Challenge 7.6: How do we use Artificial Intelligence, data and digital technologies to enable more inclusive access to public services, starting with disabled people?

Key information for applicants

Please note: you must apply for this Challenge via Public Contracts Scotland

Launch date
Tuesday 31 May 2022

Questions may be submitted until
16:00, Tuesday 21 June 2022

Closing date
Midday, Tuesday 28 June 2022

Exploration Stage interviews
Wednesday 27 July 2022

Exploration Stage
15 August to 2 September 2022

Accelerator interviews
Wednesday 07 September 2022

Accelerator Stage
3 October 2021 to 27 January 2023

Maximum contract value
£800,000

What does this mean?

Q&A session

A live Q&A session was held with the Challenge Sponsor team on Wednesday 8 June 2022 at 15:00. A recording of the session can be viewed here:

Why does this Challenge need to be solved?

The Scottish Government and Scottish public sector provide a wide range of public services to the people of Scotland. Those services might be delivered face-to-face (for example,  a medical consultation), by video (NHS Near Me), telephone (NHS 24’s 111 service), and increasingly, online services such as NHS Inform, which can incorporate content such as information text, questionnaires, and chatbots.

Ensuring inclusive access to those services is a statutory duty, including under the Equality Act 2010, Patient Rights Act, Patient Charter and in relation to Augmentative & Alternative Communication the Provision of Communication Equipment and Support legislation – Part 4 of the Health (Tobacco, Nicotine etc. and Care) (Scotland) Act 2016. A new public sector equality duty is also currently at the public consultation stage (https://www.gov.scot/publications/public-sector-equality-duty-consultation-easy-read), and one of its proposals specifically relates to inclusive communications.

Ensuring inclusive access is also a major challenge - as well as a moving target, as the balance between modes of access evolve over time, for instance due to the COVID-19 pandemic, and might affect users differentially.

As an example, some groups of disabled people will experience barriers to accessing telephone services. The wide-ranging number of conditions that can impact a person’s ability to communicate effectively over the phone makes it difficult for telephone-based service providers to anticipate and accommodate the circumstances experienced by everyone who may wish to use their services. To improve access to telephone-based services, organisations across the UK, including the public sector, promote Relay UK, helping people with hearing and speech difficulties communicate with anyone over the phone, using the national relay service. In 2012, the Scottish Government and NHS 24 began to work in partnership to develop a process that would allow Deaf British Sign Language (BSL) users to access NHS 24’s telephone services. This work eventually led to the creation of Contact SCOTLAND BSL, an organisation that now supports Deaf BSL users to contact any service provider that relies on telephone communication. This includes public, third and private sector organisations across Scotland. Deafblind BSL users are also able to use Contact SCOTLANDBSL if they have a braille reader, or residual sight. This “once for Scotland” approach has helped to remove the barriers that previously existed for Deaf and Deafblind BSL users wishing to contact service providers by phone.

However, many other groups of disabled people still continue to experience barriers to accessing telephone-based services, including but not limited to, people who:

• have lost their voice

• have difficulty speaking

• are deaf but do not use BSL

• are neurodivergent

• have a learning disability

• are living with dementia

• lack the dexterity to select numbered options

• live with lifelong conditions, such as cerebral palsy

• live with acquired conditions that can impact their ability to communicate verbally such as, motor neurone disease, a stroke or throat cancer

This is one example of the challenges for one specific mode of service delivery (the telephone), but there will be other challenges for the other modes and services listed above. This variety as well as the intersectionality between different dimensions of disability - such as cognitive (dementia); communication (blindness, deafness, speech impairment, dexterity - but also neurodivergence, age, poverty and other socio-economic factors means a range of inclusive solutions is required.

Barriers to accessing essential services have a significant impact on lives. For example, people with learning disabilities have significantly poorer health outcomes. We expect that much of that may be underpinned by challenges in lack of ability of the system to communicate in an effective way with people with learning disabilities. Similar barriers in communication impact those who have low literacy, or whose first language is not English. By addressing this challenge, you will help make lives better in Scotland.  

Additional information about the scope of the challenge:

• The focus is on improving the delivery of real-time services, but we are aware that this might be achieved using non-real-time tools as well. For example, a questionnaire filled in by the patient ahead of time (and in their own time) to help them prepare for a medical appointment, might make the real-time (and time-constrained) medical consultation itself more effective and efficient.

