Data-driven decisions for better city planning and management

From siloed data sets to AI-enabled insights

We are pleased to be helping Auckland Council get the most out of its assets and use data to inform decision-making, with the responsible use of AI playing an important role.

With the amalgamation of eight councils into the “super city” 14 years ago, Auckland Council’s Healthy Waters team set about improving the collection, standardisation and management of their data sets.

One of the challenges encountered during the amalgamation process was not the deployment of new technologies, but identifying, validating, organising and assessing the quality of the data from many siloed data sets from across the councils.

Surfacing the data naturally exposed quality issues which required iterative improvements to make it technology ready. Auckland Council needed a better way to bring important datasets together to better serve the citizens of Auckland.

Fast forward to 2025 and the strong data management discipline that Auckland Council has developed is proving valuable in the age of AI. Today, the team has implemented several AI and machine learning (ML) applications that are enhancing service delivery and greater value to Aucklanders.

Image processing for pipe blockages

We worked alongside Lynker Analytics to support Auckland Council to develop a system that combines simple wildlife outdoor cameras with AI to capture, consolidate and analyse pipe blockages or flooding at the city’s most critical stormwater structures.

By detecting blockages early, the system will enable a more sustainable and resilient stormwater network and improve the productivity of the maintenance teams, reducing costs and minimise disruptions to the public. Previously all of these discharge sites were manually checked on rotation, especially before large rainfall events.

Pipe condition prediction

Using the existing dataset of stormwater pipe inspections, our team developed a machine learning model that assesses the condition grade of pipes over their lifespan. This hybrid machine learning model predicts the age at which a pipe will deteriorate, helping Auckland Council to better plan their asset renewals and target this spending more effectively.

Flood warning systems

Auckland Council are now working to develop a solution to predict surface flooding in real-time through machine learning that uses data generated from existing hydraulic models. We are using these existing modelling assets in new ways with the aspiration of enabling the region to create relevant and targeted flood alerts that could be used for emergency response and ultimately for community warning.

The advances that Auckland is now realising through AI technologies have been made possible through the hard work put in over the past decade to organise and enhance data and build an ecosystem of aligned service providers. This approach has driven effective collaboration across the supply chain ecosystem using digital mediums to develop and deploy these new services, leveraging an organised and well-managed data system.