Big data and artificial intelligence promise to transform our cities for the better.
Whether it is to forecast water and electricity demand or to predict traffic congestion or facilitate disaster and epidemic response, cities are using a range of AI applications in a bid to protect the environment and improve living conditions.
These were the observations made at The Klosters Forum (TKF), an annual gathering which brings together high-profile participants from the fields of science, business, politics and industry to tackle some of the world’s most pressing environmental and societal challenges. The 2024 forum, which took place in late June, focused on urban resilience.
“Cities will face numerous challenges in the future, such as floods, urban heat islands and social issues,” says Martin Hullin, director of digitalisation and the common good at Bertelsmann Stiftung, a German independent foundation.
“In order to make cities more resilient, data information analysis will be an important tool, which will only reach their full potential if we break data siloes and enable leveraging both private and public data sources.”
AI urbanism
Copenhagen, Singapore and Barcelona testify to what can be achieved. Their efforts and successes include:
- Hyperlocal air quality monitoring: Copenhagen has partnered with Google to map street-by-street air quality, generating insights for municipal authorities that can then be used to develop anti-pollution policy and urban planning. For example, Copenhagen is using air quality mapping to design future neighbourhoods that include “Thrive Zones” – connected neighbourhoods with schools, playgrounds and integrated transportation networks located away from high-pollution zones. Spurred on by the Danish capital’s achievements, Hamburg, Dublin, Amsterdam, London and Bengaluru are pursuing similar initiatives.
- Energy, water and sewage management: Copenhagen is also using AI to reduce and optimise energy consumption in the municipal buildings in order to reduce carbon emissions as well as the city’s heat and electricity bill. Barcelona uses big data, machine learning and artificial intelligence to predict water consumption and maintenance needs, manage supply and model wastewater treatment in the sewer system.
- Traffic management: Singapore has pioneered the use of intelligent transport systems to optimise traffic flow and reduce congestion; it has also used big data to deliver effective pandemic responses. During the Covid crisis, the country used AI-powered technology to support digital contact tracing, efficiently roll out nation-wide vaccinations, speed up temperature scanning and promote safe-distancing. Capitalising on the well-established digital infrastructure, the city state plans to harness AI in the coming years to improve public services and encourage AI innovation and adoption in the country’s leading economic sectors such as manufacturing, financial services and biomedical sciences.
Resource efficiency gains reduce cost, but you may have a rebound effect where you may be stimulating demand.
Digital exhaustion
For all the promises AI make, participants were also cautious about a growing environmental footprint of the data industry.
The proliferation of AI is already causing a big increase in the environmental footprint of data centres for examples. Data centres form the foundations for the cutting-edge technology but compete for power and water with the rest of cities and their inhabitants.
The International Energy Agency expects electricity consumed by data centres will more than double globally by 2026 to more than 1,000 terawatt hours, an amount roughly equivalent to what Japan uses annually.
Data centres are also water guzzling. In recent years, tensions over water use by data centres have flared in water-stressed communities across the United States. Big techs like Google and Meta have come under pressure for withdrawing so much groundwater at a time when climate change is making droughts worse and more frequent.
Microsoft, in its sustainability report published in May, revealed that its total carbon emissions are almost 30 per cent higher today than in 2020 because of its global data centre expansion. This makes it more difficult for the tech giant to achieve its carbon-negative status by 2030.
“Former environmental trailblazers in the tech ecosystem are already rolling back their SGD-goals in order to cash-in on the AI hype,” Hullin says.
“The environmental impact of AI represents a currently underappreciated threat in the global discussion on AI, which given its unreasonable perception as a silver bullet should not stand. Balancing the benefits of AI with its environmental costs is essential and it’s a debate that is only just starting.”
The industry is encountering a challenge where gains in efficiency, resulting in reduced resource usage per unit of production, may be counterbalanced by increased consumption of the same product.For more, see https://am.pictet.com/ch/en/institutions/investment-views/active-equity/2023/technology-and-esg#overview
“Resource efficiency gains reduce cost, but you may have a rebound effect where you may be stimulating demand,” says Steve Freedman, head of research and sustainability of thematic equities at Pictet Asset Management.
“Benefits (of AI) are not visible yet. Data centres are grabbing land and resources and we’re seeing increasing strains on power systems.”
Municipal authorities also need to safeguard themselves against cyber vulnerabilities and attacks, which can have devastating consequences. In July this year, a faulty software update triggered what is considered to be the largest IT outage in history, causing chaos at airports, stations, banks, healthcare services and businesses around the world.
Then there are concerns about data quality – data doesn’t typically show a full picture, nor can it be a solution itself. What is more, AI-powered algorithms and machine learning can spew out more data than humans can handle; yet city planners may be ill-equipped to make a proper use of it.
"You can dump as much data as you want on the
table but it's no use if data literacy is lacking,” says Amar Rahman, global head of sustainability and climate solutions at Zurich Resilience Solutions, part of Swiss Re group.
“Data has to be made meaningful to the audience. Visualisation is one way. Data has to be integrated into the story you're trying to tell, to make the risk tangible and personal.”
TKF delegates said that being smart doesn’t necessary mean high tech. A smart and resilient city, they say, requires urban areas with better transport, water, energy and waste management infrastructure, flexible offices, logistics facilities, proper heating, ventilation and air conditioning (HVAC) systems as well as public services from healthcare to education.
“Smart can mean many things. Many answers can be low tech – such as rightly sized infrastructure, which is a more holistic way of responding and improving the quality and well-being for city residents” said Ivo Weinoehrl, senior investment manager for Pictet Asset Management’s Smart City strategy.