How AI is helping landlords act faster on damp and mould

As an arm’s-length management organisation responsible for managing most of the council homes across the city of Wolverhampton, we’ve been exploring the potential to use AI and machine learning to identify households most at risk from damp and mould.

In 2023, we took part in a pilot project with NEC Housing to find out if the technology could help us understand which factors gave certain types of properties a higher risk of damp and mould.

We also wanted to be able to prioritise repairs and maintenance in properties where our most vulnerable residents were living, before the issue became a significant problem.

Using AI, we analysed the information on residents and the location and condition of our housing assets. We also looked at external data sources, such as weather and flood patterns, to narrow down those properties most likely to be affected by damp and mould.

“AI technology was able to successfully predict the presence of damp and mould in high-risk properties with 98% accuracy”

This gave us a list of around 100 properties which we could then compare against our records, including previous reports from residents and housing officers, complaints and repairs which were required as a result of damp and mould.

We found that the AI technology was able to successfully predict the presence of damp and mould in high-risk properties with 98% accuracy.

We could see from the results that when the AI predictions are based on out-of-date or incomplete information, the rate of success is much lower. Accuracy levels from the pilot dropped to around 70% in these circumstances. This underlines how important data accuracy is for forecasting problems with damp and mould.

Where housing providers can reduce their reliance on manual re-keying of data, this could make a real difference, as it reduces errors and allows records to be updated more quickly. Our housing officers can automatically update information on our system while visiting residents, for example.

This means that if there is a new baby in the household or a sudden change of circumstances, such as a job loss or a family member with an unexpected health issue, the risk factors for vulnerability to damp and mould may also change.

“The data made available to us through the use of AI has enabled us to fix problems earlier, prioritise homes for retrofit works and put measures in place such as improved ventilation to keep our residents’ homes safe and comfortable”

If a resident calls in to report damp and mould, this information is instantly updated on their record, which can affect their risk scoring, too. A score that falls too low may trigger a priority inspection or a repair.

AI has helped us to successfully identify damp and mould risk in about two-thirds of our 21,000 homes, with accuracy levels increasing where sufficient data is available for each property.

The data made available to us through the use of AI has enabled us to fix problems earlier, prioritise homes for retrofit works and put measures in place, such as improved ventilation, to keep our residents’ homes safe and comfortable. The success of this pilot project suggests that the use of predictive technology could transform the way social housing providers manage and prevent damp and mould.

By taking a more proactive approach, vulnerable residents get the support they need. As AI systems become more refined and data-collection processes improve, there could be the potential to use the technology to boost energy efficiency or predict maintenance needs for issues like structural safety and heating systems.

The use of AI and machine learning is not just about being able to react faster when issues arise. These tools could help housing providers change the way they support people and create safer, healthier communities.  

Sam Dugmore is systems development and support manager at Wolverhampton Homes, Wolverhampton Council’s ALMO