• Within scope is also improving access to services for people whose first or preferred language is not English, and who are unable to communicate effectively in English. Under the Patient Rights Act and Patient Charter, everyone has the right to receive information in a format or language they can understand interactive voice response technology can be helpful to the public when navigating the menu of choices to get to the department or person to whom they wish to communicate with. This can be challenging for people if the information is not in a format or language that they can understand.

• Overall, we recognise this is a very large challenge domain, and we do not expect it to be fully addressed as part of this CivTech challenge. So we are looking for practical solutions to sub-domains of this challenge. What good looks like is described below. 

How will we know the Challenge has been solved?

More inclusive access to key public services, particularly for disabled people. This is expected to result in turn to better public services for all users.

The solutions maximise the users’ agency, so that they are able to conduct their personal business without having to get someone to do this on their behalf - living their own lives.

The solutions are user-centric, driven by the needs of users with lived experience and not by technologists’ pre-conceptions. Co-specification and co-design of the solutions through substantial, sustained user engagement will be required as part of the challenge to ensure that the solutions solve the right problems and in a way that works for the users (and service providers) and enables wide adoption and improvement. Careful consideration will also need to be given to ease of acquisition and use by the end users of any devices (particularly digital products and connectivity), whether existing or dedicated, that might be required as part of the solution. Although the Sponsor will support this engagement process, it is essential that applicants demonstrate their ability to do this successfully with a diverse user group. Engagement will need to manage expectations and provide the appropriate incentives for individuals and (often over-solicited) organisations to dedicate their time and expertise.

The solutions respect the users’ rights, including human rights and privacy (such as UK Human Rights Act 1998, UK Data Protection Act 2018, the UN Convention on the Rights of Persons with Disabilities, and the Charter of Patients Rights). This requirement also aligns with Scotland’s AI Strategy which commits the public sector to lead by example in the use of trustworthy, ethical and inclusive AI, and lists the principles we have adopted to achieve this (OECD principles for the responsible stewardship of trustworthy AI; UNICEF policy guidance on AI for children). Regarding privacy, some digital personalisation and/or identification mechanism is likely to be required for the solution to be effective (for instance, to “remember” communication preferences, update them as they change over time, and convey them quickly and reliably to the service. As examples, individuals who use high tech communication devices (electronic voices) might currently be blocked from getting through to telephone services as in some occasions they are being identified as “robots”; some AAC users can take longer (some much longer) to respond and can be timed out in some systems. In addition, gathering of statistical (anonymised) data about the volume, type and quality of interactions disabled people have when accessing services would be very valuable for improving service delivery and policymaking. But gaining the users’ trust is foundational to the success of any solution, and privacy is a key element of that trust. Privacy also needs to be considered from a human, not just system perspective. For instance, if the solutions require information access or relaying by other individuals, whether staff of the service, support staff, volunteers or friends and family. It should also consider constraints on the type of security that can be used (for instance, a user might not be able to use or remember a password). Regarding human rights, it will be essential to consider risks of discrimination, including those that might be related to biases in the data used for training AI, if applicable.

The solutions are safe. For instance, in a medical context, Natural Language Processing (if used) needs to respect medical terminology and be sufficiently nuanced and accurate, and potential errors need to be addressed to ensure a patient’s safety. This is even more important if translating from/to a language that is not English. Conversely, technology might offer new opportunities to make disabled people’s lives safer. For instance, disabled people can be a target of abuse (such as domestic abuse) because they have fewer opportunities and means to alert others. Perhaps the solution could incorporate a built-in alert mechanism to call for help and receive support – without their abusers knowing about it. Please note that this specific example is meant to encourage applicants to consider safety issues, and is not a requirement.

The solutions are human-centric. The aim is not to fully automate services, or replace human interactions, but to provide more inclusive public services by using technology to support users and staff. Careful consideration of how staff will operate any systems and human-machine interface is essential.

The solutions are modular in nature, and reusable for several types of services wherever possible, using a “once for Scotland” approach. This will help deliver a more consistent and seamless experience across a wider range of services, and reduce the financial and technical barriers on each service provider and user.

Who are the end users of the solution likely to be?

Disabled people are the key user group for this challenge (see some examples above). But as already mentioned, the public services in scope serve all in Scotland, and solving this challenge is expected to result in better public services for all users.

Has the Challenge Sponsor attempted to solve this problem before?

