Urban AI: three firms on how AI can help us understand our cities

Artificial intelligence promises large productivity and efficiency gains for our urban spaces. We speak to three leading firms working with Connected Places Catapult about their systems and their thoughts on the future of AI.

Intelligent AI

Imagine a future where every building in the country is mapped in 3D on a digital twin, and owners can request a detailed report pointing out its risks, flaws and inefficiencies in order to make financial and carbon savings.

Where materials used in its construction are carefully noted and a building’s current condition are tracked to allow occupants to intervene in order to lower their energy bills and reduce water consumption. And where clients are offered personalised sustainability advice using Large Language Models.

The vision may be a few years away from being realised, but is one that Anthony Peake, the Founder of Intelligent AI, is focused on helping to deliver.

“We want to get to a position where we know the layout of every room in every single building, and have a platform that can automatically calculate the carbon footprint of every building,” he says. “Most people don’t have good data about their properties, so my aim is to deliver it to them.”
Anthony Peake, Founder of Intelligent AI

In recent years, the company has been working with insurance underwriters, and risk engineers, to develop what Anthony calls a ‘digital twin of risk’ for commercial properties. Several products have been launched by Intelligent AI that calculate the risk scores and building replacement costs and allow more accurate premiums to be generated for clients.

It has created an AI engine that reads unstructured technical documents associated with commercial building surveys and extracts around 200 pieces of data necessary for insurers to allow them to determine the risks to a property in a fraction of the time.

“Really useful information detailing the construction and condition of a building and whether it has water sprinklers or alarm systems, for instance, can be hard to find in a document; often hidden in tables or random sentences. It might take a person an hour to pull out just 18 pieces of useful information from a 60 page report, but our system can extract all of the data in a second.”

Early in his career, Anthony worked for tech giants Apple and Oracle before starting a software development company and working alongside insurers. Prior to founding Intelligent AI, he also ran a project to develop a machine learning platform focused on automating council planning applications.

Just over a year ago, the company secured £15,000 from Connected Places Catapult for a project to help refine its offer for commercial property owners, managers and insurers.

“It has been great to be able to lean on the shoulders of giants to validate our offer and to establish what we need to do to deliver the most value,” he says.

Artificial intelligence is not going to replace people, Anthony adds. “As humans, 80% of what we do at work is admin. So rather than reading 80-page reports and writing another report looking at the risks, people should have more time to work with clients on strategies to reduce the risks.

“AI will help to free more people up so they can add value to their work, their companies and especially their customers.”


Recent years have seen software specialist Gentian refine an AI product that determines the quality of green roofs in urban areas, and allows for recommendations to be made to property owners about ways to improve biodiversity on the tops of buildings.

Now the technology start-up is looking to diversify its offer to monitor the quality of private gardens on the ground and tree cover in cities, as well as map the natural habitats of rural locations.

In January, the company secured a grant of £15,000 and business support from Connected Places Catapult to find ways of working more with the financial services and built environment sectors to map green roofs, and to identify and develop new use cases for the technology.

Data about the quality of flora coverage on roofs, in gardens and in the street is gathered using advanced satellite imagery, before the AI tool provides an assessment about the quality of the green coverage.

“This level of data would normally be gathered by an ecologist, spending time recording lots of information manually. But we offer the possibility of doing all that using just one snapshot from space, and showing how greenery or nature coverage is changing over time.”
Gentian’s Vice President of Business Development, Jamie Rhodes

“We cannot replicate some of the really detailed surveys that ecologists carry out, such as surveys of mammals, but we are there to help them focus their efforts on where they are most needed.”

The AI model has been taught what constitutes good quality biodiversity using ‘ground truth data’ collected by ecologists and professionals such as ‘green roof guru’ and company co-founder Dusty Gedge.

When it comes to green roofs, Gentian Lead Technologist Marcus Spiegel explains that the AI model not only characterises the components of a green roof identified – but also assesses the level of biodiversity present (using indicators such as growth of wildflowers) and allocates a score. The model can also identify roofs that may be suitable for greening.

“There is a clear demand for these kind of remote assessments,” he says. “So far we have demonstrated what we can do in different contexts, and this year we will look to scale our offer to help people improve what they are doing.”

One reason for an increased interest in green roofs is that developers in England will soon have to meet ‘biodiversity net gain’ targets. In addition, new recommendations from an international initiative known as TNFD – or Taskforce on Nature-related Financial Disclosures – encourage businesses to embrace nature.

Jamie Rhodes adds: “Our value proposition is that we can help businesses meet global regulations. Reporting on nature is going to become a massive market.” Better quality greenery – and more of it – could also help in the fight against climate change, he continues.

With so many satellites flying around the earth, far more data is being produced than humans are ever going to be able to analyse by themselves, adds Marcus. “Programmes like ours provide a way of analysing all this incredibly valuable data.

“When it comes to the environment, AI presents more of an opportunity than a threat.”

Slingshot Simulations

Use of data and analytics in place-based decision making is 25 years behind product design and development around aircraft engines, mobile phones and self-driving cars, according to Slingshot Simulations Chief Executive Robert Harwood.

But thanks to the growing adoption of artificial intelligence, advanced modelling and simulation, and digital twins in place-based applications – such as transport, energy management and environmental monitoring – “the revolution is coming, the software technologies are emerging and the required integration of models and simulations based on vast amounts of data is happening,” he says.

“These technologies need to be embraced, and people need to get on board. If they don’t they not only risk being left behind, but highly complex and interdependent policy investment decisions we need in relation to decarbonisation, net zero and climate resilience will simply not be made at the pace they need to be made.”
Slingshot Simulations Chief Executive, Robert Harwood

Slingshot Simulations’ software allows regional policy makers and their consultants to connect data and analytics around the built environment, to help them to rapidly prioritise interventions that support the development and implementation of regional strategies around local transport, energy and climate plans – and understand how they will likely evolve over time.

“Fundamentally we help them to make better decisions faster and at lower cost, by ensuring the data and analytics can be easily understood. We help them to bring the siloed information together to see what policy options work best for the benefit of the environment, regeneration and growth,” Robert says.

He adds that rapid advances in AI seen recently allow results to be generated and interrogated more quickly; meaning that an activity that may previously have taken weeks or months could be completed in hours or days.

The company’s ‘decision intelligence’ software platform Compass:Engine, for instance, has been used to show how flood and traffic simulations can be combined to develop a more climate resilient transport network, while identifying vulnerable assets. It has also been used to show the impact of displaced traffic and emissions from introducing Low Emission Zones – and featured in the Transport Research and Innovation Board Digital Twin Roadmap 2035, delivered by Connected Places Catapult on behalf of the Department for Transport.

Other projects include helping authorities to make decisions around ‘school street’ initiatives; plan for climate adaptation through future weather scenario modelling; map urban air quality; and carry out feasibility assessments for converting park and ride sites into net zero mobility hubs.

“As regions develop their strategic plans with net zero, decarbonisation, climate resilience and economic growth at their centre, they want to better understand future forecasts and any risks to people, property assets and infrastructure,” Robert says. “When we start bringing all these capabilities together, we can solve a wide range of issues.

“Our mission is to enable decision makers to make use of an exponentially increasing volume of data and accessible analytics such as AI so that they can develop a holistic view and break down decision making silos.

“In the context of addressing the place-based impacts of the climate emergency, the responsible use of AI presents a step change opportunity for us to move at the speed we need to.”