Five ways smart cities can use AI

COVID-19 has ravaged cities around the world. Shops, businesses and high streets have been turned upside down. As cities begin recovering from the economic and social fallout of COVID-19, could AI help drive improvements in urban services?

There are several motivations behind a smart city: at their most basic level, smart city projects are designed to make life more enjoyable for people who live and work there. Smart cities also have missions to become efficient, technologically-advanced, green and socially-inclusive places that are both nice places to live and work and attractive to investors.

Smart cities around the world are already deploying the latest digital solutions to assist them along this journey. In Beijing, AI is being used to reduce the severity of air pollution by analyzing data taken from coal-fueled factories, industrial sites, weather conditions and traffic congestion. San Diego seeks to make traffic congestion predictable by using AI to analyze data gathered from sensors on streetlights. Dubai uses AI to make public transport safer by monitoring bus drivers' health and condition, leading to a 65% reduction in accidents caused by fatigue.

AI can be a big step forward for smart cities and how they take technology-powered services to the next level. AI could be brought to bear on essential things related to living in cities, like traffic, pollution, waste management, energy efficiency and more.

Potential for innovative AI-powered services in cities

Speaking to the Orange Silicon Valley Hello Show last year, Shilpa Kolhatkar from Nvidia commented, "What's the difference between a regular city and a smart city? A regular city has regular lamp posts; in a smart city, there could be a whole lot more going on inside that lamp post."

The constant evolution of digitally-powered services in smart cities is a key part of ParKam's smart parking solution, according to another speaker at the Orange Silicon Valley Hello Show online event. "We can connect ParKam remotely to existing fixed cameras and use existing infrastructure – so in the IoT era, we can connect to any camera in a city – and then we use our AI engine to collect smart parking information. And then we can tell drivers about parking spot availability. Thanks to our AI tool, we can even predict which parking spots will remain available to drivers in the future. AI enables this probability," said ParKam CEO Asaf Naamani. This is central to the possibilities for AI in smart cities: the ability to learn and subsequently predict makes it easier for cities to manage resources and organize services to everybody's benefit.

Use cases where AI could positively impact cities

  • Traffic management: In today's cities, there are many thousands of private cars and also a huge movement of commercial vehicles, transporting both people and goods. Parking of these vehicles and traffic management is an area that requires constant innovative thinking and is another place where AI can help. As in the ParKam solution, AI traffic management tools learn, then predict traffic flows and space availability to make driving and parking in cities a less frustrating exercise.
  • The environment: As cities commit to a sustainable future, tools and services that leverage AI to reduce environmental impact will be essential. Last year, computer scientists at Loughborough University developed an AI and machine learning (ML) system that can predict air pollution levels hours in advance. The system uses large amounts of data from cities to learn rules and features, enabling it to make predictions. The goal is to use the AI tool to predict fine particulate matter (PM2.5) in cities and give city authorities the data they need to make informed decisions about tackling pollution.
  • Optimization of energy and water management: AI tools could help cities make more efficient use of energy, save money and enhance residents' lives. By tracking people's movement around the city and gathering data on it, AI tools could then be added to the mix to control energy allocation to busy places and locations around the city. The AI would learn patterns of energy use around the city and make supply decisions based on that.
  • Better organized public transport: AI could help cities optimize mobility by easing traffic congestion, managing people's movements better and providing citizens with real-time information updates. Singapore is an example of a city focusing on utilizing connected vehicles, with plans to launch automated public busses as early as 2022. The city has a data-rich intelligent public transportation system that delivers real-time traffic alerts to the public. AI could be layered on top of its systems to learn traffic patterns and make resource-allocation decisions accordingly. Singapore is one of the least congested cities globally, and AI could help it improve further.
  • Waste management: Innovative AI-powered solutions are already being deployed in certain cities around the world. For example, in Sydney, intelligent robots are being used for clearing waterways of plastic pollution and sorting garbage, identifying recyclables for recovery, and learning as they go.

The question of infrastructure

There are huge possibilities for AI in smart cities, but municipalities must ensure that they have the fundamental infrastructure to support the full potential of AI. According to the Erasmus Centre for Data Analytics at the Erasmus University Rotterdam, only 31% of 85 European cities currently have an operational "urban data platform" in place. With all kinds of intelligent, technology-enabled solutions likely to be needed in cities moving forward, they will need to up their game.

Five ways smart cities can use AI

The UN predicts that more than two-thirds of the world's population will live in cities by 2050. IDC has forecast that by 2021, spending on smart cities will exceed $130 billion as municipalities seek to make cities nicer places to live and work, underpinned by advanced infrastructural facilities and social management tools. That said, Gartner previously forecast that AI would become a critical feature of 30% of smart city applications by the end of 2020, up from just 5% a few years prior. It's likely that this number wasn't achieved, mainly due to the impact of the COVID-19 pandemic, but it demonstrates the potential growth of AI in cities.

For more information, discover our Smart Cities capabilities.