LITMIR.BIZ

2 Data Analytics for Smart Cities: Challenges and Promises

       Ghareh Mohammadi1, Farzan Shenavarmasouleh1, M. Hadi Amini2, and Hamid Reza Arabnia1

       1Department of Computer Science, Franklin College of Arts and Sciences, University of Georgia, Athens, Georgia

       2Knight Foundation School of Computing and Information Sciences, Florida International University, Miami, FL, USA

2.1 Introduction

      Motivation: The explosion of advancement in artificial intelligence (AI), sensor technologies like Internet of things (IoT), and wireless communication activates ubiquitous sensing through distributed sensors in various domains of networks that lead us to smart systems in healthcare, transportation, environment, and other relevant branches/networks. Having collaborative interaction among the smart systems connects end‐user devices to each other, enabling a new integrated entity called smart cities.

      In the last decade, the IoT devices have been connected among different independent agents and heterogeneous networks as with communication technologies [1, 2]. The connected high‐performance sensors and end‐user devices, IoT, are the trigger leveraging the networks in transitioning from urban cities toward smart sustainable cities. The goal of smart cities is to address upcoming challenges of conventional cities by offering integrated management systems with a combination of intelligent infrastructures [3].

Making decisions in smart cities is challenging due to the high direct/indirect dimensional factors and parameters. In this chapter, we aim to focus on one of the smart cities' important branches, namely, smart mobility and its positive ample impact on smart cities' decision‐making processes. Intelligent decision‐making systems in smart mobility offer many advantages, such as saving energy [4], relaying city traffic [5], and more importantly reducing air pollution [6] by offering real‐time useful information and imperative generated knowledge. Making an optimal decision in time in smart mobility with a wide variety of smart devices and systems is challenging. You cannot make a promising decision when your data is not frequently collected. Consequently, a training process of decision support systems still is challenging due to the lack of data [7]. In this chapter, first, we address current challenges in smart cities and provide an overview of potential solutions to these challenges. Then, we offer a framework of these solutions, called universal smart cities' decision making, having three main sections such as data capturing, data analysis, and decision making to optimize the smart mobility within smart cities.

      Interestingly, large cities are losing their own urban style and turning into smart ones. Smart cities are growing due to the advanced technology, especially AI. The more people live in large cities, the more need to have an integrated system to cost‐efficiently handle the ample growth in urbanization. The proliferation of population offers smart development challenges in such cities and enables enormous pressure on society to create innovative, smart, and sustainable environments. Therefore, today's developed cities (or smart cities) are in need of integrated smart policies and novel innovative solutions to enhance the monitoring functionality to facilitate urban living conditions [8].

      Smart cities are created to enable advanced capabilities, such as sustainable energy systems and electrified transportation networks, and interact with information and communication technologies (ICTs) to enhance the efficiency of the cities being generated [9, 10]. An example of responding to new changes for smart cities may be seen in the progress being made in smart healthcare systems for emergency care cases. For the hospitals to be able to monitor and control their patients and let the specialists offer better solutions to their patients, they need to use a smart healthcare system [11, 12].

      To develop the smart healthcare system, we need to use healthcare networks that are the inter‐ and intra‐connection among the healthcare components like IoT devices and sensors to enhance the process of monitoring and consequent services for patients [4, 13]. The performance of such a system heavily depends on the quality of this network communication or online synchronization with other associated networks (other smart systems in smart cities) that actively contribute to the service operation in a positive way. For example, a healthcare network may take advantage of networks' resources, like energy, in case the system fails to run an operation due to the lack of enough power supply. In this situation, the best solution is connecting this network with other networks for the benefit of both patients and their health and service management reliability [4].

In the processes of making and turning cities into smart cities, we can identify several representative features. According to a research study [14], the scientists investigated a period of 10 years of work from 2008 to 2018 and discovered that smart cities might have some features in common. The most interesting features in Figure 2.1 [9, 14, 15] are smart economy, smart people, smart governance, smart mobility, smart environment, smart energy, and smart living that are shown in Figure 2.1. As shown in Figure 2.1, although each feature represents the importance of itself to smart cities, they all are interrelated to each other. Each feature has direct or indirect impact on others. Furthermore, proposed decision‐making solutions used in smart cities are multi‐criteria decision making (MCDM), mathematical programming (MP), AI, and integrated method (IM). In this study, we aim to discover AI solutions, particularly machine learning and deep learning (DL) approaches rather than others, for smart mobility.