Abstract
The research develops a robust AI-powered predictive system utilizing convolution neural networks (CNNs), IoT-enabled devices, and advanced machine learning (ML) models to improve real-time monitoring, resource allocation, and decision-making. The proposed system aims to minimize emergency response times, optimize resource allocation, and enhance public safety. Using machine learning models such as convolution neural networks (CNNs) and random forests, the system integrates IoT devices, traffic cameras, and deep learning algorithms to enable proactive incident management and efficient decision-making. A methodological framework based on CRISP-DM guides the project, encompassing historical data analysis, system modeling, and evaluation against real-world data from UAE traffic authorities. Comparative analysis of ML techniques such as logistic regression, random forests, and CNNs ensures robust predictive capabilities. Metrics which include precision-recall and F1-score will validate system performance, ensuring scalability for full-scale implementation. Challenges include limited data availability, privacy regulations, IoT reliability, and AI biases, which the study addresses through preprocessing, legal compliance, and hardware optimization. The findings demonstrate the high potential of AI-driven systems to transform traffic management, achieving faster response times, efficient resource utilization, and a reduction in emergency response delays. By leveraging AI technologies for real-time traffic monitoring, this research provides a scalable, data-driven solution to reduce traffic fatalities and mitigate economic and societal impacts of RTAs. This paper investigates the transformative potential of AI to enhance public safety and emergency management in the United States. It makes a proposition for an integrated approach towards the optimization of crime prevention, disaster response, and public safety operations by leveraging AI-driven predictive analytics, ML models, and advanced threat detection algorithms. This research is based on recent advancements in predictive analytics, deep learning, and cyber security measures. Using the real-time analysis of crime statistics, social media, IoT sensors, and environmental conditions, this paper is aimed to demonstrate how AI may significantly reduce response times, detect public safety threats before time, and efficiently utilize resources. Case studies will be developed to describe the practical use of AI in real-time crime prediction, disaster management, and threat detection. It will illustrate how teams in law enforcement, emergency management, and cyber security can make the most of AI to shift from reactionary to proactive, intelligence-led operations. Second, it will focus on ethical considerations of AI use in public safety: privacy and algorithmic bias, discussing the frameworks to ensure this use is done responsibly. It will offer a strategic approach toward the integration of AI technologies into U.S. public safety infrastructure, preparing for growth, and safeguarding America's communities from rising violent crime trends, natural disasters, and emerging threats.
Library of Congress Subject Headings
Traffic accidents--Forecasting--Automation; Emergency management--Automation; Neural networks (Computer science); Machine learning
Publication Date
Fall 2024
Document Type
Thesis
Student Type
Graduate
Degree Name
Professional Studies (MS)
Department, Program, or Center
Graduate Programs & Research
Advisor
Sanjay Modak
Advisor/Committee Member
Ehsan Warriach
Recommended Citation
Al Falasi, Humaid Ahmad, "Predictive Rescue System through Real-Time Accident Monitoring Leveraging Artificial Intelligence" (2024). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/12013
Campus
RIT Dubai
Plan Codes
PROFST-MS
