Detecting and Mitigating Traffic Accidents using Machine Learning and Traffic Data

Local Project Jordan Chapter

Coordinated by the Lead of Jordan, Copeland Brandon , Almubiden Dania,

Status: Upcoming

Project Duration: 15 Dec 2022 - 12 Jan 2023

Project background.

The Jordan Directorate of Public Security reported that there were an estimated 10,857 traffic accidents in the year 2019 alone. Of those accidents, there were a reported 161,511 total serious injuries resulting in the deaths of 643 people with many more suffering severe to minor injuries and an estimated cost for damages totaling 324 million Jordanian dinars.

The problem.

We would like to find an AI solution to help reduce / mitigate the numbers of traffic accidents within the country of Jordan.

Project goals.

  • Analyze Ministry of Transportation datasets for primary causes of traffic accidents.
  • Carry out data preprocessing.
  • Develop a traditional machine learning or deep learning model to help analyze the potential causes of traffic accidents. 
  • Carry out inference with the trained model using test data.
  • Develop some suggestions on how traffic accidents could be mitigated based on data from the provided datasets.

Project plan.

  • Week 1

    Data collection

  • Week 2

    Data Cleaning / Pre-processing

  • Week 3

    Data Analysis / Modelling

  • Week 4

    Final Results / Report

Share project on: