Machine Learning and AI Techniques in Green Cities

Abstract

To meet the requirements of a smart city, efficient utilization of information and communication technologies (ICTs) are highly necessary to adequately administer the data analytics, data communications, and effective implementation of complex strategies to ensure the smooth and secure operation of a smart city. The Internet of Things (IoT) is the most important and significant constituent part of most of the smart city applications, which are responsible for generating an immense amount of data. In the presence of such amounts of big and complex data, it is difficult to precisely decide the most accurate and efficient actions. The best possible analysis of the big data can be carried out using advanced techniques like artificial intelligence (AI) and machine learning (ML) to reach an optimal decision. The preceding techniques take a long-term objective into consideration and can lead to the best possible or near-optimal control decisions. The accuracy and precision of the aforementioned techniques can be further enhanced by increasing the amount of training data to strengthen their learning capabilities and hence the automated decision efficiencies. This chapter plans to discuss the AI and ML solutions for air pollution control, on-demand intelligent transport, intelligent health care, and smart power distribution for a sustainable smart city. Air pollution in urban cities is a serious issue that is persisting in the environment and creating havoc for humans. Contamination of air is due to effluents from industries, manufacturing sectors, the burning of fossil fuels, etc. In this chapter, the context-aware air quality prediction method is discussed, which is used to evaluate the position of air pollution in urban areas. The context-aware AQ-prediction system is used for the formulating an algorithm for estimation of pollution level and is based on the available data. Prediction of forthcoming events can be formulated and necessary corrective action can be put into place. Demand-based intelligent multi-passenger communal transport options are being promoted as an influential strategy to reduce traffic crowding and emissions of CO2 and to improve the convenience and travel experience for passengers. These services, often referred to as on-demand based public transport, are aimed at meeting personal travel demands through the use of shared vehicles that run on flexible routes using advanced tools for dynamic scheduling. This chapter discusses the AI-based traffic simulation models to predict the traffic flow with existing bus service and provides on-demand public buses depending on the number of passenger waiting times, hourly vehicle utilization, total kilometers traveled, dates, and times. Depending on the requirements, AI techniques automatically give instructions to the drivers and also the passengers for bus timings and the optimum routes for travel.