQ said the U.S. Commerce award funds Large Quantitative Models / ReAQT to virtually screen millions of candidate materials and validate promising candidates with lab partners for semiconductor manufacturing materials. Source URL: https://www.sandboxaq.com/press/sand…REDACTED (sandboxaq.com)

Programmable cloud labs are turning autonomous experimentation into national infrastructure. Johns Hopkins announced a $20M NSF-backed AI-powered lab network for extreme-condition materials; UChicago / Argonne announced PoLARIS, a remotely operated robotic lab with 20+ autonomous stations; CUNY received $18.1M NSF funding for AI-BIOME, a cloud-programmable national lab for bio-inspired materials. Source URLs: https://engineering.jhu.edu/news/buil…REDACTED/ ; https://www.computerscience.uchicago.edu/news/remo…REDACTED/ ; https://www.eurekalert.org/news-releases/1137370 (engineering.jhu.edu)

BCI / human-machine boundary

Neuralink has active trial infrastructure beyond one site. Neuralink’s trials page lists active clinical trials and shows sites including Barrow, Cleveland Clinic Abu Dhabi, University Health Network Canada, University of Miami, The Miami Project, and UT Southwestern. Source URL: https://neuralink.com/trials/ (neuralink.com)

Paradromics has crossed a first-human / early-feasibility milestone. Paradromics’ news page says its June 17, 2026 company news covered first Connexus BCI implantation for the FDA-approved Connect-One early feasibility study, evaluating long-term use to restore speech and enable computer control for severe motor impairment. Source URL: https://paradromics.com/news/?e448aeee_page=1 (paradromics.com)

Precision Neuroscience is pursuing a medical-device route through Medtronic integration. Precision announced a January 2026 strategic partnership with Medtronic to integrate Layer 7 BCI technology with Medtronic’s neurosurgical platform. Source URL: https://www.globenewswire.com/news-release/2026/01/12/3217312/0/en/Prec…REDACTED.html (globenewswire.com)

CAPITAL_FLOW_MAP

Major private / public capital movements

  • Apptronik: $520M Series A-X announced Feb. 11, 2026; investors include Google, Mercedes-Benz, John Deere, QIA, AT&T Ventures. Source URL: https://apptronik.com/news-collection/appt…REDACTED (apptronik.com)
  • Skild AI: Series C announced Jan. 14, 2026; TechCrunch reported valuation around $14B. Source URLs: https://www.skild.ai/blogs/series-c ; https://techcrunch.com/2026/01/14/robo…REDACTED/ (skild.ai)
  • Physical Intelligence: reported talks for roughly $1B at $11B+ valuation; not confirmed as closed in sources fetched this run. Source URL: https://techcrunch.com/2026/03/27/phys…REDACTED/ (techcrunch.com)
  • Unitree: STAR Market IPO review passed; expected H1 2026 revenue RMB 1.052B–1.128B. Source URL: https://english.news.cn/20260601/a855…REDACTED/c.html (english.news.cn)
  • SandboxAQ: $500M CHIPS award for AI-driven semiconductor materials discovery. Source URL: https://www.sandboxaq.com/press/sand…REDACTED (sandboxaq.com)
  • DARPA / NSF / U.S. Army / Air Force: public capital is moving through RFI/program channels, UAS procurement marketplaces, CCA architecture/testing, and NSF programmable-lab grants. Source URLs: https://www.darpa.mil/news/2026/rethinking-robotics ; https://www.army.mil/article-amp/291297/us_a…REDACTED ; https://www.af.mil/News/Article-Display/Article/4405471/air-…REDACTED/ ; https://www.computerscience.uchicago.edu/news/remo…REDACTED/ (darpa.mil)

Capital-flow interpretation

Capital is clustering around four layers:

  1. Embodied foundation models / robot brains: Skild AI, Physical Intelligence, Google DeepMind, NVIDIA.
  2. Humanoid bodies + deployments: Figure, Apptronik, Agility, Unitree, Tesla.
  3. Defense autonomy procurement: Anduril, Shield AI, General Atomics, RTX Collins, Army UAS Marketplace vendors.
  4. Model-to-matter / autonomous science: SandboxAQ, Atinary, NSF cloud labs, DARPA physical compute materials.