In 2012, the Scottish Government and NHS 24 began to work in partnership to develop a process that would allow Deaf British Sign Language (BSL) users to access NHS 24’s telephone services. This work eventually led to the creation of Contact SCOTLAND BSL, an organisation that now supports Deaf BSL users to contact any service provider that relies on telephone communication. This includes public, third and private sector organisations across Scotland. Deafblind BSL users are also able to use Contact SCOTLAND BSL if they have a braille reader, or residual sight.

To improve access to telephone-based services, organisations across the UK, including the public sector, promote Relay UK, helping people with hearing and speech difficulties communicate with anyone over the phone, using the national relay service.

NHS 24 previously offered machine translations as one way to provide information, which had been published in English to be translated into different languages. A factor in withdrawing this function was the reliability of the translations provided within a health and care setting. 

Are there any interdependencies or blockers?

No, although it is clear there is a wide range of policy areas with an interest, there is buy-in from all to ensure success and a wide range of users and stakeholders covering the relevant areas, to test solutions.

Will a solution need to integrate with any existing systems or equipment?

Yes, for instance, if a proposed solution is focused on widening accessibility to NHS 24, it will have to interface with the relevant NHS 24 systems. As this will depend on the service(s) covered by a proposed solution, this will be discussed during the Exploration Stage between Participants and Sponsors.

Is this part of an existing service?

The challenge potentially concerns a wide range of public services, with a focus on health and social care in the first instance. One of the use cases could be NHS 24, but that is not exclusive.

Any technologies or features the Challenge Sponsor wishes to explore or avoid?

In addition to AI, the sponsors are open-minded in terms of technologies to be explored, as long as they meet the success criteria listed in section 2. Interfacing with current systems on the service side and on the user side will of course be an essential consideration for any technology to be useable. We can share information about augmentative equipment that is already in use.

What is the commercial opportunity beyond a CivTech contract?

Solutions to this challenge for Scottish public services are likely to be applicable to other government and private sector providers of a wide range of services, so there is potentially a considerable, international commercial market.

Less obviously, non-disabled users in extreme environments can experience similar challenges to disabled users in more day-to-day environments. For instance, an aircraft pilot doing a high-G turn at the border of consciousness, or an air traffic controller focusing on a busy airspace experience communication and cognitive bottlenecks relative to the speed at which they need to make decisions. Therefore solutions to this challenge might potentially be transferrable to extreme environments.

Finally, by providing a demanding set of design constraints, this challenge provides a creative opportunity to improve the participant’s AI techniques themselves and develop an associated competitive advantage. For instance, Machine Learning typically focuses on performing well on an “average” case and less on edge cases, which is creating significant barriers to real-world implementation – such as the behaviour of self-driving cars in rare, unexpected circumstances. In this challenge, what are normally considered “edge cases” are central to the problem we want to solve. For example, automatic captioning now performs very well for standard speech patterns, but generally very poorly in the presence of pausing/breathing and more generally non-standard patterns such as those caused by cerebral palsy. Mainstream technologies such as the typewriter, and time-stretching for audio playback speed-up originated from the disabled community, so we hope this challenge will nurture innovation for all.

Who are the stakeholders?

• Scottish Government AI team 

• Scottish Government disability policy teams

• Scottish Government Equality Policy and Mainstreaming Team

• Scottish Government Chief Nursing Officer Directorate who lead on allied health professionals

• Other relevant Scottish Government health and social care teams

• Health and social care delivery organisations, such as NHS 24

• Representatives from end user communities – and most importantly, the end user themselves, with lived experience.

Ultimately the solutions might have applicability to sectors beyond health and social care (such as education and justice) but to ensure focus, this will be out of the scope of this CivTech challenge.

Who’s in the Challenge Sponsor team?

The Scottish Government AI team will provide expertise in AI and related technologies specifically, as well as their experience sponsoring past CivTech challenges.

The Digital Health & Care policy team will ensure that this challenge is aligned to the strategic direction of digital in relation to the delivery of Health & Care in Scotland.

The NHS 24 Participation and Equalities Manager will provide information on some of the challenges that disabled people have shared with them when accessing telephone, web-based, and face-to-face services. This sponsor will also be able to describe the steps that were taken to introduce a pilot, which went on to become Contact Scotland BSL and how the active participation of people with lived experience was critical to the success of the overall programme.  

What is the policy background to the Challenge?

Please see ‘Why does this Challenge need to be solved?’ above.