
Sees the world the way people do
An ordinary frame gives the color and texture of the scene: where the path is, where there's an obstacle, where there's a person. A single frame can't tell distance, so the camera works alongside the other sensors.
Everything here is real data from the onboard systems: lidar recordings, route tracks and precise telemetry.
The robot reads its surroundings three ways at once: camera, depth map and lidar. Our algorithm fuses the data right on board, so the machine needs no internet connection or server to drive. Below is a single moment in the life of a robot in an orchard - a tough environment where everything keeps changing.

An ordinary frame gives the color and texture of the scene: where the path is, where there's an obstacle, where there's a person. A single frame can't tell distance, so the camera works alongside the other sensors.

Gives a distance for every pixel in the frame. That helps the robot tell thick grass from solid obstacles and notice objects right up against it.

Scans a full 360 degrees. Lidar picks out walls, posts and the edges of the path even in total darkness or with the sun straight ahead.
Collision avoidance is always on, whatever the machine is doing. A dedicated algorithm polls the sensors continuously and stops the robot the instant it detects danger.

Our Android app for running the machine. It is where you assign jobs, follow progress and review the results of the shift. The robot stays in constant contact with the tablet - if the signal drops, the machine comes to a safe stop.

Describe the job and the conditions, and we'll tell you whether a standard machine will handle it or it calls for custom work.