#Executive Summary
Through our in-depth analysis of car crash data in Montgomery County, Maryland, we set out to unravel the underlying causes of crashes and develop a predictive model to identify driver-at-fault incidents. The dataset, gathered from the Maryland State Police's Automated Crash Reporting System (ACRS), provided rich insights into motor vehicle collisions occurring across the county's roadways.
Our exploration uncovered several noteworthy findings:
-
Trends Over Time: We observed fluctuations in crash frequencies over the years, with a noticeable dip in reported incidents during 2020. Different municipalities and road types showed varying rates of crashes, highlighting areas of concern.
-
Location Matters: Certain municipalities, particularly those along major routes like Maryland Interstate Roads and County Routes, experienced higher crash rates.
-
Understanding Crash Severity: Most crashes resulted in no apparent injuries, underscoring the need for proactive measures to prevent severe injuries and fatalities.
-
Weather and Road Conditions: Surprisingly, the majority of crashes occurred during clear weather and on dry roads, challenging common assumptions about adverse weather conditions.
-
Vehicle Trends: Toyota emerged as the most frequently involved vehicle make in crashes, pointing towards potential areas for targeted safety initiatives and driver awareness campaigns.
-
Insights from Advanced Analysis: Our investigation into injury severity based on vehicle impact locations and collision types revealed correlations that can inform targeted road safety interventions.
-
Machine Learning Model Performance: Among the models tested, the Random Forest classifier stood out with the highest accuracy (87.71%) in predicting driver-at-fault incidents based on our dataset's features.
Building on our insights, we propose the following actions:
-
Focused Interventions: Implement targeted safety measures in municipalities with high crash rates, addressing specific road types, weather conditions, and vehicle-related factors identified in our analysis.
-
Continuous Data Improvement: Regularly update and expand the dataset to capture evolving trends and factors impacting car crashes in Montgomery County.
-
Deployment of Predictive Tools: Utilize the validated Random Forest model to proactively identify potential driver-at-fault incidents, enabling timely interventions and resource allocation.
Our project sheds light on the complexities of car crashes in Montgomery County, offering a data-driven roadmap for enhancing road safety. By leveraging these insights and recommendations, stakeholders can prioritize interventions, allocate resources effectively, and work together towards reducing accidents and improving road safety outcomes for everyone in our community.
This endeavor exemplifies our commitment to harnessing data-driven solutions to address real-world challenges and make a positive impact on road safety within Montgomery County.