Existing Design Methodologies

Our approach initiates the design process with the goals, interests and needs of the individual (Cole & Nolan, 2019) who will use a device, resonating with Ladner's notion of "design for user empowerment" (Ladner, 2015, p. 24). The goal is to reframe disabled individuals—users of disability technologies—as "valuable knowers and experts" and to "foreground the political, cultural, and social value of disability embodiments" (Hamraie, 2016, p. 20). We aim to engage disabled people as well as other stakeholders such as educators, parents or caregivers, at every stage of the design process, starting from understanding their needs in their real-world context. In this sense we look to feminist technoscience scholars (see Mauldin, 2017) to not just challenge the use of technoscientific artifacts, but to challenge the notions of who conceptualizes, designs and creates technoscientific artifacts; in this case Indigenous children, their families and caregivers who have little or no access to the supports provided by Western medical models. Our approach, therefore, has the goal of designing with and by disabled people: as researchers we play the role of facilitators, supporting individuals to define requirements, design, develop and test prototypes of their own solutions according to their changing needs. When working with disabled communities, classical user-centered design (Norman & Draper, 1986) and participatory design (Schuler & Namioka, 1993) approaches fall short where promoting self-advocacy and building capacity is of equal or greater importance than informing product development. Ladner (2015) pointed out that classical design approaches need to be rethought in order to develop greater awareness of user experiences, contexts and meaningful outcomes for disabled users. While both user-centered design and participatory design are human-centered approaches—in the sense that they follow a life-cycle of analysis, design, development and testing that aims at gaining a growing understanding of who will be using a product— they need to be extended when they target the disabled community since "typical designers and developers are not disabled" (Ladner, 2015, p. 27). User-centered design, for instance, involves the users only for usability testing and does not offer tailored solutions for specific individuals (Norman, 2005). Participatory design focuses on designing and testing with users. Users are considered as design partners and their feedback and insights are "used" to design and develop a product, being the original design team in control of the design process at all the phases. With the disabled community, we aim at engaging people to be "active producers and innovators" (Heeks 2008, p. 33) and learn from the process so that they could apply the new skills to solve emerging issues. Enabling individuals to express their own needs in increasing levels of complexity and confidence is foundational for nurturing autonomy and self-advocacy (Nolan, Raynes-Goldie & McBride, 2011); seeing every individual as both unique and uniquely positioned to challenge universal design's top down and post-disability rhetorics (Hamraie, 2017; Thumlert & Nolan, 2019). We aim to extend and enhance user-centered and participatory design to provide people with the means and expertise to deal with local conditions and individual needs. Closer to our vision, experiences from participatory design field research for designing prosthetic legs in Cambodia (Hussain et al., 2012) led the authors to propose an alternative framework for participatory design with marginalized communities that focuses on building a nuanced understanding of contextual, social and cultural factors as well as instructing people on design methods and prototyping.

Our project embraces the Do-It-Yourself philosophy with the goal of enabling people to design, build and share their personalized solutions (Hunsinger & Schrock, 2016; Kuznetsov & Paulos, 2010) and therefore support self-determination (Ladner, 2015), and increase the technical expertise of disabled people through the design and development of the technology that is intended for them. Do-It-Yourself refers to any act of creation or modification that does not engage a professional (Kuznetsov & Paulos, 2010) and fosters the creation of communities (physical or online) of "expert amateurs" that collaboratively work towards building, reconfiguring or modifying artifacts (a piece of software, physical objects, interactive devices, etc.). Flexible Do-It-Yourself practices have the potential to create individualized and easier to procure assistive technologies. Mainstream technology often marginalizes people with complex needs who require ad-hoc customizations to access mass-produced assistive devices (Lundälv et al., 2014); multi-touch input, for instance, requires fine motor skills and is nonfunctional for individuals with motor impairments (McNaughton & Light, 2013). The main challenge of implementing Do-It-Yourself for disability design relates to a lack of confidence in participant's own practical and technical skills (Hurst & Kane, 2013). This is mitigated through online (Buehler et al., 2015) and local community makerspaces (Hook et al., 2014), providing access to infrastructure and guidance that increase expertise. Access to tools and assistance allows non-professional designers to both fabricate difficult-to-find parts for broken devices (Hurst & Kane, 2013) and increase the practicality and robustness of assistive technologies (Hook et al., 2014).

