Test Pilot of using AI in Housing

One of the most persistent and costly challenges in social housing is damp and mould.  These issues not only damage properties but also pose serious health risks to residents, particularly those with respiratory conditions, the elderly, and children. According to the Scottish House Condition Survey 2022, around 9 percent of homes still report mould issues, and 49 percent have disrepair affecting essential elements.  Media coverage and tenant complaints have amplified public awareness of the risks, placing further pressure on housing associations to improve their response.

At the same time, social housing tenants often lack access to digital tools that could support them to monitor their homes or communicate issues efficiently.  A significant portion rely on mobile phones and do not have regular access to desktop devices or broadband internet.  This creates additional barriers to engagement, particularly when services are not designed with their connectivity or digital skills in mind.

Housing providers are also facing increasing expectations from regulators and tenants to evidence how they are improving service delivery.  However, the sector is known for its cautious approach to digital innovation, and concerns about data privacy, security, and change management often slow adoption.

The Hubble project was conceived to address all these interlinked issues.  It aimed to create a unified, AI-powered platform that could bring together fragmented datasets, empower tenants with real-time insights about their living conditions, and provide housing staff with early warnings of emerging problems such as damp or mould. Importantly, the platform needed to be accessible via mobile devices and operable even in households with limited broadband access, making Glasgow’s robust 5G network an essential component of the solution.

The solution focused on accessibility, ensuring that it could be used by tenants with varying levels of digital literacy and different connectivity situations.  With over 70 percent of Craigdale tenants relying on smartphones for internet access, Hubble was optimised for mobile use. For tenants, the platform presented a user-friendly dashboard allowing them to monitor conditions in their homes, receive automated alerts about risks like damp and mould, and communicate with their housing provider.  Educational tools and self-help resources were also made available, supporting tenant confidence and ownership over their living conditions.

Alpaca sensors were deployed in ten homes, measuring humidity, temperature, and air quality. These sensors transmitted data via Glasgow’s high-speed 5G infrastructure to ensure reliable, near-instant feedback.  Hubble’s AI engine analysed the data and flagged properties at risk of mould before visible signs appeared.  This enabled housing staff to schedule targeted inspections or maintenance before problems escalated, significantly improving service efficiency and tenant wellbeing.

The UNITY offer includes: 1 x 5G hub and 5 x sensors to be installed in the bathroom, kitchen, living room and up to two bedrooms. The UNITY hub requires to be connected to mains and uses 1kWh of electricity per year.  The sensors are powered by 3v lithium batteries, which require to be changed every five years.  The Hubs are installed in a safe location near an electric outlet in the tenant's home, so pose no risk to the tenants.  The data is sent directly to the Hubble dashboard. 

For housing officers, the platform provided an intuitive dashboard displaying aggregated data across properties, with built-in tools to identify high-risk cases, recurring issues, and patterns linked to building age or materials. Staff could filter data by location, property type, or reported issue to improve decision-making and reduce reliance on reactive, complaint-based interventions.

The project included training for both tenants and staff to support platform adoption. Workshops and digital literacy sessions were delivered alongside support materials to ensure that all users could confidently interact with the system.  Hubble’s layered design ensured that users with minimal tech experience could still gain value, while power users could access deeper functionality.

Real-world feedback loops were embedded throughout the pilot.  Structured interviews and app usage analytics allowed the team to refine the platform based on user behaviour and suggestions.  Feedback from tenants confirmed that the platform helped them understand damp and mould risks better, while staff reported improved ability to triage maintenance requests.

Overall, the solution provided a scalable model for how digital infrastructure, AI, and tenant engagement can work together to improve outcomes in social housing. By leveraging Glasgow’s fast 5G network, the platform ensured uninterrupted data flow and laid the groundwork for future integration with additional IoT systems and other housing associations across the Glasgow City Region.

Image

“This project has been brilliant, it has been educational, can we have more sensors for other things I please? It feels better to be able to self-serve and find solutions, rather than having to ring the office for everything.”  Participating Tenant. 

The commercial proposition for wider rollout is a Software-as-a-Service (SaaS) model with tiered pricing based on the number of properties monitored and the level of integration required.  Core platform access includes data aggregation tools, the tenant engagement app, and the AI risk analytics dashboard.  Optional add-ons include IoT sensor bundles, extended analytics packages, training support, and integration services for legacy housing management systems.

To reduce barriers for smaller housing associations, the team has proposed flexible licensing and volume discounts.  This supports the goal of increasing adoption across a wide range of housing providers, including those who may not have previously had the capacity to explore digital transformation initiatives.  The model also enables cost savings through more proactive maintenance, reduced staff time on repeat issues, and fewer tenant complaints related to damp and mould.

There is also commercial potential for wider public sector collaboration.  The project team has begun discussions with the Glasgow and West of Scotland Forum of Housing Associations (GWSF) to explore partnership frameworks and funding support.  In addition, the integration capability of Hubble enables alignment with national housing performance reporting frameworks and health data streams, opening opportunities to work with public health bodies on preventative health interventions linked to housing quality.

With more than 2,000 Registered Social Landlords in the UK, future revenue streams could also include white-labelled or branded versions of the platform for other social landlords or local authorities, bespoke data analysis consultancy, and partnerships with technology vendors looking to plug into the Hubble ecosystem.

