This thesis is about the innovation of the various Application of AI in Autonomous Vehicles with a focus on the lane detection technology. Investigating the crucial task of precisely identifying and tracking lane boundaries, which is essential to facilitate secure and efficient self-driving navigation. This study comprises a thorough examination of existing literature related to algorithms used in detecting lanes with particular emphasis on computer vision and machine learning methods. The research, via experimental analyses evaluates the state-of-the-art algorithms' performance under diverse real-lifelike scenarios in virtual or simulated eviroment. Through the use of simulations, the thesis aims to evaluate how lane detection python algorithms utilising machine learning libraries respond to different conditions and road layouts and how modern machine learning models and AI trends can be harvested in the automotive industry. The research involves the use of a simulated environments mimicking real-world conditions, allowing for controlled experimentation and analysis. The challenges of real-world deployment occlusions, dynamic road conditions and diverse lane markings types are emphasised. The thesis also investigates how dependable can lane detection algorithms be, by monitoring the accuracy and error margin values with each training and observing the simulated car behaviour. Acknowledging a crucial need for robustness and versatility within autonomous navigation systems is given special attention in this thesis. Additionally, the thesis discussed the advantages and limitations of using simulated environments for testing and validating lane detection algorithms compared to real-world testing. Consideration of ethical and safety implications in the transition from simulation to real-world deployment were taken into consideration. In conclusion, the thesis end result showed a high-accuracy prediction of the lane detection developed for this project within the simulated environment, offering valuable information for researchers and developers working on autonomous vehicle systems. The findings contribute to the understanding of how well these algorithms perform in controlled, virtual settings, aiding in the refinement and optimization of lane detection systems for future real-world applications.