Awaab’s Law: How AI holds the key to tackling deadly damp and mould in social housing

The death of Awaab Ishak put the state of damp and mould in social housing in the spotlight and a new law in his name is now forcing social landlords to fix properties more quickly. It might take tech and artificial intelligence (AI) to keep up with demand.

The Social Housing Regulation Act came into force in April requiring landlords to deal with hazards within strict fixed timeframes. The former Conservative government’s move earned praise from social housing campaigner Kwajo Tweneboa who said he hoped the legislation would mean a “more proactive stance in scrutinising providers”.  The Labour government plans to extend Awaab’s Law to the private rented sector under the Renters’ Rights Bill.

Under-resourced councils are already struggling to deal with complaints and caseloads, and housing associations are also failing to keep pace with a problem that is rife in Europe’s oldest housing stock.

As the new law came into force, the Housing Ombudsman, who investigated a record-high number of overall complaints last year, highlighted failings from a number of landlords, including one who left a family living in damp and mould for five years.

But, perhaps surprisingly, AI and machine learning could play a pivotal role in helping social landlords deal with their existing cases and predict and prevent damp and mould in their properties.

Tech firm Mobysofthas two products: RentSense, which is designed to help landlords manage rent arrears, and RepairSense, which does the same for repairs.

It launched a damp and mould-specific module of RepairSense in January, working with 10 landlords.

RepairSense uses various machine learning techniques to understand where properties are having ongoing issues with faults and repairs.

It is based on complaints – the tool notifies housing associations and tradesmen carrying out work when a complaint is raised by a tenant.

Using information from jobs carried out on the property, the module is intended to help landlords prioritise work, get to the root of the issue faster and provide a lasting fix rather than sticking-plaster solutions.

“If you’ve got 3,000 properties that are experiencing damp and mould, how do you start to attack that problem?” says Natalie Tuer, head of product for repairs solutions at Mobysoft.

“In order to manage that, they need a system to do it, which they don’t have, it doesn’t exist. So they’re doing it on spreadsheets and individuals in these organisations are doing their best to try and manage what can be thousands of properties, having lots of inspections and lots of repair jobs and ongoing cases.

“And so we’ve started to learn a lot about this from the different organisations and how they try to manage it now and what would be the best solution. That’s where our module has come into play.”

The module has a “three-phase attack” against damp and mould.

Firstly, it will use data from a tenant’s complaint to raise a case at the property and create a chain of cases showing events at the home over the last three years at a glance.

Secondly, the tool can flag any properties where landlords have failed to keep up with their own processes – where work is overdue, for example – so they can prioritise work.

Finally, the module uses historical data, asset data and modelling to predict properties where there may be a problem. Tuer says it can do this to 90% accuracy.

This gives landlords the information they need to react, she adds. Although without the resources to carry out work and take action, damp and mould will remain.

“Without technology, I can’t imagine how you’d be able to respond to that part of the regulation,” says Tuer.

“It can model hundreds of thousands of rows of data, which we just couldn’t do as human beings in a few minutes, and identify whether there’s a possible problem then.

“But without having hundreds of thousands of rows of rich data that you can learn from, I’m not sure how you would predict that information.”

Switchee takes a different approach to preventing damp and mould, using smart devices to monitor data inside and outside properties.

The tech firm works with about 130 housing associations with devices installed in about 35,000 properties. Devices are currently being deployed at a rate of around 3,000 social homes a month. 

The smart device, which residents agree to have installed when they sign their tenancy agreement, replaces the thermostat and allows the collection of data to seek out the conditions for damp and mould to form.

“Typically what we are looking for is for the moisture content or the humidity inside the property. We take a bunch of weather station data and from that we know what the external air temperature is,” explains Switchee chief executive Tom Robins. 

“We’ve got some data on the wall type of the property, if we don’t have it, we can make some gross assumptions, but we typically got a good understanding of what the age of the property is, what the wall type of the property is and, therefore, what the thermal boundary is.

“If you know the moisture content, the air’s ability to hold that moisture, the temperature of the wall, you’ve got a really good accurate indication of whether moisture is going to be forming on a wall at a temperature that is conducive to creating an environment for mould.”

Switchee formed in 2015 to fight what Robins calls “a really simple injustice” of making technology to slash heating bills available to people on lower incomes, not just higher earners.

The smart device’s algorithms can identify smart heating schedules to slash around 17% of heating bills with tenants able to work with Switchee or control their heating with an app to plot the most efficient way of keeping the property warm.

The idea is that the data can be used to proactively trigger a conversation between tenants and landlords to prevent damp and mould forming in the first place.

Robins insists a smarter use of data is key to triggering a shift in how housing hazards are handled in the first place.

“What we’re trying to do here is change the operating model of the sector, from reactive to proactive,” he says.

“Housing operates on complaints. What other service industry would operate on complaints? The only data points we have are when residents complain and you’re totally dependent on the residents that do complain. That’s why we see a lot of people suffering in silence.”

Switchee responded to the government consultation on Awaab’s Law to warn of the risk of “baking in” reactivity in response to damp and mould.

Robins says it will take a “seismic shift” to change how landlords operate and data, AI and other tech responses can play their part.

“I do believe Awaab’s Law has got the best intent, but what we’re going to end up with is a higher threshold and a lot more mould washes that are short-term interventions and don’t fundamentally solve the problems that we see in the field. We are going to make the system more reactive,” he says.

“We can solve this, we can actually just understand what’s going on. But there’s such a seismic shift in every other part of the way that housing operates. It is a complete change in operating model to move from reactive to proactive.

“It’s not going to happen fast, but I do really believe it needs to happen.”

Do you have a story to tell or opinions to share about this? Get in touch and tell us more. Big Issue exists to give homeless and marginalised people the opportunity to earn an income. To support our work buy a copy of the magazine or get the app from the App Store or Google Play.