Disability Self-Advocacy Literature

The focus of Nolan's applied work over the past decade has been in the construction of custom adaptations for children with disabilities (CBC Arts, 2016; Henderson, 2011), starting with a period of collaboration with the Adaptive Design Association in New York (Jones, 2014) and culminating in supporting the establishment of the NGO Diseñando para el Futuro in Bolivia, which is controlled and run by Bolivians who work with organizations supporting Indigenous disabled children in Cochabamba. As an autistic adult and a member of the autistic self-advocacy movement and proponent of the social model of disability, his work has focused on an approach to design that is initiated with the goals, interests and needs of the disabled individual, extending to modify the external environment to meet those goals, interests and needs (Nolan & McBride, 2015). Disability advocates and many disabled adults subscribe not only to the social model of disability, but to 'identity first language' versus 'person first language' (Gernsbacher, 2017). A search of the Internet for the terms 'identity first language' and 'person first language' shows a wide debate on the issue, and this debate continues on into academic journals (Dunn & Andrews, 2015; Peers, Spencer-Cavaliere & Eales, 2014). It appears that disabled adults tend to be more on the side of 'identity first' language (Cole & Nolan, 2019; Gernsbacher, 2017). For some individuals disabilities are defining points of identity, and we often hear disabled adults referring to themselves as autistic person, deaf person, blind person, rather than person with autism, person with deafness, person with blindness. This identity first, social model of disability perspective positions the individual as being disabled by social norms and social infrastructure that disables them more than physical or cognitive impairments. As well as seeing the individual as having been disabled by societal structures that are not inclusive and equitable, it is equally important to focus on the disability as a defining barrier to inclusion for an individual.

Disabled people struggle daily to modify the world around them to meet their needs; they must hack to live (Cole & Nolan, 2019). This has led to the central tenet of critical disability studies and disability advocacy: "nothing about us without us" (Werner, 1998). Self-advocacy is a rights movement for people disabled by societal institutions and built environments that are not designed or structured to meet the needs of the disabled to the level that they meet the needs of others. It is a complex topic with a variety of perspectives that vary across fields of inquiry, cultural contexts, and the needs and aspirations of individuals and groups (Goodley, 2010). From childhood and throughout their lives, disabled people's lives are organized around the needs of others, and lack opportunities to develop their own sense of autonomy and assert that autonomy in a meaningful manner. Pierce and Allen (1975) argue that the lack of autonomy and self-direction is the fundamental form of discrimination and oppression that starts in childhood. They termed this form of discrimination 'childism' and asserted that this organization of a child's life to meet the needs of others is the foundation of all forms of oppression and discrimination. Disabled children and adults are often infantalized by medical-model social institutions that do not support or actively restrict the development of autonomy and self-advocacy. All children require opportunities to explore and express individuality and autonomy (Horn & Kang, 2012; Nolan, Raynes-Goldie & McBride, 2011), particularly when they have multiple disabilities and are unable to express themselves verbally or move independently. The challenge of designing with/for children is to use integrated teaming where professionals and family members work in concert to ascertain the child's interests and needs in order to enact self-determination on the child's behalf and, to the greatest extent possible, enable children to do things for themselves (Horn & Kang, 2012).

Analysis and Design

We combine observation, focus groups and interviews, with the deployment of technology probes (Hutchinson et al., 2003): "tool[s] to think with" (Papert, 1981) aiming to better understand children's interest, goals and needs from their own perspective. Technology probes are intended to work in real-world settings and have simple functionality that allow children, their families, caregivers and educators to generate design ideas by interacting with the technology. The technology introduces new design possibilities and capture children's perspective in their daily context. They are not demonstration nor functional prototype, instead they are open-ended seed technologies that facilitate meaningful conversation between the team and the stakeholders and allow people to envision emerging designs.

Researchers have "much to learn about children and children's experiences from the children" (Dockett & Perry 2007, p. 48) and we advocate initiating design with children's input as early as possible. Typically, researchers investigating issues related to disabled children rely on adult interpretations of children's experiences. As a result, children's perspectives on matters that affect their lives are excluded. Children have a right to participation and to express their views about issues that affect their lives. Article 12 of the Convention on the Rights of the Child (CRC) recognizes "the potential of children to enrich decision-making processes, to share perspectives and to participate as citizens and actors of change" (United Nations Treaty Collections, 2016). The CRC is central to our engaging children as active participants and to observe them in their natural community-based surroundings, where they feel most comfortable, rather than in a lab setting (Caplan, Loomis & Di Santo, 2016).

Prototyping and Testing

We see prototyping as an interactive process where users are supported in building artifacts and determining design efficacy. To this end, we have established a fabrication lab in Cochabamba: a makerspace where people can come together and can engage in fabrication workshops (Hook et al., 2014). Our lab is a step in developing a social network that would connect disabled children, their parents and caregivers to build accessible solutions using local expertise, resources and materials, through which we can develop a nuanced understanding of Do-It-Yourself approaches and the design of assistive technologies in underserviced areas.

