Situation

With the introduction of large language models (LLMs) like ChatGPT, Envision, a company that uses Google Smart Glasses to recognize and describe the environment to blind users, saw an opportunity to enhance their technology. Their goal was to create a voice assistant for visually impaired individuals, enabling users to interact with the assistant and ask questions about their surroundings and more.

Task

My task was to design a mobile app for the voice assistant with a focus on accessibility, accommodating every user across the spectrum of visual impairment.

Action

To tackle the project, I first conducted research by reading relevant papers related to the target group, current technology, and designing for visually impaired users. This was necessary since I wasn't familiar with the problem space and needed to empathize with the target group. Next, I organized an interview session and conducted a competitive testing analysis, where participants completed a series of tasks to evaluate how current products perform. This helped me understand the pain points and identify possible directions for improvement.

I used an affinity diagram to make sense of the numerous data cards and uncover insights. To communicate the results to the team, I prepared a presentation deck and an empathy map. During the research period, I was tasked with curating the UX of a beta version of the final product to be presented at the CSUN event in California. I selected and interacted with a group of beta testers to gain valuable feedback and ensure a customer-centric approach.

The design phase followed, in which I organized a co-creative session with my colleagues, followed by an individual creative session. To converge ideas, I used an importance-difficulty matrix to evaluate and select winning concepts based on user value and development time. This led to the verbalization of the concept and use case scenarios. The iterative design process was supported by constant feedback from senior product team members, beta testers, and a subsequent testing session.

Result

The result was a product that not only met user expectations but exceeded them, making it competitive with offerings from industry giants. The current design will also be implemented in smart glasses and is scheduled for public release in late 2024. This project taught me a great deal about accessibility, an often overlooked aspect of interface design that significantly impacts the well-being of many people.