Mott MacDonald builds smart cities on Azure cloud

Applying AI in the cloud to smart city data

Mott MacDonald was an early adopter of Office 365 and, as part of the partnership, it now plans to move the Moata cloud-based analytics and digital twin platform to Azure.

According to Mott MacDonald and Microsoft, the Azure-hosted smart infrastructure platform will combine live data and analytics, AI, and machine learning to enable public services to run efficiently while also keeping the public better informed on environmental risks such as sea water quality or risks to power networks before problems occur.

“It is vital that smart infrastructure delivers social, economic and environmental benefits for society, as well as investment return for businesses,” says Oliver Hawes, Mott MacDonald’s head of smart infrastructure.

“Microsoft and Mott MacDonald believe smart solutions have a positive impact on cities and asset owners, and improve design, delivery and performance of infrastructure. Cities, government, asset owners, investors and society can all benefit hugely.”

Smart city infrastructure relies on data and networks of sensors to collect telemetry across a city. “We need to make information available to everyone, everywhere,” says Richard Shennan, group digital business development director at Mott MacDonald. For this data sharing to work, he says citizens need to agree to have data collected – and this data needs to be anonymised.

Given the infrastructure in cities can sometimes be extremely old, Shennan says smart infrastructure can be deployed, which enables those people managing this infrastructure to understand how to get the best out of what they have.

From the citizens’ perspective, he says a cloud-based platform also provides a social interface, which can be used to encourage behaviour change – such as raising awareness and using gamification to encourage informed behaviour changes, which can then lead to a reduction in demand, such as helping to reduce water wastage.

Improving water quality in Auckland

The city of Auckland in New Zealand is an example of where Mott MacDonald has deployed its smart city platform to improve the services the city provides to residents.

“We started work for the city council on traditional infrastructure, consulting on water and waste water networks. We then built a digital model to enable them to make better decisions,” says Shennan. He says this was then moved onto the Mott MacDonald Moata platform.

Sensors were added for live data, enabling city planners to identify trends and run predictive maintenance using machine learning. “We looked at the impact of specific events that would cause a deterioration in water quality,” he says,

The city-scale digital twin represents the real-time interaction between atmospheric conditions, the urban storm water and wastewater networks, and the marine environment. “We built a model with incoming rainfall, plus a live data feed covering data such as ocean currents, wind direction and speed, which we could then use to predict the quality of water,” says Shennan.

The digital model has enabled the city to improve the accuracy of water quality predictions and the associated public health risk from less than 20% to greater than 80%.

By combining engineering principles with modern AI, Mott MacDonald says the Moata smart infrastructure platform has helped Auckland’s council increase levels of engagement with the public to make better-informed decisions, driving improved outcomes for the community.

Building smart cities on a public cloud

Public clouds, like Azure, provide a way for IoT data from cities to be collected and processed. Digital twin models running in the public cloud can be updated with live data from IoT sensor data from real-world infrastructure to improve the accuracy of the simulations.

As an example, Trudy Norris-Grey, global managing director of local and regional government and City Next at Microsoft, says: “The use of IoT means we can detect potential for leakage in pipes before it happens.”

This is made possible by running a digital twin in the Azure cloud, which Norris-Grey says provides an instantiation of the infrastructure.

By using IoT sensors to update the digital twin model with real world data, the simulation enables city planners to optimise predictive maintenance schedules and run “what-if” scenarios.