Towards Context-Aware UX in Automated Mobility: BLE Based Passenger Detection via Smartphones

Understanding how people navigate urban environments is central for designing intelligent transportation systems and adaptive interfaces, particularly in the context of Human Automated Vehicle (HAV) interaction. In this paper, we investigate whether Bluetooth Low Energy (BLE) scanning on smartphones can detect a user’s mode of transport by passively sensing the presence of nearby devices. Unlike traditional approaches that rely on Global Positioning System (GPS) or motion sensors, our approach determines whether a user is traveling alone or in a shared space, such as a metro or bus, by analyzing the quantity and stability of nearby BLE signals. We developed a lightweight iOS app that logs BLE advertisements and trained a classifier to differentiate between cars, buses/trams, and metros. Our findings indicate that BLE patterns are typically consistent enough to support practical classification, demonstrating strong performance in detecting cars and other vehicles while providing valuable insights for shared transportation modes. The system avoids sensitive privacy permissions, and integrates into a user-friendly interface that provides feedback and allows for corrections. We explore how this approach can enhance passenger detection and contextual adaptation in automated vehicles. This includes adjusting interfaces, improving safety features, and facilitating adaptive behaviors. The findings reveal that although BLE-based sensing has limitations in precision, it provides a low-cost and privacy-aware alternative to traditional sensing strategies used in mobility applications.