In our process, we generate the prototypes with the users that are the results of their interactions with the technology probes. At the same time, we seek to build capacity by teaching them how to build or modify the technology by themselves. We start with the children who are supported in articulating their needs/goals, caregivers who have an intimate knowledge of the child, educators, administrators and therapists who can advocate on the child's behalf, and our lab staff who are able to manage and engage people in design/fabrication process. Our staff are responsible for standing at the periphery; we as researchers are ready to help build capacity wherever needed, observe the process, and refine our methods. In fact, we are prototyping the design of design teams, developing structures for ensuring that everyone can do the most for themselves within a context of a design team with shared goals focused on the needs of the child. We are fostering a radically individualized design strategy, in opposition to a Universal Design (Balaram, 2001) that is not tailored to the individual and cannot address complex and changing needs.

As a last step in the design cycle, we evaluate the prototypes in situ with the stakeholders to ensure that our designs addressed the original needs reveled by the probes and to observe how the introduction of a new technology changes the way children interact with their environment.

Participation and Consent

To access our sites, we obtained consent from Cochabamba region social services, and written consent from the directors, staff and volunteers working at the orphanages. We next met with staff and volunteers to explain the project and address potential questions/concerns. We developed a mechanism for consent that respected each person's literacy skills, asking if they preferred oral or written consent. All participants chose to provide oral consent. To ensure voluntary participation, people were informed that they could withdraw from the study at any time without consequences. If participants decided to withdraw, they could also choose to not have their data included in the study.

Interviews and data collection were organized around the availability of children, parents, educators, and therapists. In-person interviews provided data related to children and stakeholders' needs (see Participant Data). One focus group was also conducted at one of the orphanages in preparation for a workshop. After the workshop, we interviewed participants to gather more detailed information that arose from the focus group. Photographs were used as a data gathering tool for the researchers to document the learning process of individuals and groups to design and build custom objects as well as a way for participants to document their own progression through the project.

Including children in the project required parent/legal guardian consent and children also provided assent prior to participating in any research related activity. For the two children who participated in the interviews, consent was provided by the director of the school who had legal guardianship of the children. An interpreter who was familiar with the program worked with one of the researchers to inform the children about the activity in child appropriate language. The children were informed that they could stop the activity at any time. We also observed children closely for signs of not wanting to participate (e.g., non-verbal body language), and unless we could confirm a child's interest, we defaulted to the assumption that the child did not want to continue participation.

Participant Data

At present, 98% of the participants and families in Bolivia identify as Indigenous, the other 2% identify as mixed. At the various centers we are working in, 50% of the staff speak one of Bolivia's Indigenous languages. Over all, we are directly engaged with over 230 children, as well as 30 staff members who are indirectly impacted. Our workshops have brought 14 center staff and two parents into our space to learn techniques, and our lab director has identified 67 other individuals we have been in contact with. As well, we have been asked by the director of Cochabamba's Servicio Departamental de Gestión Social (SEDEGES) who is responsible for disability programs, to extend our work across the city of Cochabamba. This is presently beyond our capacity, however, it reflects support for our initiative at the local government level.

Figure 3. Deployment of the AAC technology probe.

The two images in figure three show our alternative and augmented communication device being field tested in Bolivia. The first image has a makeymakey touch interface hooked up to 4 copper squares. The second image shows a non-verbal Bolivian child using a different version of the input device hooked up to 4 round coins.

The AAC Toolbox

The AAC probe was developed as a toolbox for exploring different input modalities for allowing nonverbal children to input text and produce meaningful utterances (Figure 3). The main features include using H4 Huffman Codes (MacKenzie, Soukoreff & Helga, 2011), onscreen and physical text input mechanisms, open-source hardware, and multi-lingual functionality that embraces communication in local dialects and accents. Four input points (switches, buttons, soft circuits or toggles) can be placed anywhere a child can trigger them, enabling text entry. We used the Raspberry Pi 3 low-cost, open-source, single-board computer to which we attached (in a custom 3D printed enclosure) a multi-touch display, four physical buttons, a MakeyMakey input device (Collective & Shaw, 2012), a microphone, and speakers. Children can directly touch the on-screen keys or physical buttons to enter characters and compose sentences. Alligator clips from the MakeyMakey connect objects such as coins or play-doh compound, turning them into input keys; this provides open-ended affordances for text entry, thus offering insights on desirable physical input. Special commands are provided to delete last character/whole sentence, move backward and forward within sentences already introduced and "play" a sentence. The system provides different audio feedback for successful key pressed and character typed, to help children with visual impairment. The device records words using the voice of a parent, educator or caregiver, with the intention to ensure that our device speaks words in the languages, tones, dialects, slangs, idiomatic phrases that the child hears around her.

