Maintaining driving independence is crucial for older adults, yet cognitive decline poses challenges that impact their safety and decision-making on the road. Additionally, the ability to identify drivers based on unique behavioural patterns has significant implications for vehicle security, personalized driver assistance, and road safety. This thesis explores these interconnected themes through two complementary studies that leverage real-world driving data to gain insights into navigational behaviors and driver identification. The first study examines the relationship between cognitive impairment and driving patterns in older adults. Using GPS data from 246 participants, including 230 cognitively normal individuals and 16 with incident cognitive impairment, spatial clustering and hashing algorithms revealed significant differences in driving behaviours. Participants with cognitive impairment demonstrated reduced route variability and reliance on fewer distinct paths to common destinations, offering evidence for driving as a potential digital biomarker for early detection and monitoring of cognitive decline. The second study presents a privacy-preserving methodology for driver identification using gas and brake data recorded during maneuvers. An unsupervised acceleration-based maneuver detection method was developed, followed by a supervised Long Short-Term Memory (LSTM) model for predicting driver identity. Results demonstrated the efficacy of this approach, achieving an average accuracy of 90.9% across crossvalidation folds. The findings highlight the existence of personal driving signatures embedded in gas and brake patterns, offering a scalable and privacy-conscious solution for driver identification. Together, these studies aim to uncover insights into how cognitive health and individual driving behaviours shape real-world navigational patterns, advancing our understanding of driving as both a digital biomarker and a personal identifier. By integrating behavioural analysis with scalable computational models, this thesis advances the potential for personalized driver systems, digital health monitoring, and vehicle security in naturalistic driving environments.