Inception gave people an image that is hard to forget: a city folding onto itself without cuts. Orbify.eu, a two-year-old Norwegian startup, wants to bring that distortion into navigation systems. The company, co-founded by Karsten Huitfeldt, a former Snøhetta architect and project manager until 2025, is developing image processing software that applies a curvature to a 3D map model. In practice, the scene moves from a near-field street-level perspective to an overhead view while remaining on a single continuous surface.

Orbify has filed a European patent application (PCT/EP2026/058725), which has not yet been published. It is therefore unclear which claims are at stake and whether earlier implementations of curved projections or non-linear cameras could be considered prior art by patent examiners. Huitfeldt says the company wants to protect the bending algorithm, the shader and camera setup, and their use in a moving navigation context with overlaid route information. The application is in its second examination round at the European patent office.

The value is not the bend, but the change of context

Orbify's thesis is not aesthetic. Many navigation interfaces today alternate between different views, each with its own logic: a close-up, a traditional map, an aerial view. Users have to mentally reconstruct the relationship between what is ahead of them and what the map shows from above. Orbify's approach tries to remove that jump by progressively curving the scene. In cognitive terms, the promise is less mental work and the removal of three or four fragmented systems. It is not just smoother graphics: it is a change in interface architecture, with potential effects for HMI designers and for teams integrating 3D maps into embedded platforms.

The demo is a concrete technical step. It runs in the browser with PlayCanvas Engine, an open source JavaScript engine that uses WebGL2 and WebGPU; it shows Gaussian splatting in a spatial interface. Earlier prototypes were built in Unreal Engine and Unity, using Google Photorealistic 3D Tiles through Cesium. That detail says a lot about the direction: Orbify does not want to build a proprietary engine or a mapping ecosystem from scratch, but to offer itself as a platform-agnostic software/SDK layer, from web to mobile to automotive systems. The bet is on controlling rendering and interaction, not on owning geographic data.

Research before superiority

One detail separates this story from many navigation demos: Orbify has defined a controlled study protocol to compare its interface with a conventional perspective plus an aerial display, but it has not yet recruited participants. Huitfeldt is explicit: they are not claiming superiority ahead of the evidence. In a sector where interfaces are often validated with engagement metrics or completion times, the choice to subject commercial claims to a study is a sign of maturity. But it is also the main bottleneck: without evidence, the value of the patent and the willingness of engine and rendering partners remain hypotheses to test.

The patent application is not yet published, so the scope of protection remains uncertain. Initial funding comes from Innovation Norway and the Research Council of Norway. Orbify, not to be confused with the also Oslo-based Orbify.ai, sits between a working prototype and a commercial product. The technology is publicly demonstrable, but the move from visual effect to navigation tool depends on a study that has yet to begin. It is a story of real-time rendering and embedded interfaces: not about cloud, but about control of the frame the user sees.