Vision Sensor
See. Understand. React.
The nanoblue Vision Sensor brings edge AI into your spaces: it detects people, movement and gestures directly on the device – without images ever leaving the room.
AI at the edge instead of video in the cloud
With a dedicated NPU, the Vision Sensor processes image data right where it is captured: inside the sensor. Only events and numbers enter the network – people count, attention, detected gesture – never the camera image. That makes it GDPR-friendly and light on bandwidth.
Whether visitor counting in a signage network, presence detection in a meeting room or touchless gesture control in a showroom: the models run locally and can be tailored to the use case. That makes the Vision Sensor the ideal basis for intelligent signage and interactive signage – content that responds to the audience and to gestures.
What's inside
Edge AI
Neural networks run directly on the sensor – with a dedicated NPU for real-time detection.
Privacy by design
Images never leave the device. Only anonymous event data is transmitted.
Signage analytics
Audience count, dwell time and attention as metrics for your content.
Gesture control
Touchless operation of rooms and content – a wave is enough.
Room sensing
Presence and occupancy as triggers for media control and Room Analytics.
Open events
All detections are available to the entire platform as MQTT events.
At a glance
| AI performance | Dedicated NPU with 3 TOPS, quad-core ARM, 2 GB RAM |
|---|---|
| Camera | 4K30 |
| Connectivity | Gigabit Ethernet with PoE, Wi-Fi, Bluetooth |
| Events | REST, MQTT, WebSocket – detections as variables in the function logic |
| Processing | On-device, no image transmission |
