Introduction
Data-driven technologies are an increasingly common feature in everyday life. For disabled people, the use of data-driven technologies can promise opportunities to lead more independent fulfilling lives. Digital assistive technologies (DigAT), such as screen readers, wearable devices and smartphone applications, can potentially address barriers disabled people face to work, play, rest and care. However, the DigAT ecosystem can risk being technology-driven rather than driven by the needs and interests of disabled people, the primary users of these technologies. There are several points in the lifecycle of DigAT products that prevent greater adoption and more inclusive development of DigAT.
This report explores various opportunities and challenges to research, development and adoption of DigAT through the lens of how these impact disabled users. It addresses the following questions:
- What is the state-of-the-art in DigAT scientific research? What are the potential benefits and risks?
- How can we incentivise the development of transformative, sustainable, data-driven DigAT?
- What data currently exists on disability and how useful is this data for developing DigAT?
- What are the key trends and barriers in DigAT markets and how do these map onto use cases or specific needs?
Each chapter draws on evidence gathered from roundtables, workshops, interviews and commissioned research conducted for this report to answer these questions. The findings are presented as follows:
- Chapter 1 describes what digital assistive technologies are, how they are used by disabled people and what disabled and non-disabled peoples’ public attitudes are towards digital assistive technologies.
- Chapter 2 outlines the importance of disability data and the challenges of appropriate data classification. It explores advantages and disadvantages of different data classification methods and challenges leading to disability representation gaps in existing datasets.
- Chapter 3 discusses the potential of small data approaches for more inclusive data analysis and developing new DigAT. It outlines analysis techniques for small data, such as meta-learning and few-shot learning and how these can be used to support disabled people.
- Chapter 4 considers several barriers to design, development and adoption of DigAT alongside ethical concerns. This includes challenges with inclusive design, sustainable deployment and digital exclusion.
- Chapter 5 centres on suggested imagined futures for DigAT drawn from focus groups with UK-based disabled people conducted for this report.
The report also includes five case studies on the use of DigAT in employment, gaming, tourism, music and social care. These case studies outline specific examples of DigAT and explore the opportunities and challenges for disabled users.
Language
The language used to describe and refer to disability is deeply personal and can vary over time and between countries, cultures and particular disabled communities. Even amongst those who have the same disability/disabilities, there will be variety in how they experience this and the functional challenges this may entail. There are a wide range of preferences and views amongst disabled people about how to describe their own personal and group identities and disabilities. This report generally uses the term ‘disabled people’ (often described as identity-first language) which centres the collective shared identity of disabled people rather than referring to ‘people with disabilities’ (described as person-first language). However, this report does not endorse this language as the necessary ‘right’ way to describe disability and the report may use other language, especially when drawing on other sources.
Not everyone who experiences functional challenges with everyday living will consider themselves disabled, for example, older, Deaf, or neurodivergent individuals may not identify as disabled. For the purposes of this report, those who experience functional challenges that are frequently considered disabling are still priority users of DigAT given their potential to benefit from these technologies. The use of DigAT is not exclusive to those who identify as disabled and, as this report discusses, everyone can benefit from DigAT innovations.
Background
Disability is not a static category and there is no standard definition of disability. Currently in the UK, a common way to understand disability, as defined by the Equality Act 2010, is as a significant and long-term impairment which may negatively impact one’s ability to perform personal daily activities and their participation in society. However, this can be interpreted differently depending on the model of disability adopted.
Two common models of disability are medical and social models, though these are not mutually exclusive with other models combining aspects of both. Medical models of disability treat disability as abnormal, view the avoidance or elimination of disability as desirable and seek to ‘fix’ disability ()
. In this model, the barriers disabled people face are due to their individual impairments, which need to be addressed through medical interventions. In contrast, social models of disability view the barriers disabled people face as the product of society, social context and the built environment. For example, lack of mobility for a wheelchair user may be due to the barrier of inaccessible building design. In social models, disability is normal and does not need to be fixed. Instead, it is the societal barriers that disable people that need to be addressed.
In the UK, the social model has been a powerful framework for disability rights campaigns and disability advocacy organisations ()
. It also forms the basis of the United Nation’s 2006 Convention on the Rights of Persons with Disabilities ()
. However, some disabled individuals, such as those who experience chronic pain, may not fully relate to the social model of disability. Other models of disability, such as the biopsychosocial model instead combine aspects of the social and medical models to recognise the multiple factors (eg social, medical, cultural, legal) that need to be addressed to improve quality of life for disabled people ().
Aside from their importance as frameworks to understand disability, these models have consequences for technology development. The medical model focuses on the development of specialist technologies that aim to ‘cure’ disability or help a disabled person become ‘normal’. This can be at best unhelpful or at worst actively harmful for disabled people who do not consider their disability something to be ‘fixed’ by technology. However, research suggests that the medical model is more dominant in assistive technology and AI ecosystems ().
Acknowledging the social context of disability leads to a shift away from using technology to ‘fix’ disability and instead understanding technologies as tools disabled people can use if they choose to. Rather than creating specialist technologies, on this model, developers aim for universal design and integrating more accessibility features into mainstream products (). While both specialist and mainstream technologies have an important role in the DigAT ecosystem, the examples highlighted in this report focus on tools empowering people to live fulfilled lives rather than technologies that primarily seek to ‘fix’ or ‘normalise’ disabled people. This also recognises that DigAT can be beneficial to all and the users of DigAT include non-disabled people or those who don’t identify as disabled.