This commercial model is underpinned by a strong commitment to sustainability. Once established, Hubble can operate with minimal ongoing overheads, with remote technical support, cloud-based infrastructure, and an ability to scale quickly without major hardware dependencies. The SaaS approach ensures that updates, bug fixes, and new features are automatically rolled out to clients, maintaining long-term platform value while containing costs.

The UNITY system delivers substantial financial returns by significantly reducing the need for physical maintenance visits and enabling smarter, data-led decisions.  By replacing one in-person survey (estimated at £170) and one reactive maintenance visit (£80) per property each year through remote environmental monitoring, UNITY can deliver £250 in annual savings per property equating to £1,250 over five years, well above the £899 installation cost.

Beyond this, UNITY provides further efficiencies through insight-driven repair validation. Continuous monitoring enables early detection of issues like leaks, mould risk, or heating failures, allowing repairs to be proactively scheduled before they escalate into costly emergencies. This not only prevents unnecessary reactive visits but ensures that repairs are validated, targeted, and appropriate reducing over-servicing.  Similar systems have shown that proactive interventions can cut annual maintenance costs by 20–30%, representing an additional £100–£150 in savings per property each year.

Image

Operational Efficiencies for Housing Providers
Craigdale Housing Association reported early improvements in maintenance request efficiency and prioritisation.  The AI component of Hubble enabled early detection of risk factors and proactive interventions, helping to reduce repeat complaints and reactive maintenance visits. Housing officers were also able to make better-informed decisions, reducing staff time spent manually triaging complex housing issues.

Reduction in Mould Incidents
Within the five-month pilot period, there was a 10% decrease in mould-related issues in homes where sensors were deployed.  While longer-term data is needed to confirm impact at scale, early indications show that using layered data and AI-driven alerts can significantly support mould prevention strategies.  Five early intervention visits took place based on platform-generated insights, demonstrating tangible results from the integrated approach.

Increased Digital Confidence and Inclusion
Tenants participating in the pilot described the Hubble app as easy to use, with 90% finding it 'very easy' and the remainder finding it 'easy'.  This is a significant indicator of the platform’s accessibility, particularly given that digital literacy varies across social housing populations. For some participants, this was their first engagement with a digital service related to their tenancy, suggesting the platform can be a gateway to broader digital inclusion.

Cross-sector Innovation and Partnership
Hubble’s design brought together housing, technology, and community engagement.  The partnership with Alpaca Global Solutions and Craigdale Housing Association, alongside community-led design through Tenants Hub CIC, created a replicable blueprint for other Registered Social Landlords.  The approach demonstrated how data from different systems, when securely aggregated, can drive both operational improvements and improved quality of life for residents.

Supports Net Zero and Health Equity Agendas
By enabling earlier interventions, reducing the need for site visits, and helping prevent health-related housing issues, Hubble contributes towards net zero goals and reduces strain on public services.  Healthier homes directly support national ambitions to reduce health inequalities and long-term NHS demand.  The use of connectivity infrastructure like 5G ensures environmental sustainability through remote data management and targeted responses.

Scalability and Replicability
The platform’s modular architecture, combined with a cloud-based infrastructure and flexible commercial model, ensures that Hubble can be adapted to different RSLs, property types, and levels of digital readiness.  Its interoperability with existing housing management systems makes it especially attractive to providers seeking to modernise without major overhaul costs.

Image

2. High-Quality Connectivity Underpins Success
Reliable access to the platform depended on good mobile and broadband infrastructure. Glasgow’s 5G network played a key role in ensuring real-time sensor data and tenant access to the Hubble app.  This highlights the importance of strong digital infrastructure for smart housing services, particularly when rolling out across areas with variable coverage.

3. Digital Literacy Varies, and Can Be Improved
Tenants showed strong engagement when supported through simple interfaces and training. While some users were confident from the start, others needed more encouragement.  A personalised onboarding approach, combined with clear explanations of data use and value, helped build trust.  Digital inclusion must remain a design priority for future expansion.

4. Agile, Iterative Development Builds Resilience
The project’s timeline was ambitious, requiring flexibility.  Using an agile development method enabled the team to adapt as challenges arose, particularly during data integration and staff onboarding.  Early deployment of a minimum viable product (MVP) allowed the team to gather feedback quickly and refine features in response.

5. Community Engagement is Critical
The collaborative design process with Craigdale tenants and staff was vital to platform relevance and success.  Workshops and feedback sessions not only improved usability but created a sense of ownership among users. When tenants see their input reflected in design, adoption and impact improve significantly.

6. AI Can Deliver Early Wins, Even at Small Scale
The AI insights, even during a short pilot, added measurable value. Identifying homes at risk of mould through patterns in temperature and humidity allowed Craigdale to act before issues escalated. This shows the promise of AI in housing when paired with real-world data and staff who understand how to act on it.

7. Sustainability Requires More than Tech
While Hubble’s architecture is built for scale, true sustainability will rely on partnerships, ongoing training, funding models, and continued evaluation.  Support from leadership, resident buy-in, and integration with other systems will be key for long-term impact.  Conversations are already underway with other housing associations to expand the platform’s reach.

Image