Supporting and connecting people using Artificial Intelligence
The problem that we intend to address is whether we can support and connect people with social care needs through the use of personal AI devices. The capabilities of these devices to augment and enhance our lives is almost limitless, and their value is being proven by the dramatic uptake of their use. However, harnessing their potential in a care setting is yet to be fully realised and we see this as an opportunity to improve the lives of those that we support.
Our objectives will be to understand if we can provide better outcomes for people by utilising this technology and the connectivity that it offers. We will also want to understand if providing this offer can realise savings elsewhere (financial, resource, time) and reduce demand in to other services.
The mechanism we will use is a companion “robot” (GenieConnect®) which uses an AI interface and personal diagnostic to interact both reactively and proactively with the person using it. This is also linked to a response centre, giving another level of support and engagement for the user. As with other AI interfaces, GenieConnect® can be linked to any number of IoT devices and sensors to further support the user and enhance the offer provided through use. GenieConnect® uses commercial spec technologies, and so if the trial is successful then it could be deployed at scale at reasonable subscription cost.
Success would be measured through qualitative user engagement; success would be that users felt that by using GenieConnect®, their wellbeing had improved and that they felt it had a positive impact in their life. We would hope that they felt they had more control over managing their lives and that they felt better connected (i.e. not isolated or lonely). We would also want to understand quantitative measures i.e. how would the cost of providing a monthly subscription compare to providing other support provision such as domiciliary care, and whether it is a viable model to support people and reduce demand.
The project will be co-led by Suffolk County Council, Norfolk County Council and Service Robotics, with resourcing and infrastructure split across the stakeholders. Qualitative and quantitative performance indicators will be agreed prior to the project commencement by Suffolk and Norfolk, in discussion with Service Robotics. Resource and infrastructure arrangements will be finalised prior to project commencement dependent on agreed start date and most efficient way to manage deployment (for example, project manager could be virtual, based between localities or within one).
The total project duration will be around 6 months, with the live deployment of devices lasting up to 4 months. It is anticipated that the first 4-6 weeks will be planning and then up to 4 months live deployment and review. The project would be reviewed weekly during live deployment to ensure effective uptake and use of devices, and to respond proactively to any developing issues. We would ensure that an initial findings report was published by 31st March 2019, with a full report being finalised following the completion of the live deployment. In these reports we would assess, following our regular reviews and final audit, whether the objectives had been met and what learning needs to be taken forward.
Our key events and milestones would be:
· Project plan (including KPIs) defined and agreed: 20/12/2018
· Project team assembled: 7/1/2019
· User cohort identified: 14/1/2019
· User engagement, research and upskilling: 28/01/2019
· Deployment of devices: 4/2/2019
· Initial findings report published: 31/03/2019
· Final report published: 3/6/2019
– Quantitative will be savings against other provision (i.e. care), demand reduction to other services and the resource associated with this (including admissions to hospital etc).
– Qualitative will be improving wellbeing and connectivity, better physical and mental wellbeing, feeling in more control and more independent. Reducing loneliness and social isolation which can prevent further need developing in the future.
– Outputs will be final report on viability of using this type of solution to support those with social care needs, and opportunities to do this at scale (if any). Body of intelligence for public sector around deploying these kind of devices and technologies and the impact that they have on user outcomes and demands across the wider system.
– Delivery model and service design will be published with learnings and recommendations. As the project will be co-produced with local authorities, a technologist and with users, it will ensure that the outcomes are fit for purpose to apply to the wider system.
– Barriers or issues will be outlined to support others with similar issues (i.e. risk, consent, upskilling, sentiment, comms)
Reference points of discovery for Suffolk are its Digital Care strategy and the outcomes from a recent consultation around demand management in Adult Social Care by imPOWER.
The external consultation and review by imPOWER identified that Suffolk could provide a better “pre-front door” offer to support prevention and self-service, which could support managing demand more efficiently. Linking in with this was the opportunity to increase use of assistive/digital care to meet needs.
