Technology
Inside DOS.
For agronomists, engineers, partners and anyone curious about the details. DOS combines computer vision, onboard computing and electronically controlled valves. All components are in development.
System architecture
Six Parts, One Add-On.
DOS is being developed as an adaptable system that integrates with conventional sprayers rather than replacing them.
- 1
Camera modules
Capture the orchard floor ahead of the nozzles.
- 2
Onboard processing computer
Runs detection on the machine itself.
- 3
Detection software
Computer-vision model that flags weeds.
- 4
Spraying controller
Turns detections into valve commands.
- 5
Electrically controlled valves
Open only over areas marked for treatment.
- 6
Existing agricultural sprayer
The farmer's current machine stays in service.
Processing pipeline
From Image to Valve Command.
- 01
Camera Scanning
Cameras capture images of the ground as equipment moves through the orchard.
- 02
AI Image Processing
Computer-vision software analyzes images and identifies weeds.
- 03
Detection Decision
The system determines whether a detected area requires treatment.
- 04
Smart Valve Control
The planned control mechanism opens and closes selected spraying valves to target identified weeds.
- 05
Continuous Improvement
Testing and feedback help the team refine detection reliability and system performance.
Research roadmap
One Platform. Expanding Possibilities.
Beyond weed detection, we're exploring how intelligent systems could help monitor and manage orchards. Items beyond the first are long-term research directions.
- 01In development
Precision targeted weed spraying
- 02Planned
Advanced weed-detection intelligence
- 03Research Direction
Orchard health, disease, and pest monitoring
- 04Research Direction
Seasonal harvest prediction
- 05Research Direction
Semi-autonomous navigation
- 06Research Direction
A broader orchard intelligence ecosystem