Vangrid, an Amsterdam-based startup turning ordinary smartphones into a crowdsourced network for capturing spatial data, announced in early August it secured $9 million in seed funding led by HashKey Capital. The round drew backing from Borderless Capital, Crypto.com Capital, Animoca Brands, Gate Labs and Mapleblock Capital.
The company's pitch centers on a straightforward idea: recruit everyday users to film their surroundings, then stitch those clips into three-dimensional models that autonomous robots and physical AI systems can use to navigate the real world. Each capture gets a cryptographic fingerprint stored on Coinbase's Base blockchain, creating what the startup describes as a verifiable audit trail for spatial data.
Robert Brighton, who joined as CEO, told The Block the round closed through a simple agreement for future tokens and ran across several tranches over recent months. He declined to share the company's valuation or specific closure dates. Vangrid waited to announce the funding, he said, until its capture network was live and demonstrated early signs of traction. The startup reported it had logged nearly 100,000 verified captures, 200,000 grid events and more than 1,300 batch attestations by the time of the announcement. A team of 14 runs the operation.
How it works in practice
Contributors download Vangrid's Android app—available since late summer with over 1,000 downloads on Google Play—or access a web interface to record short videos of physical locations. When buyers post bounties for specific areas, users who capture those spots earn rewards. Vangrid reconstructs the footage into spatial models, blurring faces and license plates on the device itself to address privacy concerns. Each capture gets a fingerprint at the moment of recording; those fingerprints are batched into Merkle trees and anchored on Base using the Ethereum Attestation Service. Buyers can verify the provenance through Vangrid's public explorer.
"Contributors capture real-world locations on their smartphones, we reconstruct them into 3D spatial models and verify everything onchain," Brighton said. The company is working on an iOS version and offers an enterprise spatial API for clients who need bulk data access.
The market opportunity, according to Vangrid
Brighton positions the network as infrastructure for companies building robots and autonomous systems that operate in environments where satellite imagery or static maps prove inadequate. He said Vangrid has been in discussions with defense contractors and autonomous-systems developers, though he did not name them. The new capital will fund deeper partnerships with those clients, expand the startup's edge-computation pipeline so more processing happens locally, and build network density across geographies.
"Without a decentralised perception grid, billion-dollar robots will remain blind to dynamic human environments," Brighton wrote in an interview published by Tech.eu. The company frames its addressable market as builders of physical AI—robots, autonomous agents, digital twins—and government or defense users working in GPS-denied or rapidly changing terrain.
Brighton described the investor roster as a blend of financial backers and strategic partners who contributed infrastructure, technology and exchange relationships beyond capital. Gate Labs, for instance, integrated Vangrid into Gate Wallet in late summer, according to posts from both companies.
A crowded field, with different approaches
Vangrid enters a decentralized physical infrastructure mapping category that Hivemapper has dominated by incentivizing drivers to mount dashcams and collect street-level imagery. NATIX Network, which operated a smartphone-based Drive& app before shutting it down, has since repositioned around VX360, a multi-camera offering. DIMO runs a vehicle data layer focused on drive-to-earn mechanics but does not reconstruct 3D spatial models. Vangrid's bounty system differs from always-on dashcam collection; buyers request specific locations and contributors respond, creating a more targeted, on-demand model.
A Crunchbase News analysis published recently noted rising investor interest in "physical and spatial intelligence" startups, alongside broader robotics and defense funding. CB Insights, in a market map of the physical-AI sector, observed that control of real-world robotic datasets creates defensibility for infrastructure providers—a dynamic Vangrid appears to be betting on.
Founder backgrounds and corporate structure
Khalid Benrafik founded the company and serves as CTO; Brighton joined as co-founder and CEO. Vangrid operates from Amsterdam and lists a developer entity, Aurevion Group Ltd, in the British Virgin Islands on its Google Play page. The startup's website footer shows a copyright for Vangrid Inc but does not specify jurisdiction, an organizational structure that's not uncommon among blockchain-oriented ventures but leaves some questions about regulatory domicile.
The company said it plans to launch a token later this year with reward mechanics tied to capture contributions. It has not released a whitepaper or set a token-generation date. Brighton said the near-term focus is growing the contributor base to improve coverage density and proving demand from enterprise customers testing the spatial API. "We are the eyes for the next generation of AI," he wrote—a lofty promise that the market will test as physical AI moves from labs into the field.
