EcoTraffic BCN is an EdgeAI smart traffic system that reduces CO₂ emissions and optimizes urban traffic flow in Barcelona in real time.

Three intelligent nodes communicate via MQTT: the Vehicle Node detects cars and calculates waiting times using YoloX-Nano running locally on Arduino UNO Q. The Traffic Light Node receives alerts and decides when to change the cycle using a dynamic demand score. The Dashboard visualizes everything in real time with an interactive map, global metrics and CO₂ saved.

All intelligence runs locally on Arduino UNO Q devices — no cloud, zero latency, under 50€ per node.

Key features:

  • Real-time vehicle detection with YoloX-Nano (COCO dataset)
  • Smart alert system: only sends MQTT alerts when anomalies detected
  • Priority management: emergency vehicles get immediate green light
  • Rush hour adaptation: thresholds adjust automatically
  • Privacy by design: images never leave the device
  • 30% reduction in waiting times
  • 18% reduction in CO₂ emissions

Built With

Share this project:

Updates