Serverless Digital Mobility: A Case Study on Event-Triggered Passenger Counting.

Digital mobility providers are among the largest producers of real-time public data. Locations of vehicles, passenger movement and activity, connection inquiries, ticketing and related data sources cause high demand on the computation architecture which has to operate under elastic scalability, yielding predictable latency to satisfy consumers of real-time insights. Serverless architectures are meant to address such concerns at low cost and at acceptable simplicity for application engineers. In this paper summarising an international research collaboration, we aim to validate this statement. We claim that our work is the second one to study the use of serverless application technologies in the domain of digital mobility, and the first to run a field study with empirical evaluation. Our contributions include a real-time sensing, observation and fusion architecture, a field experiment to link the architecture to ground operations by mobility providers, and a synthetic upscaling of the obtained results.