Highest conviction: robotics infrastructure and defense-autonomy rails. Lower conviction but high upside: individual humanoid OEM winners.

CONVERSION_LAYER

What is converting from hype to deployment?

  • Figure → BMW: strongest humanoid conversion signal because BMW confirms plant deployment / logistics sequencing context. Source URL: https://www.press.bmwgroup.com/global/article/detail/T0458778EN/bmw-…REDACTED (press.bmwgroup.com)
  • Agility → GXO: real logistics RaaS deployment structure, useful as benchmark for robots-as-service economics. Source URL: https://www.agilityrobotics.com/content/gxo-…REDACTED (agilityrobotics.com)
  • Defense drones → marketplace procurement: Army UAS Marketplace and JTAARS contract show acquisition process conversion, not just lab testing. Source URLs: https://www.army.mil/article-amp/291297/us_a…REDACTED ; https://www.army.mil/article-amp/292147/pm_u…REDACTED (army.mil)
  • Materials AI → government-backed validation: SandboxAQ’s award requires virtual screening plus lab validation partners, which matters because pure simulation is not enough. Source URL: https://www.sandboxaq.com/press/sand…REDACTED (sandboxaq.com)

What remains mostly optionality?

  • Tesla Optimus: factory construction / line conversion is significant, but no fetched source confirms broad external customer revenue or independent productivity metrics. Source URLs: https://ir.tesla.com/ ; https://evwire.com/p/tesl…REDACTED (ir.tesla.com)
  • Google Gemini Robotics 2: capability signal; deployment and monetization still early-access / partner-driven. Source URL: https://www.axios.com/2026/07/30/goog…REDACTED (axios.com)
  • BCI: clinical milestone density rising, but commercial reimbursement / durable safety / surgical adoption remain unresolved. Source URLs: https://neuralink.com/trials/ ; https://paradromics.com/news/?e448aeee_page=1 ; https://www.globenewswire.com/news-release/2026/01/12/3217312/0/en/Prec…REDACTED.html (neuralink.com)

TIME_MAP

0–6 months

  • Watch BMW/Figure for factory-hours worked, number of robots deployed, task uptime, safety incidents, labor displacement/substitution metrics, and commercial expansion beyond pilot scope. Source basis: BMW/Figure deployment updates. URLs: https://www.press.bmwgroup.com/global/article/detail/T0458778EN/bmw-…REDACTED ; https://www.figure.ai/news/f-03-at-bmw (press.bmwgroup.com)
  • Watch U.S. Air Force CCA for flight demonstrations, vendor down-selects, FY2026 production decision trajectory, and autonomy safety architecture. Source basis: Air Force CCA updates. URLs: https://www.af.mil/News/Article-Display/Article/4405471/air-…REDACTED/ ; https://www.af.mil/News/Article-Display/Article/4412705/coll…REDACTED/ (af.mil)
  • Watch Unitree IPO pricing / allocation / post-listing valuation as the first large public humanoid-market comp from China. Source URL: https://english.news.cn/20260601/a855…REDACTED/c.html (english.news.cn)

6–18 months

  • Humanoid winners should show repeatable unit economics, not just “more robots.” Key metrics: uptime, MTBF, intervention rate, task cycle time, hand durability, maintenance burden, and customer renewal.
  • Defense autonomy should show open-architecture interoperability and procurement scale through marketplaces / production awards. Source URLs: https://www.army.mil/article-amp/291297/us_a…REDACTED ; https://www.army.mil/article-amp/292147/pm_u…REDACTED (army.mil)
  • AI materials labs should move from grant announcements to closed-loop discovered materials with external industrial validation. Source URLs: https://www.sandboxaq.com/press/sand…REDACTED ; https://www.computerscience.uchicago.edu/news/remo…REDACTED/ (sandboxaq.com)

18–36 months

  • Public-market exposure may consolidate around platform providers: NVIDIA, Google/Alphabet, ABB robotics, defense primes/autonomy providers, and AI-science infrastructure.
  • Private humanoid valuations may compress if deployments do not show labor substitution or safety-certified repeatability.

