Scouting Robots in Beijing and Paying the Community to Scan the World

Back in Beijing

This week’s community update comes from Beijing. Nils is back for the first time since COVID, returning to a city where he lived for seven years.

The trip has two main purposes:

• Lecture at Peking University – A session with MBA students about what we are building and where robotics is heading.

• Robot scouting – Hands-on time with two Chinese semi-humanoids we’re very interested in:
  • Robotera Q5
  • Galbot G1

Galbot is one of Nils’ favorite Chinese robot companies, known for impressive live demos. Roboterra is newer to us in person, but looks promising on paper.

Securing 500 semi-humanoids for 2026

Our 2026 goal has become pretty concrete: roughly 500 semi-humanoid robots doing real work in grocery stores.

That means hardware budgeting is getting real too:

• Expected cost per robot: USD 40–70k
• For the Q5, at volume (100+ units), we already have offers just under 40k per robot

To pull this off, we’ve been talking with Taiwanese banks and financial institutions about credit lines specifically to finance these deployments.

The contrast with Europe is stark. “It would certainly not be easy for me to get funding to do this from a Swedish bank,” Nils says. “Really excited to be in Asia where people believe in the robots.”

Between Taiwan’s financiers and China’s robot OEMs, Asia is the natural launchpad for scaling our robot fleet.

Hosting a European Fortune 500 in Hong Kong

While Nils has been on the road, the Hong Kong lab has been busy too. This week we:

• Hosted a delegation of one of Europe’s Fortune 500 companies
  • Roughly 50 people visiting Level 10
  • Market cap around $7B
• Ran a joint event in the lab showcasing:
  • Our own tech and demos
  • HKRITA (a Hong Kong research institute focused on textiles and related tech)
  • Terragrit, a startup working in the same ecosystem

The takeaway: it's nice to have reached the stage where large, traditional enterprises are flying teams halfway across the world to see how our network, AI Copilots and robots fit into their operations.

December crunch: signatures, rollouts and a year in review

End of year is always chaotic, and this one is no exception.

The big one: 17 December

Next stop for Nils is Europe, where he’ll:

• Visit our current largest customer, and
• Meet the C-level of our next major customer to secure the final signature on a large deployment.

“Hopefully December 17th is going to be a very important day for us… our last missing signature on a very important deal.”

If that signature lands, there’s a real chance we go into 2026 already profitable.

Given where we were a year ago, that’s a huge shift in our position and negotiating power.

New business that can’t wait

On top of that, we’re seeing new deals in Asia come in with a strong sense of urgency:

• Prospects that want to start right away, still in December
• More locations to roll out before year-end

Good news for the company, slightly worse news for anyone dreaming of a long, quiet Christmas. “People were hoping to take more time off for Christmas this year, and it looks like we’re going to be very busy for Christmas rolling out these new locations.”

Year-in-review: 19 December

We’ve finally committed to a date for a proper year-in-review community update:

• December 19th – after the Europe trip and (hopefully) the 17th signature
• Nils will be in Europe, which also gives a friendlier time zone for many community members

“It’s been a crazy year… we are just in such a different position now than we were a year ago.” Expect that session to cover the arc from “are we crazy?” to “this might actually work,” plus where we are heading next.

A new rendering layer: Gaussian scenes for domains

We also got a quick preview of a big technical step: Gaussian scene reconstruction as the rendering layer for domains.

Until now, many of our demos showed point clouds – useful for robots, but not exactly human-friendly. The new work lets you:

• Move around freely inside a 3D reconstruction
• Generated from our DMT recordings
• With much more realistic, continuous visuals than a point cloud

The goal is to let anyone:

• Scan their apartment, office, or store,
• Turn it into a navigable 3D scene, and
• Do all of this before the end of the year.

“This is the upcoming rendering layer of the domain… our goal is for you guys to be able to go play with this before the end of the year and set this up.”

This ties directly into the community initiative.

Paying the community to build the world’s largest open scene dataset

We’re planning to let the community create domains and sell them to us for $AUKI tokens, under a clear rule:

• You scan and create a domain
• You can view it online, share it, and
• If you choose, sell that domain to us for tokens
• We then open source it as part of a public dataset

The ambition is bold but realistic: “Hopefully, as a community, we can come together and create the world’s largest open source data set of scanned environments so that people can do real-to-sim-to-real pipelines.”

Why this matters:

• AI and robotics labs are hungry for high-quality real-world 3D data
• Our network already has:
  • A motivated community
  • A growing node infrastructure
  • A treasury to reward contributors
• If we execute, this can:
  • Cement Auki as a core data source for embodied AI research
  • Deepen our relationships with labs and companies building physical AI

It is also a very clean example of what we always wanted the token to be used for: rewarding real work in the physical world that strengthens the network.

Intercognitive meetups and 2026 vibes

Finally, a quick note on Intercognitive, the space where AI, robotics, spatial computing, and human cognition overlap.

• We recently hosted an Intercognitive meetup in Shenzhen, connecting people from robotics, computer vision, and AI.
• Next up is a Beijing Intercognitive meetup (this Thursday), during Nils’ current visit.

“I think that 2026 is going to be a great year not just for us but also for Intercognitive.”

As always, the community update ended with the usual split:

• A public part streamed to X, and
• A private, off-the-record conversation on Discord, where we talk more openly about what’s happening and what’s next.

If you want to be in the room when those conversations happen, you know where to find us.

About Auki

Auki is making the physical world accessible to AI by building the real world web: a way for robots and digital devices like smart glasses and phones to browse, navigate, and search physical locations.

70% of the world economy is still tied to physical locations and labor, so making the physical world accessible to AI represents a 3X increase in the TAM of AI in general. Auki's goal is to become the decentralized nervous system of AI in the physical world, providing collaborative spatial reasoning for the next 100bn devices on Earth and beyond.

X | Discord | LinkedIn | YouTube | Whitepaper | auki.com