Niantic Spatial and Spexi Geospatial have formed a strategic partnership that pairs Spexi's decentralised drone-imagery network with Niantic Spatial's reconstruction technology, aiming to convert raw aerial captures into city-scale 3D intelligence for physical AI. The deal makes Spexi a preferred drone imagery provider for training Niantic Spatial's real-world foundation models.
A Drone-to-3D Pipeline
Under the agreement, customers can commission drone captures through Spexi and route the imagery through Niantic Spatial's Reconstruction API to generate high-fidelity, geometrically accurate 3D models in the form of 3D Gaussian splats. The outputs are viewable inside a Niantic Spatial Viewer and measurement tool embedded in the Spexi World platform, giving enterprise teams a single workflow from raw captures to actionable 3D intelligence.
From One Building to a Whole City
"For physical AI to work in the real world, it needs a foundation grounded in reality," Niantic Spatial CEO Inhi Cho Suh said. "Combining Spexi's capture network with our reconstruction technology and real-world models gets us significantly closer to that." The companies say the collaboration is designed to push 3D reconstruction from individual objects and buildings to entire urban environments, a step change for digital twins, simulation pipelines and physical AI training data.
10,000 Pilots, Six Million Acres
Spexi operates a drone imagery network of more than 10,000 pilots and has mapped over six million acres at roughly 2.8 cm resolution, an order of magnitude sharper than satellite imagery. Its standardised autonomous flight protocols are specifically tuned for machine-learning workflows, which the company says lets Niantic Spatial calibrate its reconstruction pipeline directly against Spexi's capture standards. "Together, Spexi and Niantic Spatial deliver a drone-to-3D pipeline that will redefine the next generation of physical AI," said Spexi CEO Bill Lakeland.
Use Cases Beyond Gaming
Niantic Spatial said the joint pipeline targets infrastructure inspection, insurance risk assessment, energy site analysis, asset management, 3D measurement and simulation. That puts the duo squarely in front of warehouse, factory and construction customers already buying spatial AI for ground-level applications, such as Slamcore's recent $14M raise from Rockwell to scale visual AI in warehouses and Torc Robotics' physical-AI partnership with Mila for autonomous trucking.
Training Data Is the New Moat
The Niantic-Spexi alliance also reflects a broader race to build real-world training corpora for physical AI. Vision-language-action models are increasingly bottlenecked by the quantity and fidelity of grounded real-world data, a theme echoed at top venues this year including CVPR 2026, where Apple is set to present 14 computer-vision papers. Drone-derived 3D scenes give foundation-model developers an inexpensive way to scale spatial ground truth without standing up bespoke capture fleets.
What Comes Next
The companies did not disclose financial terms but said the integrated pipeline is available now via the Spexi World platform. With both teams pushing toward city-scale reconstructions, the partnership is a signal that the physical-AI stack is maturing from research demo to production pipeline.
Reporting based on coverage from AI Insider, DRONELIFE, GamesBeat and Sensors & Systems.