Spatial intelligence
The subject of study is the legibility of spaces, to devices, and to the people for whom legibility is the precondition of independence.
The world, however, is dimensional
It has distance, velocity, weight, light, friction, texture, motion, orientation, presence, and consequence. To understand it, intelligence must learn to perceive space.
Where the inquiry is taking shape
Field notes and essays on computation becoming situated in the world, measured, remembered, and met at the place it concerns
The interface leaves the screen
Computation met at the place it concerns, present because it knows where it is.
Read →Spatial memory and the return to place
Relocalization, anchors, and the difference between an overlay and a layer that lasts.
Read →Measurement carries responsibility
Distance, clearance, closing speed, the numbers a system stakes an action on.
Read →The research corpus
The corpus behind the field, filed by domain and open to search.
Browse →Built to remain accountable to reality
Reality before representation
Voir begins with the physical field itself: the surface, the boundary, the body, the motion, the measurement, the constraint
The digital layer must remain accountable to reality.
Measurement carries responsibility
To measure is to claim precision; to guide action is to enter consequence. Spatial intelligence must be built with humility, knowing when it is certain, when it is approximating, and when uncertainty should stay visible. In autonomous driving, LiDAR already proved the point: it turned the environment into a measurable field of lanes, curbs, vehicles, and clearance.
A system that measures poorly does not merely err. It weakens trust in the intelligence itself.
Intelligence must become situated
An answer changes when the environment changes. A recommendation changes when the user is standing in a room, crossing a street, lifting a weight, repairing a machine, entering a hospital, or navigating a city.
Context is not metadata. Context is the field.
The body is part of the interface
Human intelligence is embodied. We reach, turn, walk, point, look, balance, hesitate, and move through space. Spatial systems must understand the body not as an input device, but as a living participant in the environment.
The future is mixed
The physical world will not disappear into screens. A more powerful transformation is the deepening relationship between computation and reality. Voir studies the mixed field where physical and digital systems begin to cooperate.
Seven doors into one inquiry
Environmental Understanding
We study how machines perceive rooms, buildings, streets, landscapes, and shared spaces. This includes object detection, scene segmentation, room reconstruction, surface understanding, depth estimation, mapping, relocalization, and persistent spatial memory.
The aim is to understand the environment as a structured field of meaning, constraint, motion, and possibility.
Enter →Embodied Intelligence
We study how intelligent systems reason from the perspective of an agent situated in space. This includes movement, orientation, reachability, affordances, navigation, physical constraints, body pose, and human-object interaction.
A spatial system must know what exists, what can be done, and what should be avoided.
Enter →Augmented Reality Systems
We investigate how digital objects, instructions, measurements, simulations, and annotations can be placed into the physical world with coherence and persistence. We are interested in AR as a new layer of cognition: perception made visible, memory made spatial, instruction placed directly into the field of action.
Enter →Spatial Memory and Persistence
We study how places are remembered across sessions, devices, users, and time. A useful spatial system recognizes when it has returned to the same location, what has changed, and what should remain anchored.
Persistence is the difference between a temporary overlay and a meaningful spatial layer.
Enter →Measurement and Verification
We study how devices estimate distance, scale, height, motion, position, alignment, and change. Measurement connects computation to trust. It is the bridge between perception and evidence.
When intelligence enters the physical world, it must know how much confidence its measurements deserve.
Enter →World Models and Simulation
We study how machines form internal models of environments, not merely as visual scenes, but as dynamic systems with objects, forces, constraints, paths, probabilities, and futures.
The world is not a still image. It is a field of possible events.
Enter →Human Spatial Experience
We study how people perceive space, remember place, navigate environments, make decisions, and experience presence. Spatial intelligence must serve human perception without overwhelming it. The best interface is not always the one that shows the most. It is often the one that reveals what matters.
Enter →How does intelligence become present in the world?
- Not as spectacle.
- Not as a labeling machine.
- As perception, memory, measurement, embodiment, and action.
The necessary foundations are converging
Cameras, depth sensors, neural perception, AR anchoring, robotics, world models, the pieces are forming a single field
Spatial intelligence integrates these capacities into coherent perception and action. The next era of computing will not be defined only by smarter models. It will be defined by models that understand where they are.
Advanced by more than engineers
Those who move between science and experience, measurement and meaning, computation and embodiment.
Collaborate →Signals from the field
Essays, field notes, and a research corpus on the state of spatial intelligence.
Read the Ledger →An inquiry into how computation becomes situated in the world
Not built around a single product. Spatial intelligence belongs to the future of computing itself, to how machines perceive reality, and how digital systems become situated in the physical world.