BOTTLENECK_INDEX

  1. Dexterous manipulation — still the central bottleneck for useful humanoid economics; ABB/PSYONIC collaboration validates the importance of tactile/human-derived manipulation data. Source URL: https://www.abb.com/global/en/news/136690/prsr…REDACTED (abb.com)
  2. Robot data scarcity — Skild explicitly frames the absence of an “Internet of robot data” as a core challenge; NVIDIA/Cosmos and synthetic training stacks attempt to solve this. Source URLs: https://www.globenewswire.com/news-release/2026/03/17/3256839/0/en/skil…REDACTED.html ; https://investor.nvidia.com/news/press-release-details/2026/NVID…REDACTED/ (globenewswire.com)
  3. Sim-to-real reliability — world models help, but the investment signal should follow measured real-world task success, not generated video quality. Source URL: https://blogs.nvidia.com/blog/cosm…REDACTED/ (blogs.nvidia.com)
  4. Hardware durability / maintenance — hands, actuators, batteries, thermal management and field maintenance determine whether robots beat human labor or fixed automation.
  5. Safety and liability — critical for humanoids around workers, CCA weapons integration, and BCI implants. Air Force weapons-integration testing and BCI early feasibility studies are examples of safety gating. Source URLs: https://www.af.mil/News/Article-Display/Article/4412705/coll…REDACTED/ ; https://paradromics.com/news/?e448aeee_page=1 (af.mil)
  6. Commercial proof vs valuation — Skild / Physical Intelligence / humanoid OEM valuations are ahead of visible revenue; keep position sizing disciplined.

THESIS_STATE

Thesis: Physical AI is becoming the next AI infrastructure layer.

Status: strengthened. Evidence: NVIDIA Cosmos/Isaac/GR00T ecosystem, Google Gemini Robotics 2, Skild / PI capital flow, and growing deployment trials. Source URLs: https://nvidianews.nvidia.com/news/nvid…REDACTED ; https://www.axios.com/2026/07/30/goog…REDACTED ; https://www.skild.ai/blogs/series-c (nvidianews.nvidia.com)

Thesis: Humanoid robots are entering commercial sites before they are broadly economical.

Status: strengthened but still early. BMW/Figure and GXO/Agility are real deployment signals; the missing piece is repeatable economics at scale. Source URLs: https://www.press.bmwgroup.com/global/article/detail/T0458778EN/bmw-…REDACTED ; https://www.agilityrobotics.com/content/gxo-…REDACTED (press.bmwgroup.com)

Thesis: Defense autonomy will commercialize faster than general-purpose home humanoids.

Status: strengthened. Defense has procurement urgency, budgets, doctrine and defined mission needs; UAS Marketplace / CCA / JTAARS validate conversion. Source URLs: https://www.army.mil/article-amp/291297/us_a…REDACTED ; https://www.af.mil/News/Article-Display/Article/4405471/air-…REDACTED/ ; https://www.army.mil/article-amp/292147/pm_u…REDACTED (army.mil)

Thesis: Model-to-matter is an investable category, not just research.

Status: strengthened. SandboxAQ CHIPS award and NSF programmable lab grants show sovereign capital moving into AI-driven discovery and autonomous experimentation. Source URLs: https://www.sandboxaq.com/press/sand…REDACTED ; https://engineering.jhu.edu/news/buil…REDACTED/ ; https://www.eurekalert.org/news-releases/1137370 (sandboxaq.com)

Thesis: BCI is strategically important but not yet a liquid public-market thesis.

Status: unchanged / watchlist. Clinical milestones are real, but revenue, reimbursement, durability and risk remain gating. Source URLs: https://neuralink.com/trials/ ; https://paradromics.com/news/?e448aeee_page=1 ; https://www.globenewswire.com/news-release/2026/01/12/3217312/0/en/Prec…REDACTED.html (neuralink.com)

STRATEGIC_INTERPRETATION

The most actionable conclusion is not “buy humanoid hype.” It is:

Physical AI is splitting into three investable layers:

  1. Infrastructure layer: compute, simulation, world models, robot foundation models, synthetic data, edge deployment. This is where public-market exposure is clearest through NVIDIA and Alphabet.
  2. Deployment layer: humanoids/logistics/industrial robotics where proof depends on hours worked, safety, reliability and customer renewals. Figure/BMW and Agility/GXO are the key validation patterns.
  3. Sovereign / defense / model-to-matter layer: defense autonomy and AI materials are getting procurement/grant pathways faster than consumer humanoids are getting broad demand.

