Building smart cities that will think for themselves

Imagine a city that ensures deliveries arrive the first time, enables emergency vehicles to get to their destinations by ‘parting’ traffic, and reunites people with their lost pets. This is the next generation of smart cities, also known as ‘cognitive cities’ that think for themselves.

Whilst the first era of smart cities could sense, they were not able to react. With sensors and edge computing distributed through streets, cognitive cities will sense and respond. Many of the future smart cities will be ‘greenfield’, where cities are newly built from the ground up to incorporate intelligence, with edge computing baked into everything including streetlamps and bins. For citizens living in these areas, real, measurable improvements will be delivered to their lives, such as instantly finding parking spaces, and using predictive intelligence to reduce energy bills.

The role of edge in building cognitive cities

When creating a cognitive city, the fundamental need is to move the computing power to where data is generated: where people live, work and travel. That applies whether you’re building a totally new smart city or retrofitting technology to a pre-existing ‘brownfield’ city. Either way, edge is key here. You're dealing with information from sensors in rubbish bins, drains, and cameras in traffic lights. You need to react to these in real time, for example to address water problems or get frontline responders to the site of a traffic incident.

In today’s smart cities, the focus has always been on capturing data: whether that’s for monitoring traffic hotspots or looking for water leaks. But in years to come the city itself will respond dynamically to the changing physical world, adjusting energy use in real-time to respond to the weather, for example.

The evolution of monitoring has come from a machine-to-machine foundation, with the introduction of the Internet of Things (IoT) and now artificial intelligence (AI) becoming transformational in enabling smart technologies to become dynamic. Emerging AI technologies such as large language models will also play a role going forward, making it easy for both city planners and ordinary citizens to interact with the city they live in. Edge will be the key ingredient which gives us effective control of these cities of the future.

To enable this sort of rapid, responsive service, edge is crucial: you need to move the computing power to the streets themselves. It’s part of a wider shift away from single-use analogue sensors, such as traffic or smoke sensors, and into the use of smart cameras, built to generate data but also preserve privacy.

Getting smarter on the road

In the smart cities of the future, technology will be built to respond to human needs. Sustainability is the biggest problem facing cities - and by far the biggest contributor is the automobile. Smart cities will enable the move towards reducing traffic, and towards autonomous vehicles directed efficiently through the streets. Deliveries which are not successful the first time are one example. These are a key driver of congestion, as drivers have to return to the same address repeatedly. In a cognitive city, location data that shows

when a customer is home can be shared anonymously with delivery companies - with their consent - so that more deliveries arrive on the first attempt.

Smart parking will be another important way to reduce congestion and make the streets more efficient. Edge computing nodes will sense empty parking spaces and direct cars there in real-time. They will also be a key enabler for autonomous driving, delivering more data points to autonomous systems in cars. In the smart cities of the future, roads will be designed around autonomy, with vehicle-to-vehicle and vehicle-to-infrastructure communication.

Edge computing can also help to get first responders to the scene of incidents faster. Smart city infrastructure can detect a fire in a building through vision-based sensors and trigger an alarm. Emergency services are alerted, and AI can preconfigure the safest and quickest route to get to the scene, rerouting other vehicles if necessary.

Prioritising privacy

Video is used to offer situational awareness of everything from overflowing bins to traffic movements, rather than for surveillance. Smart cameras will help reunite people with lost pets, for example, with AI identifying the pets as they move between cameras. In any smart city, privacy is the number one concern. The smart cities of the future won’t capture data for the sake of it, but for the purpose of delivering improved services. If citizens are trusting city planners with their information, they need to get back more than they give.

Edge can also help with sustainability in the home. Even the smartest of smart homes detects occupancy and only turns off the air conditioning when people leave. You could gain a huge amount by slowly turning it down in the hour before someone leaves, using sensors and AI to predict this. Cities will use advanced computing technology to monitor real-time activities in buildings, allowing authorities to match energy supply with demand.

The cognitive cities of the future will also provide augmented reality experiences which will help people who are visually or hearing impaired, delivering text-to-speech and voice-to-text with the help of large language models. Edge will be key here too: when it comes to a visually impaired person crossing a road, milliseconds count. Computing cannot be confined to a data centre - in a truly cognitive city, the streets themselves will pulse with data.

The complex computing nodes

When building a smart city from scratch, you’re talking about hundreds of thousands of computing nodes, located all over the city. This requires planning. The nodes need to be part of the built environment, such as inside lamps on the street. Each device also has different requirements around cooling and latency: you can’t just throw a PC in a cabinet. When you build a new smart city, you’re able to integrate these far more effectively, so that they are less visible and easily accessible for service engineers.

But if you’re retrofitting, there are still ways to deliver computing power to where it needs to be, both in devices like smart streetlamps and in cabinets. In cities such as Barcelona, London or Paris, there is an abundance of service cabinets that are providing either mobile or telco services - or even access to water. Service engineers can adapt them to interact with the network. Believe it or not, these are your future data centres.

Human-centric smart cities

The future smart cities will be built with a truly human-centric approach, whether fitted to existing cities or built-in ‘greenfield’ to fully incorporate smart technology. With citizens’ engagement in mind, these large-scale infrastructure projects will bring key foundational blocks together to support the cities with everything, including communication networks, transportation, public safety, and energy efficiency.

To achieve this, it is crucial to ensure edge computing nodes are on every road, enabling the city to ‘think’ for itself by delivering insights harvested from sensors and cameras. The smart cities today are just the beginning, and the cognitive city in the future will create a safer, happier, and more sustainable way to live.