)
)
Nearmap was proud to sponsor AI.Con 2025, and brought to you an exclusive workshop, presented by Dr. Mike Bewley, VP of AI and Computer Vision at Nearmap.
We explored how artificial intelligence is transforming the way we understand, map, and measure the evolving urban landscape.
In this interactive session, Mike pulled back the curtain on Nearmap AI technology, showcasing how petabyte-scale aerial imagery, data science, and machine learning converge to deliver powerful, actionable insights into how cities change over time.
Curious about the future of geospatial intelligence after joining Mike’s workshop? Contact Mike at michael.bewley@nearmap.com to continue the conversation — just be sure to include your name, role, and organisation in your email
Because the event was 5 years ago (before the formal ImpactResponse program and realtime damage classification products), commercial usage would look very different for an event today. For recent major events (remember these are global products), Hurricanes Frances, Helene and Milton, and the LA Wildfires had a number of the largest US property insurers and disaster response agencies using the imagery and AI to respond in realtime to the event.
There are many! We have studies Australia’s leafiest suburbs. Nearmap AI data was also featured in an ABC story which touched on leafy suburbs and socioeconomic status.
Nearmap camera systems are designed and built exclusively by Nearmap and expertly installed in aircraft (piloted fixed-wing, not drones) to capture aerial location intelligence covering 95% of Australia’s populated areas. No other aerial imagery provider in the world has the same vision and data technology stack.
Nearmap is subject to air traffic control in terms of where and when to fly, and at times we are restricted from flying certain locations. The imagery resolution is in the 4-7cm per pixel range, which means personally identifiable features are not visible. We track assets, not people. We focus on understanding the urban environment and infrastructure, where communities are made up of homes, businesses, and the natural environment, and always comply with local laws and regulations.
More information on our risk models (which use our semantic maps as inputs to calculate risks at a per-property or building level) can be found here.
Our vision is ambitious — we want to accurately model the changing nature of critical assets in the urban world — from roof condition to changes in tree cover, ground surfaces and environmental risks. Further, we have been building out end-to-end applications for the property insurance industry to ensure clean integrations, and including other 3rd party information as needed to richly satisfy use cases.
They are trained to work extremely well for commercial properties, and we have a range of customers using our data and apps for that purpose. However, it is worth noting that commercial properties tend to be much more complex than residential ones, so a higher portion of the errors in the data will naturally be present in a commercial complex compared to a simple freestanding home.
There are so many factors involved in repairs, and we wouldn’t want to comment on what standard should be expected. We aim to represent the truth on the ground, so industry experts can understand and improve their businesses. Another suburb with a similar event was the Christmas hailstorms in Berowra (Sydney).
We have captured post-catastrophe imagery of a range of floods to help show the broad scale community impact. We also have digital surface models, digital terrain models, impervious surface Nearmap AI layers, Which can allow customers to do fairly detailed flood modelling.
Learn about the AI-derived insights and location intelligence solutions that are helping insurers write more profitable business, reduce expenses, and mitigate risk — with tailored insights for underwriting, claims and more.
)
)