Ordnance Survey used Microsoft AI to ‘see’ roofs – and it could save you money

Artificial intelligence could soon help homeowners save money on insuring their house, after Ordnance Survey (OS) taught Microsoft technology to identify roofs.

People who own properties with flat roofs can be charged more for insurance due to potential problems with sagging or damp, as rainwater may not run off properly.

OS said there was a 5% error rate when humans used maps to identify roof types, so if homes are judged to have flat roof when they do not, this could lead to people paying more for home insurance policies.

According to the AA’s British Insurance Premium Index, the home insurance shoparound quote (an average of the five cheapest quotes identified) hit £117.11 per annum in the last three months of 2017. It also revealed that the costliest region to insure a home is London, with a combined contents and buildings policy of £172.26.

It is hoped that by uploading images of roof types to AI programs, the technology can be used to avoid errors and save people money. A recent trial by OS using Microsoft’s machine learning technology to identify different roofs saw the AI’s accuracy rise from 0% to 87% in just a week.

While OS admitted there was further work to do, the mapping agency was excited about the potential of using AI in its work.

“We now have a machine training system that can be trained to classify almost anything, such as bodies of water for example, and on top of this, the machine can be set to run hundreds of these classification queries simultaneously,” said Isabel Sargent, Senior Research and Development Scientist at OS.

“This has the possibility to significantly increase the efficiency of our surveying operations, freeing up the surveyors we have on the ground to focus on those complex tasks for which there is no AI assistance. We would not have been able to achieve this breakthrough so quickly without Microsoft, and we’re thrilled for the future by the potential of what our collaboration has achieved.”

Everything you need to know about AI

In the time it takes a person to identify roof types in a single image, an AI algorithm can process thousands of images. It is thought it would take Microsoft’s technology less than a day to classify all 35.7 million properties in Britain and their roof types.

Microsoft’s computer vision API is being used by companies to analyse images, read text in pictures and recognise famous people and landmarks.

The technology firm is one of the global leaders in AI. Since the creation of the company’s research labs in the early 1990s, it has pioneered innovations in computer vision, speech recognition, natural language processing and machine learning with the goal of developing systems that anticipate people’s needs instead of responding solely to our commands.

“The hack we’ve conducted with OS demonstrates the potential benefits that AI can bring to people in a broad mix of roles,” said Gina Dragulin, Audience Evangelist at Microsoft. “Our own research found that workers believe more than a third (36%) of tasks they carry out on a regular basis would benefit from intelligent automation.

“The OS hack is a perfect example of the value that could be created for OS surveyors if they were able to eliminate time-intensive tasks, such as roof-type identification, from their day-to-day activities, enabling them to focus on more complex tasks that require the type of high-level decision making and judgment calls that only an expert can make.”

The OS has previously used machine learning to identify and map more than 230,000 miles of England’s farmland hedges.