The idea
The company wanted its cameras to answer three questions: how many cars entered the lot, how long they stayed, and how many drivers paid. With those numbers, landlords and property managers could calculate stall usage, earn from underused space and run the lot better.
What we did
Using computer vision and deep learning, we built an advanced proof of concept that analyzes video feeds in real time to detect vehicle movement, with the analytics delivered in an AI-enabled mobile app. It pairs low-cost hardware with custom software.
The result
Our first attempt reached a 95% accuracy rate.
