Project Abstract

This project is motivated by the pressing necessity to improve cybersecurity in the automobile industry, with a specific emphasis on driver profiling and car-theft detection. With the growing connectivity and autonomy of automobiles, there is an increasing risk of unauthorised access and theft. Therefore, enhanced security measures are necessary. The main objective of this project is to create a strong system utilising the Driving Dataset that not only characterises driver behaviour but also identifies possible car thefts, so making a substantial contribution to vehicle safety and security.The study suggests an innovative method that combines machine learning techniques, including clustering algorithms and deep learning models, to examine and categorise driver behaviours into several groups such as aggressive, cautious, and distracted driving. In addition, the research investigates anomaly detection techniques, including statistical and unsupervised learning algorithms to identify suspicious activities that are indicative of auto theft. This work is positioned uniquely at the junction of driving behaviour analysis and theft detection due to its dual methodology.The project entails the systematic development and incorporation of driver profile and car-theft detection modules into a cohesive cybersecurity framework. This framework is designed to be compatible with current automobile systems, guaranteeing smooth interoperability. The system’s performance will be assessed by analysing real-world driving data in several scenarios, including simulated car theft attempts, to measure accuracy, false positive rates, and response times.The anticipated results seek to emphasise significant patterns in driving conduct and the efficacy of anomaly detection in identifying prospective thefts. The knowledge acquired from these discoveries will be crucial in creating customised interventions and improving vehicle security protocols.Upon successful completion, this research will greatly enhance the field of automobile cybersecurity. This will yield useful observations on driver behaviour patterns and the effectiveness of anomaly detection tools, providing a holistic solution to contemporary difficulties in vehicle security. The project aims to provide proven findings that showcase the system’s efficacy in enhancing safety, security, and user experience in automotive environments. This will in turn contribute to the development of future innovations and initiatives in the sector.

Keywords: Artificial intelligence, data science, security

 Conference Details

Session: Presentation Stream 13 at Presentation Slot 3

Location: GH049 at Wednesday 8th 09:00 – 12:30

Markers: Chen Hu (GTA), Hassan Eshkiki

Course: MSc Advanced Computer Science, Masters PG

Future Plans: I’m looking for work