
G120
Widest field of view
sense
The perception layer of Physical AI.
XINK SENSE gives machines the ability to perceive the real world through 3D vision, depth sensing, and spatial awareness — a full lineup of stereo depth cameras built on eYs3D's own silicon.

Four form factors, one depth engine — pick by range, field of view, and environment. Naming rule: XINK + Function + Series.

Widest field of view

The SLAM workhorse

Rugged & cost-effective

Smallest footprint
Every decision a machine makes starts with what it can perceive. XINK SENSE captures dense, accurate spatial data in real time — stereo and ToF depth that turns a flat camera feed into geometry a robot can plan against.

Depth on SENSE cameras is computed by eYs3D's own eSP-series image processors — a proprietary stereo depth engine refined across seven IC generations, paired with active IR dot projectors for texture-poor scenes. Depth arrives as a hardware output, with no host GPU required.

SENSE output flows straight into XINK CORE for on-device reasoning, and registers with the platform for one-click install, OTA updates, and remote debugging over WebRTC. Perception is the first stage of the loop — never a dead end.
| ToF module | Hybrid ToF sensing for precision depth at range |
|---|---|
| Multi-camera | Seven-sensor bridge ICs synchronize multi-camera perception rigs |
| eCV5546 | Edge vision module with on-board AI acceleration |
| Interfaces | USB / MIPI; SDK for Windows, Linux, Python, ROS |
Register in minutes, install apps and models over the air, and control remotely over WebRTC — SENSE devices are fleet citizens from day one.
XINK CONNECT