Reducing uneven crowd distribution in the Stockholm metro system using data driven design.

Abstract: The purpose of this project was to develop several information system concepts with the goal of reducing uneven crowd distribution in the Stockholm subway system, using a data driven, user centred, and iterative design approach. These systems were designed to be implemented on-site in the subway. The project also sought to evaluate the efficacy of different types of information systems. The project was initiated with a background study, involving a study of available literature from previously conducted studies, as well as a state-of-the-art analysis of systems currently available on the market. The background study was followed by an analysis of train load data in order to find patterns in traveller behaviour, as well as stations of particular interest for user studies. Initial user studies were conducted through contextual and in-depth interviews, an online survey, and passive observations onsite in the subway. Findings from the load data and user study analyses were then utilised in the concept development phase, during which several concepts were created. Concepts were evaluated in an iterative manner, using continuous user feedback to drive development. In total, nine different concepts were developed and tested at some stage. Ultimately, one concept was chosen for field tests with a prototype implemented at Tekniska Högskolan subway station, which further drove development while also generating objective data through the load measuring systems built into the train. The test involved delivering real-time crowding information to travellers using a system of signs, and RGB-LED modules mounted in the ceiling of the platform. The findings generated throughout the project were condensed as a set of design considerations to be used when designing information systems for use in the subway. Such considerations include factors such as information detail and abstraction, information placement, and visual design principles. The field tests indicated that the delivery of real-time crowding information is highly valued by travellers and has the potential to both increase user satisfaction and reduce uneven crowd distribution.