MotoWatch: Computer Vision Tracking
A computer-vision-assisted tracking concept that explores automated monitoring, predictions, and a modern operations dashboard.
Overview
MotoWatch builds on the tracking ideas explored in SFS50. It combines a React dashboard with Python-based computer vision and prediction experiments to reduce manual logging and surface operational trends.
Development process
The concept separates the React client from a lightweight Python service. The dashboard consumes model outputs to present predictions and potential issues alongside ordinary operational data.
Challenges solved
The key design problem was presenting model outputs as useful signals rather than distractions. I explored clear status states and dashboard placement that keep people in control of the final decision.
UI/UX approach
The dashboard uses strong hierarchy, status cues, and data summaries to separate routine tasks from computer-vision and prediction signals.
Inside the product.
