Finding a balance between pluses and minuses of AI in sustainable cities

A scheme to use #AI to prioritise cyclists over cars at traffic light has grabbed the headlines. AI has huge potential to aid #urban #sustainability but will involve trade-offs between economics, society and the #environment.

Solihull, a town in the UK’s West Midlands, this month unveiled a pilot schemes that uses smart sensor technology to make a busy road crossing safer for cyclists and pedestrians.

The sensors made by VivaCity use intelligent video analytics and advanced algorithms to detect cyclists from up to 30 metres away, triggering the lights at a crossing of the A34 road to turn red for motorists. This early detection enables the traffic signals to go green quicker, giving cyclists a “smoother, uninterrupted journey” on their bikes.

The initiative will be welcomed by cycling campaigners and environmental groups alike for its potential to encourage active travel and environmentally friendly travel while reducing the number of car journeys.

This is likely to deliver secondary benefits such as cleaner air and support the goal of urban sustainability that is often defined as enabling its citizens to meet their current needs without harming the condition of future generations. That goal seen as based on the pillars of economic equity, social resilience, environmental gains and good governance.

This balance is hard to achieve at the best of times, particularly when implementing measures that favour the interests of one group over another. Inevitably the pilot, which could serve as a model for future projects in the West Midlands and perhaps further afield, has encountered some criticism.

One report cited privacy campaigners angry at the widespread use of AI cameras against motorists although they pointed to their use to spot drivers breaking the law on the use of mobile phones while driving. The Solihull scheme could also perhaps be criticised for causing cars to stop more frequently and for longer, creating more delays and more emissions while idling at red lights.

The point is that AI can be a powerful tool to impact human behaviour, either to enhance or restrict what we do. For instance staying within road transport, AI has been used to improve the flow of cars at junctions. An AI system developed by Aston University was developed to read live camera footage and adapt the lights to compensate, keeping the traffic flowing and reducing congestion in a dozen cities in four continents.

WIRED magazine has looked at how Google AI claims that its analysis of changes to traffic light timings could cut as many as 30 per cent of stops and 10 per cent of emissions for 30 million cars a month. On the other hand, it could be said that a potential downside is that by improving traffic flow, it makes driving more popular thus creating further problems down the road, as it were.

There is no doubt that AI tools are very powerful, as anyone who has asked it to write a letter and seen the fluent result delivered in less than a second will know. That has led to a debate over the potential input that a “smart city” can make towards sustainability goals on the one hand, and the risks and negatives impacts on the other.

A quick summary is that AI presents an exciting chance to create intelligent systems that can produce vital knowledge for protecting and sustaining the quality of city life. However, more effort is needed to limit the ethical and governance concerns raised by the greater use of AI, for instance by diverting resources such as traffic management away from other areas or by benefitting one group (drivers) over others (cyclists, pedestrians).

Given that more than two-thirds of people will live in cities by 2050, according to United Nations forecasts, there is an urgent need to ensure that cities are sustainable organisms that do not add to the environmental burden of future generations. AI clearly has the potential to reduce that threat, but at what cost?

UN-Habitat, the UN agency that has led much of the debate and research into urban sustainability is now working to help councils harness digital technologies while ensuring marginalised and vulnerable groups are included in smart cities and digital transformation processes

It is working on drawing up international guidelines on people-centred smart cities to be approved as a non-binding framework by the end of next year.

Technologists always move faster than regulators, but the hope is that policymakers can attempt to catch up by taking a proactive, rather than reactive stance, and ensure that cities can balance the needs of different groups of their citizens.