The brief
Hi-Res Limited needed a faster, more consistent way to analyse traffic on Uganda's roads. Manual observation is slow, expensive and hard to repeat at scale. Timos was built to automate the collection and analysis, using computer vision and machine learning instead of clipboards and stopwatches.
What we built
- A desktop application that ingests road footage and processes it automatically.
- Computer vision models that detect and track vehicles in real traffic scenes.
- A structured data output designed for traffic analysis and reporting.
- A workflow that lets an analyst review results instead of counting vehicles by hand.
How it works
The application combines a computer-vision pipeline with a machine-learning layer trained on real road scenes. Video is processed frame by frame, vehicles are detected and tracked across frames, and the results are aggregated into traffic counts and movement data. The desktop interface keeps the operation local, so large video files do not need to be uploaded before analysis can start.
Outcome
Timos moved traffic studies from manual observation to an automated, repeatable process, giving the team consistent data across surveys and freeing analysts from hours of frame-by-frame counting.
Technology at a glance
Building something similar?
Tell us what you have in mind. We will reply with a straight answer, not a sales deck.