Mark Donovan says a digital project has helped identify those most in need of public services.
When we think about digital services, we must be ambitious and with the potential reorganisation of public services underway, it will be important to move past the basic ‘digital first’ services, where citizens can apply for licenses online, book waste collection or identify potholes that need filling. All vital services, but let’s think about what we could do if we really broke down barriers between people and services, and between more than one organisation.
Imagine if you could start again, joining up all the data available about how someone interacts with government services, and finding a way of redesigning services based on what people need, when they need it. The data would likely show that some people rely very heavily on government-funded services – perhaps they live alone and need social services support, or they are regularly in and out of hospital, or maybe they are a focus of police services because they are victims of violence, or indeed are regularly involved in anti-social behavior.
Surely that would be digital innovation for the right reasons, and a concept that we in Wales would highly value. We know that many of our communities have suffered significant social deprivation, and our instincts would be to provide support for those who need it most.
Sounds like a difficult task? It would need a fresh approach from people across the public sector. How would you get the different organisations to share the data they hold in its many forms and formats? Quite tricky to do in Wales?
Well I am pleased to say we have already tried it out. It works. And the potential is enormous.
The concept came about through innovative thinking from committed public servants within the Welsh Government, local authorities – Blaenau Gwent, Torfaen and Caerphilly, Gwent Police and the Aneurin Bevan University Health Board. The idea was to develop an intelligence system that could identify particularly vulnerable people who needed more support than most, and this led to a project, which was lucky enough to secure a grant from the Local Service Board European Social Fund programme to explore the concept.
Imagine if you could predict who is most likely to be vulnerable to falls and trips. The cost to the public purse for the support required in those aged 65+ has been shown to be in excess of £12,000 (King’s Fund) for the twelve months following an admission to hospital due to a fall. If you could divert some resources to preventative measures, based on real information and good evidenced data, through home adaptations or just help with shopping, you could better support people to lead full, independent lives and reduce costs across the public sector.
It sounds easy, but sharing data across agencies is technologically fraught, and this is where Atos’ specialist expertise in data management, analytics and digital came in. We worked with the multi-agency team to draw in all sorts of raw data from all sorts of systems, matched it to individual records from multiple sources, and derived insights from citizens engaging with local public services. There were issues around sharing protocols and governance arrangements, and of course such collaboration needs political buy-in.
“We don’t have a baseline for success with this project” Jonathan Pinkney at Regional Project Coordinator at Blaenau Gwent Council said at the outset of the project, “But if at the end of the pilot we have good quality, accurate and consistent information that we can take to a social worker, and say: how can you use this data to perform your duties better – to stop someone being abused, or help someone more quickly? That will be fantastic success.”
This six month pilot project has effectively proved that it can be done. The Vulnerability Intelligence Project shows that data sharing across complex organisation in Wales is possible, and that we now have the tools available to be able to redesign services around the people who need most help, in the best way we can.
What happens next is now being discussed and debated – do we create a new notification system, so social services can be alerted if the police are called or if someone is hospitalised, so the support required can be quickly targeted, or do we focus on prevention? It’s an exciting time because the potential is astounding, and we look forward to playing our part in truly making the most of digital innovation for all the right reasons in Wales.
I smell the huge ‘astounding’ potential for an Orwellian ‘Big Brother’ in this. Like all schemes like this the motives are very ‘worthy’ at the outset. Give it ten years of roll out and there will be cameras hooked up to the Internet in every ‘problem home’ (ie.the whole of Hirwaun) and big data banks of computers flagging up instant messages to be relayed to fleets of coppers on motorbikes and wifi connected nurse robot drones. Talk about the ‘rise of the machines!’ No doubt someone is already working on a social services Terminator somewhere that will be set into motion by a computer generated alert and which can never ever be stopped!
So in the sad world of government IT even single agency projects consistently fail to meet requirement, go massively over budget and come in years late, Projects are sold as visions to transient politicians with no clear understanding of measurable deliverables. Low grade/overworked government officials are seduced by the latest management consultancy manufactured bauble/fad which turns out to be hopelessly expensive and conveniently extends the project scope. When projects finally go live all the original team have long since been moved on. Meanwhile industry and public look on incredulously. Taxes slip away to their final resting place – the bank accounts of Partners in the global service industries.
Against that backdrop you are suggesting that a we pursue a larger multi-agency integrated technological solution on the basis of a six month trial without a baseline tainted with ‘lets just spend it before we loose it ‘ euro-funding.
Maybe stand up and take a good deep breath……….you might just detect a whiff of coffee.
George Orwell wrote about this – shame he got the year wrong! Nanny state or digital nanny state – I don’t want either! Where is the opt-out button?
The comments about this article are so terribly sad to read. Instant naysaying negativity.
The global evidence is clear: joined up and open data provides greater effectiveness and also greater efficiency.
The fact that we have seen failures in the past is no excuse to not try and try again.
Sadly I fear that the naysaying mindset will prevail and we will be stuck in a round of sub-optimal effectiveness and continued inefficiency.
@alan davies. Pointing out the potential problems in any scheme is not ‘nay saying’. As for ‘trying and trying again’, the phrase ‘if in a hole stop digging’ springs to mind. As for the evidence that ‘open data’ of the sort envisaged in the article would lead to greater efficiency and effectiveness (for whom or what?) I would be grateful for a link to the evidence. I may be wrong but I have a very suspicious ‘libertarian’ view of this.
Further to my last comment the ‘most efficient’ and most effective’ governments (at spying and collecting information on their citizens) tend to be totalitarian (or American). Give me an ineffective, sub-optimal government where we all ‘muddle along’ any day.
This very general description left me totally confused as to the actual result achieved by the pilot project. You “derived insights from citizens engaging with local public services” with a view to “redesign services around the people who need most help”. Does that actually mean training more doctors or parking ambulances at strategic points?
Is there a summary of results or some sort of formal statement of intent that we can read?
@alan davies
Computers are very good at crunching regular, complete, empirical data but notoriously bad at crunching random/complex and incomplete data which is the kind we tend to need to input from living or real-world physical systems. In these situations the output is only as good as the data and/or the algorithms/modelling used. Most people are familiar with the concept of GIGO = garbage in, garbage out but it can actually be, and often is, much worse than that! We can have incorrect modelling deliberately passed off as correct, we can have biased modelling, we can get the algorithms completely wrong, so the output does more harm than good. Computers are now being used as tools of mass deception. I don’t want this garbage to be used for social or political engineering – period!
There are plenty of current examples of computing power doing more harm than good in this world. And I can’t see any evidence that the political class has the skills necessary to specify outcomes that do more good than harm. Nor do I have a great deal of confidence in the current standards of analysis or programming based on the computer systems I have to engage with.
We employ professionals to decide who the vulnerable people are using intuition far more accurate than any artificial intelligence I have ever come across. I would like to keep it that way until I can see some evidence that computation of living systems doesn’t add up to ‘negative skill’.