LOTI are publishing two reports written by Faculty detailing a range of possible applications of Artificial Intelligence (AI) that might help with specific elements of service delivery in both adult social care (ASC) and housing services. These reports are aimed at a variety of audiences, mostly with the goal of guiding genuinely useful innovation:

1. Council officers working in these service areas;
2. Those working in housing or social care but from outside of local government;
3. Technology or data professionals in the public sector who are interested in the potential of AI;
4. Technology companies who might be interested in developing genuinely useful products for local government.

Councils are under incredible pressure to deliver some of their core services, with funding having plummeted in spite of rising demand for those services. As a result, officers are increasingly desperate to find new and innovative ways to deliver these services. Accompanied with the excitement around AI, there has been particular interest in how AI might help with designing some of these services.

However, as a sector, local government is often quite poor at identifying opportunities for innovation, especially when it comes to emerging technology. We don’t have in-house technical expertise (especially compared to those trying to sell us technology) to evaluate the feasibility of different ideas, and sometimes struggle to marry our short-term needs with actions that fit a long-term strategy of where we want to get to with technology.

Therefore, under the steer of LOTI members, we identified adult social care and housing as two priority areas where we wanted to work with our members to identify and evaluate ‘what are the actual AI opportunities’ that we might want to focus on individually or as a sector. Therefore, we commissioned Faculty to do this research, bringing their service design and technical expertise to help answer this question for us as a sector.

A collaborative, user-led approach to identifying AI opportunities

Rather than start with the technology and look for solutions, we believed it is important that we have a user-centred approach, just as councils should for any innovation project: start with the problem and the needs of the users, and then consider if AI in this case is a relevant technology to help solve these problems.

To do this, LOTI connected Faculty researchers with officers from six LOTI member councils. Faculty ran mini workshops with these officers to understand what pain points exist in their job, what technology and data they use, what the objectives are for the service as as a whole and more. The findings from this engagement informed a set of ideas for AI solutions, that were then prioritised by the council officers themselves, using a framework that Faculty developed in consultation with those boroughs (see below).

What were the suggested opportunities that we prioritised?

In each service area, the solutions that ranked highest for impact and feasibility were then fleshed out by Faculty with a service designer and a data scientist – including imagining the solution in practice, and what some of the specific risks and mitigations might be to using it well.

One thing to note from a LOTI perspective – most of the solutions that officers suggested that were higher impact were more short term ‘quick’ (although not necessarily easy) wins. This makes sense and probably is right given the short-term financial pressures that officers are under. However, as a sector, we probably should still be paying attention to more of the long term transformational opportunities of AI, things that might require more imagination and research to know if they are even possible and so which maybe our officers didn’t prioritise right now.

Nonetheless, our findings were as follows:

Housing Services Solutions Shortlist:

1. Optimising resource planning and forecasting risks and their implications.
2. Regular and automated updates and responses to maintenance requests.
3. Extract and summarise maintenance requests.
4. Preventative maintenance of properties, including boiler maintenance.
5. Mould and Damp Detection.
6. Smart sensors, IoT (Internet of Things) devices and how to utilise these for monitoring faults at scale.
7. AI matching algorithm for allocating the right properties to the right households.

ASC Services Solutions Shortlist:

1. Automating the transcription of meeting notes.
2. Converting existing documentation into learning disability formats.
3. Matching adult social care data to gather insights for decision making.
4. Triaging and prioritising requests to deal with demand across the front door.
5. Improve and encourage self service for direct payments.
6. Predictive forecasting to allow for early intervention in care.
7. Interactive tool for ASC care providers for case management and personalised care.

What’s next?

In line with some of the recommendations in the report, we would love for these bits of work to precipitate councils to start looking into whether these solutions might be feasible for them. We hope to assist any of our members who are curious about whether any of the use cases identified might be something they look at, either individually or collectively.

Please visit our Responsible AI project page for updates on our work and get in touch if you are interested in any of these use cases.