Mobile crowdsensing of parking space using geofencing and activity recognition

New mobile phones come with an increasing array of sensors. Recently mobile operating systems have started to incorporate means to offer contextual information derived from measurements of multiple sensors. A phone can be aware of whether it is transported in a vehicle or carried on foot. We investigated how these sensing capabilities could be used to derive information about available parking places in the absence of parking sensor infrastructure and tested the method with a dedicated mobile client connected to a network server. In this document we present the sensing method and the application, and qualitatively analyze the pros and cons of the approach.