Associate
Every observation is checked: do time, route and appearance fit a known person, vehicle or object? If the evidence is ambiguous, several assignments remain side by side, each with its probability.
Cameras, drones, sensors and field reports deliver more observations than ever. Detecting people and vehicles is largely solved. Linking the detections of many sources into a reliable situational picture is still done by people.
Polaris is meant to take over that layer. It assigns every observation to real people, vehicles and objects, derives relationships and events, and corrects its conclusions as soon as new information contradicts them.
During the first outdoor flight test, a stationary camera and, shortly after, the drone each detected a person. Whether it was the same person, only the operator could judge. Read the report
Order from chaos.
Connect
People, vehicles and objects turn into relationships, groups and events. Patterns that only emerge across several sources or over longer periods become visible, such as a vehicle that keeps reappearing at different gates.
Revise
When a later observation contradicts an assignment, Polaris corrects it together with every conclusion built on it. Early mistakes do not slip unnoticed into decisions.
When the picture changes
Principles
Uncertainty made visible
Every assignment, relationship and event carries a probability. Polaris never pretends to know more than it does.
Back to the source
Every assessment leads to the observations it rests on. Every correction stays traceable in the history.
People decide
Polaris assesses, the operator decides on operational consequences. Machine assessment and human decision are stored separately.
On your infrastructure
Runs entirely on the operator's own infrastructure, independent of the sensor vendor.
A layer between sensors and decisions
Polaris builds on existing sensors. Cameras, drones and systems stay as they are. What Polaris derives flows on as a single datafeed: people, vehicles, objects, relationships and events, each with its probability and provenance.
Shape the project with us
For emergency services, critical infrastructure operators and integrators who know the problem from operations.
