One Year on with the Innovation Fund : Presenting our First Graduating Class
Teaching an AI About Empathy -- and Learning About Ourselves

UNICEF’s Innovation Fund goes beyond just offering seed funding to its companies. We connect them to UNICEF’s community of global experts and partners – and to each other, providing pathways to accelerate their solutions to market and connecting solutions to each other to build globally scalable platforms.

On February 5-7, the UNICEF Innovation Fund held a three-day intensive workshop in New York for its newly onboarded accessibility cohort. Companies from Argentina, China, Kenya, and India received one-on-one mentorship from technical, business and industry (accessibility) experts to discuss how to improve the usability of their applications, conduct risk assessments of their business plans, and develop pathways to diversify their revenue models. Networking sessions were also built in-between sessions to support companies to discover synergies between their products and to identify how frontier technologies, such as AI could be leveraged to improve their solutions.

Questions were raised and challenges were posed during the workshop — leading companies to identify opportunities for collaboration to support the broader accessibility ecosystem. These include:

What is the potential of Artificial Intelligence and Data Science to solve big challenges in the AAC field like the lack of interoperability between symbol systems?

There are no viable AAC systems or technologies which translate how a concept is represented in one symbol set to how it may be represented in another symbol set, preventing greater communication mobility for AAC users. AI technologies have shown great potential in translating between languages as well as in generating automatic descriptions of complex photographs. In a similar way, this technology can play a valuable role in bridging the communication gaps between AAC symbol sets to support children acquire literacy and communication competencies. Through bringing together members of the UNICEF Innovation Data Science team already at work on using AI and machine learning for big data challenges such as Magic Box, and UNICEF Innovation Fund accessibility companies developing and piloting products, viable technical solutions can be developed which promote universal communication for AAC users.

AAC Cohort startups’ CIREHA and Ninaad Digital technologies brainstorming ideas on how to address the ongoing challenges within the AAC field. ©UNICEF/AriamMogos

What new opportunities can emerge for AAC users from making different symbol sets and communication boards publicly available and open source?

AAC symbol sets are either not open source or not shared/easy to locate on the web under a creative commons license. This can create a duplication of efforts around the development of symbols sets, and additional challenges for multicultural communities who need access to multiple symbols sets. Moreover, without this publicly available data, the translation between symbol sets becomes an impossible task. The UNICEF Innovation Fund accessibility cohort will work with their mentors to contribute their symbols sets to a shared digital space through the development of a REST API, where symbol sets and communication boards can be curated, remixed and shared by AAC users.

AAC Startup Dakoudai, on a one-on-one mentorship session ©UNICEF/AriamMogos

Can we contribute to and build on an Open Board Standard that creates greater interoperability between AAC applications?

Most AAC applications make it difficult for speech and language therapists, parents and other users to import or export communication boardsfrom one application to another due to incompatible board formats. This can create additional work and obstacles for AAC users to personalize their communication boards. The UNICEF Innovation Fund accessibility cohort will collaborate with other companies in the AAC field and the Open Board Format team to understand the technical challenges around working toward an open board standard for portability. This might include a feasibility study to review the technical architecture of different AAC boards, developing a minimum viable product which adheres to the Open Board Format, and testing portability between AAC systems.

UNICEF Innovation Fund portfolio member, Ekitabu talks about AI & Machine Learning integration opportunities with our tech specialist, Atul.


How are we collecting the data we need to show evidence of impact?

When using AAC applications, speech and language therapists are required to evaluate an AAC user’s progress — but that data is rarely collected through an AAC application (usually paper-based), making it difficult for an AAC company to demonstrate evidence of impact. The UNICEF Innovation Fund accessibility cohort will work with mentors to develop and integrate a web-based form of the PIADS scale (Psychosocial Impact of Assistive Devices Scales) — a validated questionnaire designed to assess the effects of an assistive device on functional independence, well-being, and quality of life — into their applications for data collection.

Call to action
While the February gathering produced a number of successful outcomes, there were challenges related to big data, maintaining the privacy of AAC user communications, developing a universal and efficient process for localizing symbol sets, and more that still needs to be addressed.

Collaborate with Us
Are you working on an AAC symbol set and have data to contribute? Tell us more and join our AAC symbol set data initiative.

Join Our Cohort
Are you working on addressing similar gaps and challenges? Are you a start-up in the accessibility field with a solution which complements our solutions and can add value to the cohort? Apply to the UNICEF Innovation Fund’s rolling call for applications.

Follow Our Journey
This post is part of a series of blog posts that document our journey building the UNICEF Innovation Fund’s Accessibility Portfolio. Read more below and follow us @unicefinnovate on Twitter.

One Year on with the Innovation Fund : Presenting our First Graduating Class
Teaching an AI About Empathy -- and Learning About Ourselves