← Back to all projects
Bottle sorting with computer vision
The need
« I need to automatically sort the bottles moving along a conveyor belt »
A camera and a servo motor that sort every bottle by type
Vento connects the camera, trigger sensor, Jetson computer and servo motor on a conveyor. The vision model runs beside the machine, classifies each bottle as A, B or C and tells the servo to divert it towards the corresponding exit. The dashboard records how many bottles pass through each category.
How Vento helps
- 01Connects the camera, trigger sensor, Jetson computer and servo motor
- 02Runs the vision model beside the conveyor without sending every image to the cloud
- 03Turns each classification into a servo position: left, centre or right
- 04Records the count for each bottle type in a dashboard
How it works
01
Connect the camera and sensor
The sensor detects each bottle and triggers the capture as it reaches the sorting area.
02
Run computer vision locally
The Jetson processes the image beside the machine and classifies the bottle as type A, B or C.
03
Sort each bottle with the servo
Vento turns the result into a servo position and updates the corresponding count.
What's inside
- 01Camera over the conveyor
- 02Vision model running locally
- 03Jetson-class local computer
- 04Three-way servo diverter
- 05Belt trigger sensor
- 06Live count dashboard
Highlights
- Classifies on the line itself, no cloud round-trip
- Three outputs: type A, type B and type C
- The model runs on an edge computer beside the camera
- A reusable pattern for any sorting line
Notes from the maker
« The challenge was turning each prediction into a physical action and testing the complete cycle on the conveyor. With Vento, we connected the camera, Jetson, servo and counts in one system. »
Laura Gómez, Vision engineer