A Smart City IoT Crowdsensing System Based on Data Streaming Architecture

Abstract

The subject of this paper is data streaming in IoT crowdsensing systems. The goal of this paper is to present a way of designing a scalable IoT crowdsensing system that enables design of various business models for smart city projects. The system designed in such a way is capable of handling an increasing number of users while maintaining acceptable performance. Performance of the system can be measured in response latency, which allows for real-time tracking of crowdsensing parameters. The first part of the paper deals with data streaming concepts and software solutions, with a particular focus on Apache Kafka. The second part presents the designed system for crowdsensing in smart city environments. The designed system allows for use of mobile and Arduino devices as input data for the Kafka cloud cluster in order to provide crowdsourcing insights in real-time. The primary way that users can utilize these insights is through a web or mobile application, where various data visualizations can be presented. The development of a system based on the proposed model can allow for easy access to recent crowdsourced data, and real-time smart city indicators such as air pollution.