SafeRoadsNSW

Team Name

O.R.G

Team Members

Project Description

Performing exploratory analysis of crash data of the ARDD dataset, to gain insights into how to make NSW roads safer for high school aged kids. Based on these insights, we describe what a NSW tailored action plan could look like. Based on the analysis, our final action plan consisted of:
1. Education and Training: partnering with schools to integrate advanced driver training modules to address the 17 year old deaths spike.
2. Public Awareness: launching state-wide campaigns during high-risk months, using media, billboards, and social platforms.
3. Infrastructure and Enforcement: collaborating with local councils to audit various arterial roads and understand why so many NSW highschoolers are dying on them. Monitoring these roads more vigorously to detect reckless driving.
4. Community Engagement: organising community workshops, emphasising the Safe Systems approach.

The python notebook used to analyse the data, alongside the dataset, can be found on the github. Simply look at the notebook 'analysis2.ipynb' in a notebook editor, or run the code cells yourself, making sure you have installed the packages found in the requirements.txt file.

Data Story

We used the Australian Road Deaths Database (ARDD), a vast repository detailing road-related deaths across Australia. It has attributes such as age, location, crash type, and date, among others, allowing us to investigate the relevant factors of fatal road crashes.
We used a Python notebook, with packages numpy, pandas, and seaborn. We cleaned and structured the data, removing data discrepancies for accuracy. We filtered to only include students aged 13-18, focusing on the last 5 years. To derive state-specific insights, we split the data into two segments: NSW and the rest of Australia.
We had the following insights:
1. The legal driving age brought with it a spike in deaths. Remarkably, while an increase at age 18 would be anticipated due to more students getting licenses, the numbers plateaued, spotlighting the increased risks for early licensees.
2. April and December were months of heightened risk, possibly correlating with Easter and Christmas festivities. Weekends all over Australia are found to be much more dangerous than weekdays.
3. Unlike other states where highways were predominant zones of accidents, NSW's arterial roads stood out as significant trauma hotspots.