In developing the technology probe, we chose a low-tech approach to speech generation by recording human voices. In terms of the multilingual functionality, we abandoned the idea of computer speech synthesis, due to the cost and inflexibility. Speech synthesis such as SpeakJet (https://sparkfun.com/products/9578) or textspeak (http://textspeak.com) are still rudimentary and limited for our purposes, and the kind of systems available in, for example, Apple and Android devices, are tied with technology that exceeds our price point and design goals. The rationale for eschewing speech synthesis, in favor of human generated speech is manifold, and largely situated in perspectives from critical disability studies that highlight the need for autonomy and self-advocacy (Nolan, Raynes-Goldie & McBride, 2011). Available commercial tools for speech synthesis offer predefined set of mostly English voices, or at best, standardized and depersonalized voices that have little connection to the kinds of speech sounds people hear in their daily lives. Our technology probe, instead, provides a functionality to record words using the voice of a parent, friends, siblings, teachers or caregivers at the orphanage. This way the system would allow the child to speak with the voice of someone familiar to her. New words can be pre-recorded (creating a custom vocabulary) as well as added any time the system recognizes that a word is not in the vocabulary. The intention is to ensure that our device speaks words in the languages, tones, dialects, slangs, idiomatic phrases that the child hears around her.

We chose orthographic as opposed to other types of symbol sets (e.g., icons or pictograms) for a few reasons. Based on interviews with parents and educators during the first visit, we learned about the kinds of educational opportunities and interventions undertaken in the schools. We observed that these particular nonverbal children were gaining a facility with written language, in this case Spanish, and educators indicated that they wanted tools that support their direction of language development. At the same time, we were curious about how we could build on our prior research experience in providing children with tools to communicate outside of immediate family and caregivers, to express their own needs in an open-ended manner (Gaston, 2011). We hypothesize that being able to provide children with technologies that would help them communicate real words that are already familiar to them would play an important role in helping them establish a sense of self and identity as a social individual, and it may encourage greater self-efficacy than using a preset suite of close-ended images.

Figure 4. Visual description of the problem and design solution.

Figure 4 contains two photographs. The first shows the leg of a disabled Bolivian child on a horse in riding stirrups with the hand of an adult indicating an aspect of the design they want made to modify the stirrup to help the child better sit on a horse. The second image is a 3D paper mockup of what the modified stirrup would look like.

Figure 5. Prototypes of interactive tangible toys to support blind children in the development of pre-braille skills.

Figure 5 contains 2 photos of prototypes for a tool to teach pre-braille to young children. The first image is a tablet device and a laser cut wooden box with six buttons. The second image is the same except that the buttons have been replace by wooden doweling with magnets on the bottom of them and the box now has holes with metal rings in them that connect with the doweling.

Support children's autonomous expression

By offering assistive technologies and customized designs based on the children's individual needs, we help stakeholders support children's right to freedom of expression. We suggest that the design of assistive technologies should be left open and should be localized by the user (children, parents, educators, etc.) to achieve high repurposability. This represents a significant design challenge with respect to the trade-offs between tailorability/configurability and the demands of non-technical users. Designing for open-ended affordances and custom adaptations rely on user creativity and input; however, when the end user of an object is a child, researchers often turn to key stakeholders in the child's life for input on how children engage with their world. Stakeholders' observations of children are key to understanding children's social environment and how they construct meaning of objects and experiences in their environments. However, it is paramount to directly engage children in their natural settings as they interact with technology. Gathering data from "realistic settings and situations that reflect children's actual performance" (NAEYC, 2003) will document their activities in the context of their daily routine. Our endeavor to build capacity for local fabrication showed potential to support assistive technologies design that reflects children's need for self-determination and independent/autonomous action (Horn & Kang, 2012). As people learn the fabrication techniques, the objects and processes themselves become probes, enabling a shared understanding of design thinking that in turn leads to a greater ability on their part to advocate on behalf of particular needs/objects, and most importantly to see themselves as sources of ideas, insights and growing design expertise.

Help children help others

During our second visit we discovered an iron workshop in a school, where students with mild cognitive disabilities learn arc welding and working scrap iron. They were repairing iron gates and building flower stands from discarded iron, which they sell locally. We had the chance to include several of these children in our study and obtained valuable feedback on the design and use of our devices. We see an opportunity for the Do-It-Yourself philosophy to be embraced for the design of our devices–and AT in general–in underserved communities in a way to include young individuals with mild cognitive disabilities in the fabrication process so that they build or assemble the physical parts of the device themselves. This shifts the focus from "design for disabled" to supporting them making things for each other. We plan to explore the possibility of these children being given the opportunity to learn soldering skills so that they may assemble and maintain assistive devices. This would include the attaching of switches at various locations on a user's wheelchair or bed frame so that the assistive devices could be used in accordance with a child's limited physical movements. As well, using their metalworking skills, the children could fabricate components to repair damaged assistive devices locally as well as custom mounts to attach communication devices to wheelchairs.