Hacki
About the project
Calgary Parking helps drivers answer one simple question:
“Can I park here right now?”
An empty space may look available, but it could be a driveway, a loading zone, or an area where parking is not allowed.
Our app shows parking locations on a map of Calgary. When the user scans an area, it retrieves an image from a public traffic camera and analyzes it directly on the phone.
Using YOLO26s with QVAC ONNX, the app detects cars, checks their position, and estimates the space between them.
The system is designed to be careful. It checks for dark, blurry, blocked, or outdated images. A space must appear free in three scans before it is confirmed. If the evidence is unclear, the app says Review or Refuse instead of guessing.
Inference runs locally. We only use external services for public camera images and OpenStreetMap data—there is no central inference server.
QVAC already supports peer-to-peer delegated inference. In the future, Calgary Parking could use it to send a scan to a configured QVAC peer, while keeping local inference as a fallback.
Calgary Parking does not promise a perfect answer every time. It gives drivers a more private, careful, and honest way to find parking.