To make an optimal movement of vehicles and to reduce the accident rate, the government has installed traffic lights at almost every intersection. Traffic lights are intended to decrease congestion. However, the dynamic nature of traffic movement causes congestion always. This congestion leads to increased waiting times for every vehicle. In this chapter, two machine learning-based approaches used to improve in the congested traffic environment. The first part of the work is Deep-Learning based traffic signaling, which identifies the congestion on all sides of the intersection with the help of image processing techniques. By analyzing the congestion, the algorithm proposes dynamic green-light times rather than the traditional fixed lighting system. In the second part, a Q-learning-based approach has been suggested in which an agent decides the state of the traffic light based on a cumulative reward. Further, these algorithms have been tested on different traffic simulated environments using SUMO, and detailed analysis has been carried out.
TopIntroduction
To improve the living standards of citizens, smart cities are the future of any country across the globe. This concept is prospering step by step to accomplish the infrastructure advancements and facilities levels, above and beyond (Smart city definition, 2019) (Smart city mission, 2019). To make a city or country smart, the idea is to integrate ICT in all our daily routine life services. Services such as water supply, energy consumption, waste management, healthcare, traffic and transportation, education, agriculture, road infrastructure, surveillance systems and security of citizens are few among many more that can be integrated.
The Internet of things (IoT), sensors can gather information from gadgets, peoples and their belongings. This information would be further assessed using data analytics tools and intelligence could be added using Artificial Intelligence techniques. It could be utilized towards ongoing environmental challenges, for resource optimization, service administration and many more. City authorities can screen real-time information about the ongoing activities of the city. This could be one of the ways where government or the board individuals can keep a track about the troubles faced by their citizens and improve the quality standards. Active participation of citizens and governments is widely expected which will lead to a positive change in the general public standards (Smart city, 2019). A solution development towards designing it should be done by keeping in mind the two important objectives: sustainability and citizen-friendly.
Urbanization growth is also one of the important factors which depict the need for smart cities. Due to very few facilities in rural areas, more and more people are migrating towards urban areas. As per the UN report, two out of every three people will live in cities by 2050 (UN report, 2019). To handle this much of urban population, detailed planning of urban areas concerning to resource constraints, economic and environmental demands concerning air, water resources; health, education and transportation services, etc. is highly required.
From all these highlighted areas and current need to cater population and governance, it can be concluded that smart cities are the need for the development of any growing country. Thus all over the world, government officials are spending huge money to drive social transformation and make these projects realistic in upcoming times. Across the globe, Governments launched several projects. Smart city strategy is adopted by many cities such as Manchester, Dubai, Singapore, New York, Taipei, Amsterdam, City of Columbus and many more. If we talk about India, Prime Minister Narendra Modi announced his mission in 2015, about making 100 smart cities across the country (Smart Cities Mission, 2016). Government prepared a budget of approx. Rs 201,981 crore on various projects under this mission and it was expected that 20 cities would accomplish the objective of the smart city by 2021 (Hindustan Times, 2019). Not only government bodies, researchers and academicians also wanted to be the part of this success story of transformation. Some solutions in terms of digitization of data, online education, e-health services, affordable housing, better road connectivity, mobile computing, etc. are already developed but still there is a need of more adaptive and sustainable solutions (Features, 2019) (Rana, 2018) (Yin, 2015).