Civil and Environmental Engineering

CEE1: 1-1

Providen Providence: Traffic Incident Management Support System / A vision-language model for automated detection and assessment

1-1 project image
Traffic AI Team at the Order of the Engineer Ceremony. From left: David Oloye, Bruno De Falco, Ariyan Aghamiri, Micheal Ongango

Project Description:

This project automates traffic incident reporting by integrating computer vision and large language models to process camera feeds. For system efficiency, YOLOv11 is deployed as an initial threshold filter that detects potential incidents and prevents unnecessary processing. Once an event is flagged, Google Gemini analyzes the crash data to assess the situation and automatically generates a detailed incident report. This report is then forwarded to a CHART officer who reviews the AI generated insights to make final operational decisions, effectively replacing manual identification tasks and streamlining the entire response workflow. 

Advisor/Instructor:

Dr. Deb Niemeier & Dr. Gretchen Bella & Dr. Terry (Xianfeng) Yang

Sponsor:

Dr. Terry (Xianfeng) Yang

Team Members:

Ariyan Aghamiri Civil and Environmental Engineering
Bruno De Falco Civil and Environmental Engineering
David Oloye Civil and Environmental Engineering
Micheal Ongango Civil and Environmental Engineering

Poster:

Design EXPO POSTER.pptx_.pdf (3.57 MB)

Table #:

G3
Back to Top