A distributed system architecture for pedestrian flock detection with participatory sensing

We present distributed system architecture for smartphone-based participatory sensing applications, where the computational load is distributed between the participating devices. The system allows using of static data to reduce the computational load further. Secondly, we present a conceptual participatory sensing application for detecting pedestrian flocks moving into the same direction in certain area, using this architecture. The flock detection is based on selecting an energy efficient set of sensors in the smartphone and available location-based static data, provided by previous computations by the smartphones or by external Web services. Initially, we believe this method reduces energy consumption in the participating devices beyond the previous approaches.