In recent years, the idea of personal mobility devices (PMD) has gained prominence globally for different contexts, for diverse types and extent of uses. The advantages of owning a PMD allows users to cover the short distance in between stops where they have access to long distance transportation, establishing a full end to end transport system for many. The rise in usage of PMDs also came along the rise in accidents. One of the reasons that could result in this phenomenon is the lack of calibration of PMD towards how users use it. Currently, most user experience (UX) methodologies are based on subjective questionnaires rather than by objective quantitative data. While there exists a few that studies wheelchair and electronic bicycles, UX concerning this specific device is a field not many studies have delved into. Therefore, in this project, we seek to propose a data-driven model to explore electronic scooter user’s riding profile based on psychophysiological data such as galvanic skin response (GSR) and kinematics data such as the speed and acceleration. Upon retrieving the stress status of the user when he or she is riding, the dataset undergoes a data analysis pipeline that cleans, process and analyse data with Random Forest machine learning algorithms. With the ability to create customised profiles, the model can be adopted to serve the needs of PMD sharing service stakeholders or PMD design companies to ensure good user experience for their customers in the future.