The Manual Entry Problem
Every delivery driver knows the frustration: you arrive at the depot, pick up your parcels, and then spend 30-60 minutes typing addresses into your navigation app one by one. For drivers handling 80-120 stops per day, manual data entry is a massive time sink that delays the first delivery and cuts into earning potential.
Enter AI-Powered Scanning
RouteMate's AI label scanner uses Google's Gemini multimodal model to extract addresses directly from shipping labels, handwritten notes, and printed manifests. Simply point your phone camera at a label, and the address is parsed, validated, and added to your route in seconds.
How It Works
- Capture: Take a photo of one or more shipping labels
- Extract: Our AI identifies and parses all addresses in the image
- Validate: Each address is geocoded and verified against Google Maps
- Add: Confirmed addresses are added as stops to your active route
- Optimize: The full route is re-optimized with the new stops included
Handling Edge Cases
Real-world labels are messy. They have barcodes, logos, sender details, and sometimes handwriting that would challenge a human reader. Our AI model has been trained on thousands of Australian and international shipping labels to handle:
- Multiple label formats: Australia Post, StarTrack, Aramex, DHL, FedEx, and more
- Handwritten addresses: The model can parse most legible handwriting
- Partial information: Missing postcodes or suburbs are inferred from context
- Bulk scanning: Photograph multiple labels at once for batch processing
Real-World Impact
Beta testers reported saving an average of 2 hours per day on data entry. For a driver completing 100 stops, that means:
- First delivery happens 45 minutes earlier
- 12% more stops completed per shift
- Virtually zero address entry errors
- Less frustration and mental fatigue
Privacy and Security
All image processing happens through a secure Supabase Edge Function. Images are processed in memory and never stored. Only the extracted address text is saved to your route — the original photos are discarded immediately after processing.
The AI scanner is available on both our web app and iOS app, with Android support coming soon.