Robust IoT Based Parking Information System

Abstract

In this paper, we propose an adaptive parking information system. In this system, sensors and cameras monitor choke points in a parking space. Using process discovery algorithm, and by adhering to assumptions described in this paper, a Petri Net model is generated which tracks the occupancy of the parking space. This model automatically adapts to adding or removing sensors and to sensor failures. We also discuss how to detect sensor failures using relatively small amount of data. A use case is also provided which demonstrates the benefits of parking information system to smart cities.