Spranic builds spatial memory layer for physical AI

Spranic's pitch rests on a gap it sees in physical AI: machines still struggle to adapt to changing environments, and founder Jordi Spranger argues reusable spatial context can reduce the task-specific planning that robot deployments currently require. He compares the need to "Google Maps on steroids" so robots can replan across many tasks rather than follow a statically planned one.

Beyond navigation, Spranger ties flexibility to provenance — knowing where a robot's information comes from. Knowledge graphs, in his telling, are well suited to connecting domains and preserving that origin data, which he links to auditable and safe robot behavior. Spranic's approach appears aimed at giving physical AI the same kind of context digital agents already use, though it remains a build rather than a shipped product.