Suffolk developed its Digital Care strategy in response to the rise of technology and connectivity, and the outcomes from the imPOWER report. The Digital Care strategy outlines that there is potential (both known and unknown) to harness the use of technology and digital pathways to support social care needs. This is with the ambition to provide better outcomes to people and aim to reduce demand where possible to free resource in other areas. Using savings from existing technology provision to meet care needs, and applying this to future provision that could be replaced or complimented with “digital care”, it is estimated that by successful implementation of the strategy Suffolk could save £4.8 million in the next three years.
In addition, Suffolk are finalising feedback from their recent Family Carer’s Survey, and are waiting for sign off for a report on the impact of Loneliness. The output from these reports will be used to further inform what role technologies such as artificial intelligence could play in supporting people.
There are also existing programs around this area to cite, namely the work that Hampshire have completed with Argenti in using Amazon Alexa enabled devices to support people with care needs. The initial outcome report for this piece of work has found that users found having the technology and connectivity a largely positive influence in their lives, and although it was always as a compliment to other provision, it did support better outcomes and more independence.
With councils facing reduced budgets and the numbers of older people set to increase to unprecedented levels in the next decade, it is vital that we start to think creatively about how we can support people using technology and connectivity to maintain independence and wellbeing.
With the increased uptake of digital devices in everyday lives and the benefits that they can bring, our project’s findings will help to inform other local authorities about the potential benefits and pitfalls of using devices with artificial intelligence to support citizens with care needs.
Suffolk and Norfolk County Councils will be working in partnership to achieve shared outcomes and learning, which we hope will encourage others to take a collaborative approach to shared problems. We will present our learnings and experience with others to ensure system wide benefits.
If proved successful, the service is easily scalable and could be implemented by other local authorities based on our learning and outputs.
We will compile information from our existing discovery and research assets to present a business case evidencing indicative savings that could be achieved both locally and nationally to the social care system as well as improved outcomes to users. This will take in to account using a preventative approach, improving social inclusion and supporting wellbeing by using technology and connectivity.
Our project outcomes, to understand if this is a viable way of supporting needs, will be published in an initial and then final report with recommendations. We are happy to share these reports widely to support other local authorities in similar projects, and to evidence the benefits (if any) as well as outline any barriers or issues that we encountered.
Our project will inform future iterations of Service Robotic’s initial offer, as well as their overall roadmap of delivery. The learnings and intentions for next release of product would be included as part of final report recommendations and learning. Working in partnership with Service Robotics would ensure that the product created was co-produced for both users and local authorities. If proved successful, the service is easily scalable and could be implemented by other local authorities based on our learning.
This would form part of our final report and recommendations, outlining what barriers and issues there were during the project lifespan and what the resolution would be in the next phase of the project lifecycle.
For the purposes of the project, our user cohort will be anyone with eligible social care needs. However, we would start with low level needs to reduce risk and understand effectiveness at entry levels initially, and users would have to have internet access and preferably some level of understanding with technology to ensure that the research objectives can be measured more effectively.
We intend to explore the possibility of creating a feedback mechanism in GenieConnect® itself, so that users can provide feedback at any time. We would also regularly survey our users, and pull together a sample from the cohort to attend a forum to share what went well and what we could improve. There would also be a proactive element to the project, where we would “check in” with the users each week to understand if there were any immediate issues that they needed support with to try and facilitate a successful deployment. Lastly, we would explore providing the ability for users to engage with each other through the device to facilitate peer to peer support, and provide an online community where users can engage and interact virtually.
The proposed user research objectives would be to understand whether this is a viable or desirable way to support needs;
· do they like it? If not, why?
· do they use it? If not, what are they barriers?
· is it helpful? If not, how can we overcome this?
· does it add value to their lives, and if so in what areas?
· does it help them stay connected to others, and if so who?
· is it being used to keep connected with the wider world, and if so what connections are being made?
· is it helping them to live healthier or better lives i.e. has their wellbeing improved?
These objectives will support us to identify any barriers or issues that need to be understood and overcome in the next phase of the project, and whether the device is actually viable as a means of supporting individuals with care needs.
· Ability to send comms via MHCLG channels to share progress and success
· Help with sharing the outputs with the local gov sector
No previous funding