By 2050, more than two-thirds of the world’s population will live in cities.
The challenges of efficiently running a major urban center are many and varied, including (but by no means limited to) accessibility, mobility, sustainability, efficiency, quality of life, and the management of public spaces. Artificial intelligence (AI) has emerged in recent years as a powerful technology for rebooting the world’s infrastructure, providing urban leaders with proven tools for dealing with many of the challenges governments and civil service organizations face every day. Whether energy, water, sewage, transport, or public safety, AI brings a wide range of capabilities to improve the quality of life and operating efficiency that modern cities demand.
The scope of opportunity for urban AI
Day-to-day management of a large urban center is fraught with countless operational challenges in areas as diverse as energy, water (drinking and waste), mobility, safety/crime, food supplies, waste management, and infrastructure operation (roads, tunnels, bridges, pipelines, sewage, etc.). The challenges associated with installing and maintaining these expensive assets often overlap, demanding innovative solutions that balance many simultaneous factors.
Examples of opportunity areas using AI are widespread and diverse.
Aging water networks
Rome’s water grid endures about 44% leakage annually, an extraordinary challenge given the city’s increasing water demand and recent droughts resulting in rationing. AI applications can contribute to tackling this challenge by proactively monitoring, detecting, and preventing leakage.
Leveraging existing infrastructure
Large cities are often equipped with 50 million or more street lights, accounting for 20-50% of a city’s energy bill. As many as 75% of these streetlights are older than 25 years and are ripe for innovation. Optimizing the value of public lighting using remote monitoring and control, lampposts can be ideal infrastructure elements to equip with devices that can collect, communicate and locally process data on traffic, pedestrian flows, environmental factors like air quality, temperature, wind speed, and humidity, and even the ability to listen for and localize gunshots, urban noise, etc. In Barcelona, Spain, for example, smart lighting infrastructure is used to monitor the occupation of beaches and public areas for crowd management.
Waste collection and processing
By installing sensors that measure filling rate and irregularities in public waste containers, efficiency gains can be realized by intelligently routing collection vehicles according to real-time container status. This saves money and minimizes nuisances for citizens due to overfilled or defective containers. AI can be used to not only predict how, when, and where waste is discarded—providing the ability for urban planners to implement efficient waste handling—AI-based computer vision can be used to separate waste flows for recycling programs.
Urban infrastructure planning
AI applied to spatial object recognition from satellite imagery, in combination with machine learning for route optimization, can improve infrastructure planning in the limited space cities have available, supporting and prioritizing work while aligning construction and maintenance demands with limited labor and materials.
Large municipalities around the globe have employed AI to address their urban management challenges, for example:
- Surat, India saw a 27% drop in crime rate after they implemented AI-based safety measures. Surat Police is the first law enforcement organization in India to get a Picture Intelligence Unit (PIU) and is also the first to deploy facial recognition technology. The city police are now using intelligent, video-based analytics deployed at their command and control center.
- Seoul, South Korea has built an integrated public transport system that uses smart cameras in subways to gather information on passenger volumes and uses the data to modify the speed and frequency of trains in real-time.
- São Paulo, Brazil has developed a system that estimates and predicts air quality using AI to analyze data from the mobile network complemented with information from weather, traffic, and pollution sensors. The system allows local planners to calculate pollution levels 24-48 hours in advance, enabling more effective preventive actions.
- San Francisco, California startup Plenty has developed an AI-based vertical farm that has resulted in a 99% reduction in land use and 95% less water usage. In Amsterdam, vertical farming company GROWx has built an AI-driven robotic vertical farm that automates crop cultivation.
- In Helsinki, Finland, Ramboll Group has developed machine learning algorithms that use AI-based time-series forecasting to predict the quality of water leaving their water utilities. Water quality is analyzed in the context of environmental permissions and regulations, providing the ability to monitor the treatment process continuously. This shifts the focus from troubleshooting to predictive risk assessment and dynamic optimization of the facilities.
The role of centralized control and digital twins
An emerging trend in urban infrastructure management is digital platforms, sometimes known as urban data platforms (UDPs). Such platforms monitor and control all the city’s operational activity, providing a holistic view and allowing for event correlation, fast and effective root-cause analysis, predictive recommendations and incident management, and operational insights through visualization. Such UDP systems have been implemented in:
- Cascais, Portugal—Managed service digital command center, predictive management through event correlation, smart waste management systems, etc.
- Vienna, Austria—Data aggregation and analysis, environmental monitoring, urban mobility planning, energy efficiency improvement.
- Hong Kong—Chatbots for citizen communication, real-time traffic management using sensors on 80% of urban thoroughfares, sensor-based disaster preparedness, and visa/immigration document scanning with AI.
In fact, major cities, including Dublin and Singapore, have created digital twins—dynamic digital replicas of each city’s physical assets and environments and their interdependencies—for urban planning purposes, using machine learning to predict future events and trends. A digital twin can be used to support day-to-day operations, simulate a natural disaster and its potential impact on the city, or evaluate the flow of breezes that cool the city and the trees that provide shade on streets and parks. With the evolution of new AI-powered technologies coupled with massive data availability and ever-growing computer processing speeds, digital twins will become increasingly valuable enablers of data-driven decisions. Researchers expect to see increasing adoption rates among city governments wanting to make their infrastructure more resilient. ABI Research has predicted that by 2025 the number of digital-twin-equipped cities will exceed 500 and that by 2030 these cities will be saving more than $280B per year as a result.
Artificial intelligence is the core technological capability enabling the development of smart urban solutions designed to increase safety and efficiency, reduce traffic congestion, improve air and noise pollution, and reduce ongoing operating costs. AI-based solutions like Renewable Suite and Oil & Gas Maintenance Advisor are also essential to continuing efforts to decarbonize the transport sector and reach ambitious emission reduction goals.
AI is a powerful tool to drive the sustainable transition to a more resource-efficient, liveable, and citizen-focused urban infrastructure. Applied to urban infrastructure management and planning, AI can maximize the value of data provided by existing infrastructure (e.g. traffic controllers, sensors, video data, etc.), fleet data, public transport, and even third-party data. The opportunity is vast; the tools and solutions exist today.