AI Double Detection System
Stop losing throughput to double feeds.
AI double detection that holds up on a real line — trained on parcel data from 54+ camera sources worldwide. Robust to lighting, blur and item variation, with no endless parameter tuning.
Numbers your operations team can bank on.
One camera. One and a half metres. Decision made.
The system runs at 30 frames per second, detecting items on high-speed conveyors and making rapid decisions over short distances. That responsiveness allows a compact detection footprint and optimal space utilisation.
The detection area is determined by your largest item: a typical setup uses roughly 1.5 metres of conveyor, suitable for items up to 1200 × 800 × 600 mm — and it integrates seamlessly into new or existing systems.
- No line rebuild. Retrofits onto conveyors you already run.
- No tuning marathon. Pre-trained on 54+ real camera sources.
- No blind spots. Reads dark, black and shiny items alike.
A metre and a half is all the room it needs.
- One camera above the belt. No encoder hardware, no line rebuild.
- 1.5 m decision window. Sized to your largest item, not a fixed spec.
- Doubles flagged in-flight. Handed straight to your diverter or splitter.
Built for the messy reality of real parcels.
Shiny tape, black poly bags, overlapping items, changing light. The failure modes that break conventional vision are the ones we trained on.
- 01
Real-time image acquisition & processing
High-speed capture and on-the-fly processing at 30 fps for instant decisions.
- 02
Multi-parameter analysis
Size, shape, position and more — analysed together for reliable double detection.
- 03
Dark & black object processing
Robust detection even on dark and black items that defeat conventional vision.
- 04
Virtual encoder tracking
Precise item tracking along the conveyor without extra encoder hardware.
- 05
Image storage & remote diagnostics
Stored detections and remote access for fast troubleshooting and audits.
- 06
Material-based detection
Distinguishes items by material characteristics for smarter singulation.
One system, six configurations.
The same detection core, mounted and actuated to suit your line. Explore each configuration in interactive 3D — drag to rotate.
VSU (Vertical Split Unit)
Vertical diversion of detected doubles — separate flagged items without stopping the line.
Induction station flow optimizer
Optimises parcel flow at the induction point, keeping singulated items feeding the sorter at full rate.
Circular cross belt sorter
DDS camera integration for circular cross-belt sorters — detection built into the loop.
Linear cross belt sorter
The same detection intelligence configured for linear cross-belt sortation lines.
Plug and play solution
A self-contained camera unit with CHD — drop-in detection for new or existing conveyors.
Line splitter with detection & action
Detects doubles and acts — camera plus switch to split the line automatically.
Stay in control from anywhere on site.
A smart, web-based dashboard. Track detections as they happen, export reports, and open the stored images behind any decision — without a site visit.
- Live detectionsEvery read as it happens.
- Image storageStored frames per detection.
- Remote diagnosticsTroubleshoot without a site visit.
- Report exportPull records for audits.
- Real-time monitoringLine status at a glance.
From first call to full-scale deployment.
- 01
Analyse operational needs
Understand your line, items and throughput targets.
- 02
Select optimal camera setup
Configure AI-driven vision for precision detection.
- 03
Deploy initial models
Run preliminary AI models to gauge performance.
- 04
Proof of concept
Collect data and validate accuracy with real-world testing.
- 05
Full-scale deployment
Integrate and optimise the solution for production.
See 
DDS on your line.
Book a demo and we'll show you double detection on real parcel data — then map the setup, footprint and ROI for your sorter.