UC Berkeley Master of Design - Making Mental Health Support Accessible For Blind Children

As part of a UC Berkeley technology class, I created an interactive, calming, and tactile bedside toy that automatically provides mental health strategies based on the emotions that a visually impaired/blind child is feeling through human voice, dog barks, and vibrations.
Role
Product Designer/Engineer
Timeline
3 weeks (October '21')
Core Responsibilities
Programming, digital fabrication

Problem

Blind/visually impaired children (below age 10) aren't aware of how to navigate emotions.

When children are distressed, they often turn to stuffed animals that comfort them. To support their accessibility needs, I worked on making it easily detectable to interact with the friendly toy, which responds to the child's emotions.

User Input

Designing and programming a button that triggers speech input and uses NLP

I fabricated a button with braille so children could easily access the soft toy and feel soothed.1

. Blind/visually-impaired child presses button
2. Child can state emotion and script detects the keywords that trigger a vibration and unique audio output, the mental health strategy for cope with said emotion
Our team met with social workers and non-profit leadership to create journey maps of the fostering process. This user research helped us understand the different types of parents that sign up and the complicated journey that occurs for resource families. Altogether, this research helped us streamline the signup process for parents and the application review process for Angels Foster Care.

Output

Designing conditional statements to provide tailored mental health strategies

Example mental health strategy: Input = Button, speech (“happy) Output = Bark sound (happy), Mental health strategy (happy) “I’m so happy that you’re feeling happy today. You should be proud of yourself. Give yourself a pat on the back.”

Solution

An interactive mental health toy that's engaging and helpful to blind children

We were given some starter code, and I modified/revised/added code of my own to produce the following demo.

Conclusion

Exploring and programming AI for the first time

This was uncharted territory for me, but I learned to tackle technical challenges in ways I never did before. I got to learn the mechanics of how AI works through NLP and build my own interactive product outside the usual digital realm, what I've been so used to for the past years.