Lane detection is a critical component of autonomous vehicles, enabling them to identify road boundaries and maintain safe navigation within traffic lanes. Human drivers naturally rely on visual lane markings to make steering and directional decisions. Similarly, autonomous vehicles depend on AI-enabled frameworks such as Deep Learning algorithms to detect and track lane lines accurately. Despite the apparent visibility of lane markings, reliable lane detection remains a significant challenge due to varying road conditions, illumination changes, reflections, traffic congestion, occlusions, and differences in lane marker types. This study investigates methods for improving the robustness of lane detection systems for safer autonomous navigation. The research aims to improve lane detection accuracy and reliability under complex real-world traffic scenarios. By enhancing the robustness of lane recognition systems, this study contributes explore approaches to reducing navigation errors, minimising the risk of lane departure, and improving the overall safety and effectiveness of autonomous vehicles. Through an interdisciplinary lens, this study analyzes recent amendments to AV regulations in California and argues that the proposed lane detection framework could be used by manufacturers to comply with these more stringent safety requirements.