The near-term portfolio mistake would be over-weighting charismatic humanoid OEM claims while under-weighting infrastructure and defense procurement rails.

SOURCES

  1. BMW — Figure 03 at BMW Plant Spartanburg: https://www.press.bmwgroup.com/global/article/detail/T0458778EN/bmw-…REDACTED
  2. Figure — F.03 arrives at BMW: https://www.figure.ai/news/f-03-at-bmw
  3. Figure — Helix 02 full-body autonomy: https://www.figure.ai/news/helix-02
  4. NVIDIA — physical AI models / Cosmos / GR00T / Isaac: https://nvidianews.nvidia.com/news/nvid…REDACTED
  5. NVIDIA Investor — robotics leaders / Cosmos / Isaac: https://investor.nvidia.com/news/press-release-details/2026/NVID…REDACTED/
  6. Apptronik — $520M Series A-X: https://apptronik.com/news-collection/appt…REDACTED
  7. Agility Robotics — GXO multi-year agreement: https://www.agilityrobotics.com/content/gxo-…REDACTED
  8. Unitree / Xinhua — STAR Market IPO review: https://english.news.cn/20260601/a855…REDACTED/c.html
  9. Tesla Investor Relations Q2 2026 documents: https://ir.tesla.com/
  10. EVWire summary of Tesla Q2 2026 Optimus line details: https://evwire.com/p/tesl…REDACTED
  11. Skild AI — Series C: https://www.skild.ai/blogs/series-c
  12. TechCrunch — Skild valuation: https://techcrunch.com/2026/01/14/robo…REDACTED/
  13. Physical Intelligence site: https://www.pi.website/?trk=public_post-text
  14. TechCrunch — Physical Intelligence reported funding talks: https://techcrunch.com/2026/03/27/phys…REDACTED/
  15. Axios — Google Gemini Robotics 2: https://www.axios.com/2026/07/30/goog…REDACTED
  16. ABB — PSYONIC dexterity collaboration: https://www.abb.com/global/en/news/136690/prsr…REDACTED
  17. PSYONIC — ABB / NVIDIA robotics collaboration: https://www.psyonic.io/news/pres…REDACTED
  18. Ottobock — Blue Arbor investment: https://corporate.ottobock.com/en/media/newsroom/otto…REDACTED
  19. U.S. Army — UAS Marketplace: https://www.army.mil/article-amp/291297/us_a…REDACTED
  20. U.S. Army — JTAARS production contract: https://www.army.mil/article-amp/292147/pm_u…REDACTED
  21. U.S. Air Force — CCA open architecture / autonomy vendors: https://www.af.mil/News/Article-Display/Article/4405471/air-…REDACTED/
  22. U.S. Air Force — CCA weapons integration testing: https://www.af.mil/News/Article-Display/Article/4412705/coll…REDACTED/
  23. Shield AI — CCA mission autonomy provider: https://shield.ai/shie…REDACTED/
  24. DARPA — rethinking robotics with physical intelligence: https://www.darpa.mil/news/2026/rethinking-robotics
  25. DARPA RFI — materials for physical compute: https://www.darpa.mil/work/-with-us/opportunities/darpa-sn-26-76
  26. SandboxAQ — $500M CHIPS award: https://www.sandboxaq.com/press/sand…REDACTED
  27. Johns Hopkins — $20M AI-powered materials lab network: https://engineering.jhu.edu/news/buil…REDACTED/
  28. UChicago / Argonne — PoLARIS robotic lab: https://www.computerscience.uchicago.edu/news/remo…REDACTED/
  29. CUNY / EurekAlert — AI-BIOME NSF award: https://www.eurekalert.org/news-releases/1137370
  30. Neuralink trials: https://neuralink.com/trials/
  31. Paradromics news / first Connexus implantation: https://paradromics.com/news/?e448aeee_page=1
  32. Precision Neuroscience / Medtronic partnership: https://www.globenewswire.com/news-release/2026/01/12/3217312/0/en/Prec…REDACTED.html
Source in the house: eth-memory/sol/reports/2026-08-02-physical-ai-model-to-matter.md☉ Sol
← Previous · 2026-08-01 — Task 1 + Task 2: Physical AI / Model → Matter ScanNext · 2026-08-03 — Task 1 / Task 2: Physical AI / Embodiment Scan + Et →