Tackling damp and mould using AI

Like many housing associations with thousands of homes to look after, Greatwell Homes admits that dealing with damp and mould cases can be complex and invariably sometimes things can slip through the net. In June 2024, the percentage of households with cases of damp and mould was 16%. In 2025, Greatwell Homes, with over 5,000 homes across Wellingborough and Northampton, says the rate has dropped significantly to just 2%. 

To achieve this, it didn’t change its property portfolio, nor did it go on a hiring spree. Instead, in June last year, the housing provider implemented predictive analytics software to combine its existing data with machine learning and AI to spot damp and mould issues and predict where such issues might arise in the future. Instead of being reactive, it means it is being proactive and can solve issues before they escalate. 

“The key factor for us was about minimising risk,” says James Norton, repairs and maintenance manager at Greatwell Homes. “We are aware, naturally, of the media coverage in our sector, especially around damp and mould. What we wanted was an overarching, wraparound system, which would pick up where there could be any opportunity for problems.” 

Tracking repairs and maintenance with AI 

The system, RepairSense by Mobysoft, tracks all repairs raised, then screens the severity and geographical risk. For example, if there are several flats that have had an issue, it will flag up that the others might be at risk. 

The software is also able to quickly look back at the history of the block. “We have just over 5,000 properties and one [damp] case might have called six months ago,” Mr Norton explains. “It can really look into places on the system that a human might not without downloading multiple spreadsheets and joining the dots up.” 

Also, he says it can create systematic follow-ups, tracking cancellations and unanswered visits to make sure the customer doesn’t fall through the cracks among the 300 repairs delivered a week.  

“It will look through the data and do all the relevant checks to make sure that jobs have been booked in within the relevant time frames,” says Greatwell Homes’ Lisa Harley, a senior planner managing a team who work closely with customers and operatives, ensuring repairs are scheduled for a convenient time. 

“Then it will look at your past history, for example how many times you’ve called in, and suggest what the next step might be,” she says. “You might have only moved in last month, and a previous customer had issues, and so for us to organise a mould wash is pointless. You actually need a surveyor, or you need X, Y and Z, and so it will list the suggestions.” 

At Notting Hill Genesis, a social housing provider with 67,000 properties in London, the organisation is trialling detecting and predicting issues such as damp and mould using Google Gemini and OpenAI. 

For example, it can detect damp and mould issues when residents send in pictures of issues in their property, look at patterns of what has happened in the past, and predict what might happen in the future. 

“For example, with a high degree of accuracy it can say this property next summer might have this kind of problem, so it pre-empts and is getting more and more accurate with what will happen and when, and we can line up the resources to deal with that problem way before someone notices it,” says its chief information officer Rajiv Peter. 

“My objective is that residents should not need to call us,” says Mr Peter. “Instead, we deal with the problems before they contact us. Over time there should be less spent on maintenance. Costs will be lower, and we will have a higher surplus to reinvest in more homes.” 

Predicting and forecasting problems 

Every 24 hours, the Mobysoft RepairSense system drills through the last three years of Greatwell Homes’ damp and mould data across its portfolio of properties and is able to forecast potential problems, so that Greatwell can action preventative works. The machine learning is able to sift through all the data more than a person and flag up where there might be wider problems in the block, that a human might have missed. 

Although talk of AI often comes with fears of job losses, Mr Norton says it’s actually the opposite. “It’s meant that we can actually do our work,” he says. “Rather than visiting a property three times and spending £30 a time wiping down some mould, we go out there once,” Mr Norton says. “We might spend a thousand pounds, but we’re rectifying the issue with the gutter, for example, and it’s a resolved case. Although it’s costing us more in the long run, it’s resolved the issue.”

Mr Norton advises other housing associations not to be afraid of AI. “There’s a massive stigma around it,” he says. “People think that it could take jobs and all the rest. But it’s a very user-friendly system. And if you’re using it regularly, and using it correctly, it’s fairly straightforward to keep on top of. Now everyone who uses it [at Greatwell Homes] can really understand the benefit around helping minimise safety risks. It’s actually saving time.” 

But for Mr Norton, one of the most positive upshots is that it’s given him and his team much more assurance. “It makes me feel a lot calmer and I have the confidence that we’ve gone through all of our processes, because the system ensures that we go through them,” he says. “When you’re working with data, you’re working with facts, which makes it a lot easier to be confident when speaking with our customers. We’re not saying things can’t go wrong, but it just gives you much more reassurance.” 

Subscribe to Inside Housing Management and sign up to the newsletter

Inside Housing Management is the go-to source for learning, information and ideas for housing managers.

Subscribe here to read the articles.

Already have an account? Click here to manage your newsletters.