Open-Source Big Data Platform for Real-Time Geolocation in Smart Cities

Nowadays, big data analytic tools and Internet of Things applications boost productivity in Intelligent Transportation Systems in the context of smart cities. Each day, location mobility data are generated continuously from Global Positioning System devices in a high temporal granularity. This article introduces a framework for public transportation mobility analysis. The proposed big data platform uses open source components for real-time geolocation tracking processing. The platform collects location information over Message Queue Telemetry Transport protocol to Apache Kafka, and then information is processed using Apache Storm, which guarantees fault tolerance, horizontal scalability, and low latency. Experimental evaluation is performed for a case study considering 10357 taxi tours (17 million GPS timestamps) using problem instances of different sizes. Results demonstrate that the proposed open-source big data platform is capable of processing a significantly large number of GPS timestamps of tested instances in reasonable execution times.