I’ve been sitting with XMAQUINA’s July digest for a few days now and honestly, this month clarified something I’ve been chewing on since TGE.
The humanoid robotics thesis isn’t a story anymore. It’s a spreadsheet.
$8.6 billion has been raised across the sector so far in 2026. That’s 1.8x the entire industry’s 2025 total, and we’re only halfway through the year. Five years ago the whole category raised $61M for a full year. Now individual companies are pulling that in every few weeks.
I remember when robotics was the sector everyone talked about but nobody funded. That era is done.
Let me walk through what actually caught my eye this month, and why I think this is the moment XMAQUINA’s positioning starts to matter differently than it did at launch.
The industry is fully in commercial deployment mode
The rounds that hit in July weren’t research grants. They were commercial scale-up rounds tied to real deployments.
- Humanoid raised $152M at a $1.35B valuation, making them Europe’s first humanoid unicorn. But the number I actually care about is the Schaeffler deal — a four-digit robot fleet across manufacturing by 2032, plus Schaeffler becoming their preferred actuator supplier. That’s not a pilot. That’s a supply chain integration.
- Walden Robotics came out of stealth with $300M co-led by Toyota and Deviation Capital. Their robots went from pilot to production work inside Toyota facilities in under two months. NVIDIA, Boeing Ventures, and Samsung joined the round. When Toyota is both an investor and a customer, that’s the signal.
- FieldAI crossed $100M in revenue and customer commitments in under three years. 30+ enterprise customers across construction, mining, energy, and data centers. FieldAI is worth watching closely because they’re playing the intelligence-layer game — foundation models for existing machines, not building their own hardware. Different bet, but a compounding one.
- Holiday Robotics raised $105M for FRIDAY, one of the biggest Korean humanoid rounds to date. Factory floor focus.
- Figure hit four years old, confirmed the F.04 design cycle is complete, and is moving toward its next production robot. Three humanoid generations, BMW and UPS commercial deals, and one of the largest humanoid manufacturing facilities in the industry. Figure is quietly executing at a pace most projects only talk about.
- 1X dropped a new tendon-driven hand for NEO — 25 degrees of freedom, nearing human-level dexterity. And they’re targeting up to 10,000 robots in production this year. Ten thousand. Real units, shipping to real customers.
Then there’s ATOMS
Travis Kalanick’s new physical AI company just raised $1.7B led by a16z, with Ben Horowitz taking a board seat.
The framing is fascinating — Kalanick’s calling it the completion of the “bits to atoms” story he started at Uber. The vision is atoms-based computers for mining, construction, transport, and food production. Uber becomes a shareholder again after nearly a decade.
I don’t have a strong take on whether ATOMS specifically wins or loses. But $1.7B into a physical-AI thesis from a16z with Kalanick at the helm tells you where serious capital thinks the frontier is. It’s not chatbots anymore.
Why XMAQUINA’s angle keeps making more sense to me
Here’s the thing that keeps bringing me back to this project.
Everyone reading this can see the same headlines I’m seeing. Everyone knows Figure, 1X, Apptronik, Agility are the names to watch. Everyone knows humanoid robotics is going to be one of the defining sectors of the next decade.
What almost nobody has is a way to actually own it before IPO.
Figure’s valuation went from $2.6B (Feb 2024 Series B) to $39B (Sept 2025 Series C). Apptronik went from ~$1.75B to $5.5B+ in a single funding cycle. 1X reportedly targeting $10B+ from ~$820M in eighteen months.
If you tried to buy exposure to these companies at those valuations, you couldn’t. The rounds are oversubscribed, the cap tables are gated behind venture funds, and secondary markets for private equity are illiquid and expensive.
XMAQUINA is one of the only ways to hold governance-weight exposure to that basket without being a licensed institutional investor. Apptronik, Figure AI, 1X, Agility Robotics, and Neura are already in the treasury via SPVs held by regulated third parties. And now with BOT-13 passing at 99%+ approval, Gecko Robotics gets added — AI-powered robotic inspection systems for energy, manufacturing, and defense infrastructure.
Every one of those companies is compounding. And every proposal that passes expands the basket.
The BOT-13 vote itself is a signal
99%+ approval on a real capital allocation. This isn’t performative governance. The DAO is functionally moving.
Gecko Robotics is a smart add — they’re not competing with the humanoid platform names, they’re playing the inspection-and-monitoring layer. Different segment of the robotics value chain, different revenue model, real enterprise customers. Portfolio construction that respects that “robotics” isn’t one bet, it’s a category of bets.
The Execution Engine now handles the actual deployment. The community voted, the mechanism runs, the position gets built.
My honest read
I’ve been through the post-TGE volatility. I’ve watched the price action, engaged with the confusion around vesting and the Humanoid Card mint rules, and pushed back on the FUD when it deserved pushing back on.
None of that changes what actually matters here.
The robotics sector is entering the commercial deployment phase. The capital is flowing. The companies XMAQUINA already holds are compounding on paper markups that would take years to unwind through IPOs. The DAO is executing new allocations through a governance process that actually functions.
If you believe humanoids and physical AI become one of the defining sectors of the 2026 to 2035 decade — and I do, more strongly every month — then having governance-weight exposure to a curated basket of the leaders through a DAO you can actually participate in is a genuinely rare positioning.
Most people are either watching the sector from the sidelines or trying to buy exposure through the wrong tickers. XMAQUINA holders are already positioned.
That’s what July confirmed for me.
We’re early.
nfa, dyor as always — this is my read, not advice. everything I mentioned is public: humanoid.ai, waldenrobotics.com, atoms.co, figure.ai, 1x.tech, fieldai.com, holiday-robotics.com, geckorobotics.com, and dao.xmaquina.io. do your own homework.
