AI-Powered Cybersecurity Threat Detection In Automotive Vehicles(CARS)

AI-Powered Cybersecurity Threat Detection In Automotive Vehicles(CARS)

Manas M N*, Pavan kumar R1, Praveen Prakash Hebbal1, Nithin Gowda L1, Prajwal M Biradar1

* Associate Professor, Department of Computer Science and Engineering, R V College of Engineering

1 B.E Students, Department of Computer Science and Engineering, R V College of Engineering

Abstract - The increasing integration of digital technologies in modern vehicles has led to significant advancements in autonomous driving, connected car services, and in-vehicle communication networks. However, this progress has also introduced a new array of cybersecurity vulnerabilities, making automotive systems prime targets for cyber threats, including malware attacks, unauthorized access, and data breaches. Traditional cybersecurity approaches often fail to detect sophisticated and evolving threats in real time, necessitating advanced solutions that leverage artificial intelligence (AI) for enhanced security.

This project proposes an AI-powered cybersecurity threat detection system designed to identify and mitigate cyber threats in automotive environments. The system employs a multi-layered approach combining real-time data collection, anomaly detection, and machine learning algorithms to detect suspicious activities within in-vehicle networks. Key features such as behavioral analysis, deep learning-based intrusion detection, and real-time anomaly monitoring allow the system to distinguish between normal and malicious activities with high accuracy. AI models, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Isolation Forests, are trained on vast datasets of vehicular network traffic to improve threat detection capabilities dynamically.

A major innovation in this project is the implementation of federated learning, enabling AI models to be trained across multiple vehicles without sharing raw data, thus preserving user privacy while enhancing threat intelligence. The system also integrates blockchain technology for secure logging and validation of detected threats, ensuring data integrity and reducing the risk of tampering. By combining AI-driven analytics and decentralized security mechanisms, this approach enhances cybersecurity resilience in modern vehicles.

The results of this project demonstrate a significant improvement in threat detection accuracy, with AI models achieving over 95% detection rates while minimizing false positives. The integration of blockchain ensures secure data exchange among connected vehicles, strengthening the automotive cybersecurity framework. These findings highlight the potential of AI-powered solutions in safeguarding automotive networks against emerging cyber threats.

In conclusion, this research presents a robust AI-based cybersecurity system for modern vehicles, addressing key vulnerabilities in automotive networks. By leveraging real-time AI threat detection and blockchain-enhanced security, the proposed system improves vehicle safety, reduces cyberattack risks, and contributes to the development of resilient intelligent transportation systems. Future work may involve expanding the system's capabilities to detect zero-day attacks and integrating it with industry-wide cybersecurity standards for enhanced interoperability.

Keywords: Automotive Cybersecurity, AI-Powered Threat Detection, Machine Learning, Intrusion Detection, Blockchain Security, Connected Vehicles, Federated Learning.