Join the movement and participate on Uniswap.
Participate
Auki ran its first community update live on X to improve screen sharing and demo quality. The session focused on the new open-source SDK, major gains in robot scanning speed, cleaner depth sensing, and a significant new inbound RFQ from a large Asian retailer.
The team demonstrated the new open-source Auki SDK treating a phone exactly like a robot in the domain. JD walked through a store while the store manager viewed his live camera feed with an AR overlay showing the AI’s understanding of the environment (purple tags on products).
Any device with the right permissions — robot, glasses, or phone — can now share sensor data and appear in the shared spatial model. The AR view is not just the raw camera feed; it shows the world as the AI sees it.
The demo also included Jason’s 3D shelf visualization pulled from product dimension data, with a 2D planogram-style view for users more comfortable with traditional shelf planning tools. Editing the digital information layer on top of the physical store is now possible in real time.
The SDK is public and anyone can try it. It handles the networking and sensor sharing so developers can treat phones as additional participants in the domain.
Scanning performance continues to improve rapidly:
These gains come from tighter integration between perception, planning, and execution.
The computer vision lab ran experiments fusing raw depth sensor data with monocular depth estimation from Depth Anything 3. Pure depth sensors are noisy and can cause collisions or awkward end-effector movements. Pure monocular depth lacks metric scale.
The fused approach produces results that remain metrically accurate while looking almost as clean as the monocular estimates. Early visuals show a dramatic reduction in noise while preserving real-world distances.
Phil designed a very low-profile gripper because the standard Galbot arms are too large to operate effectively inside retail shelves. The team is filing a patent and therefore cannot show the prototype yet.
The team used the public Understand Anything tool to generate a knowledge graph and guided tours for the Auki SDK. New developers or contributors can walk through the tours to quickly understand the repo’s structure and purpose. Agents can also query the knowledge graph directly instead of reading every file.
The generated tour is live at: https://aukilabs.github.io/auki-sdk/
A very large Asian retailer with 800 stores in one country has issued an RFQ after seeing BracketBot work shared by a partner. This would be Auki’s first major robotics deployment in Asia and one of its larger implementations to date.
The Galbot is too big for this particular use case, so the pilot is expected to use BracketBot. Negotiations are now underway.
The switch to live X streaming gave better screen-sharing control and higher-quality demos than streaming from Discord. The weekly Discord hangout will continue after each update for off-record discussion.
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
