U.S. DOT Inclusive Design Challenge seeks to improve passenger vehicle accessibility. As one of the 10 semifinalists, our team created the smartphone-based interface for controlling autonomous vehicles and validated its usability with an online simulator.
September 2021 ~ Current
Product designers, User researchers, Simulator developers, Mobile app developers (Propel)
Propel is an external company that partnered with our team and they developed iOS and Android versions of our mobile application.
Product Design Lead
I am responsible for managing a team of 3 members to build an online simulator. I lead the design sprints and collaborate with the development team for implementation.
Built online user testing environment for remote user testing sessions
Funded by the U.S. Department of Transportation (USDOT), the project aims to create a design system to improve the level of accessibility for autonomous vehicles. Our team has built an autonomous ridesharing app (Unigo) that leverages smartphone accessibility systems. With Unigo, people with visual, hearing and motor impairments can have full control over the autonomous car and travel independently.
As the user testing is planned to be conducted remotely, we've decided to build a web-based simulator that can display real-time input from Unigo. Also, to mimic the journey of Unigo, we've divided it into two situations - inside and outside the vehicle - and depicted them with animation and UIs respectively.
Using real-time data introduced us to the world of state machine. The main purpose of state machine is signal databases about the feedback to send, given an input and its corresponding state. I've collaborated with the developers to translate all of the user's journey into states and made sure those were resulted from either a user or simulator-generated event.
I have collaborated with the user testing team to identify the tasks to test with interviewees. From the 4 sections that the journey of Unigo consists of - ordering a ride, onboarding, during the ride, and drop off - we've specified the features that contain different usability issues.
Our goal is to make the simulator satisfy Level AA standard of Web Content Accessibility Guidelines (WCAG). To set a clear design guide of our own, I specified the key accessibility norms consulting WCAG 2.1. We took the guide into account consistently during the design process.
As the simulator intends to react to the interviewees' input on Unigo, its IA should be more concise than that of the mobile app. I constructed the IA by sorting out essential information and grouping them in a way that helps people to comprehend the simulator (e.g. what did they select and what's the result of the interaction.)
To display both context and the result of the input simultaneously, we've decided to use the sidebar UI. It not only helped us to present hierarchical content clearly but also place information in logical order so that screen reader users could easily navigate the content.
Maintaining a good color contrast was one of our fundamental rules in building an accessible simulator. We've started designing in grayscale to focus on the value of colors over hue or saturation.
One concern in regards to colors was choosing the theme of UI. To remain consistent with the mobile app, dark theme should have been applied on the controls menu. However, our final decision was light theme, as it ensured a good contrast between the design elements and the background which increased visibility.
Visual feedback should be the most striking element while controlling the inside of the car. To maximize its visibility, close-up images were selected for each control and located at the center of the screen.
One challenge we've faced was that using animation causes significant delays when developed. Thus, we've used static visual cues that implies the directionality of the controls.
For screen reader users, we've labeled the visual feedback. The terms were carefully chosen to be concise so that people can quickly distinguish what state they are at.
To test the people with motor disabilities, we created an independent flow for wheelchair users. During this stage, we collaborated with the VUI design team to enable the voice input, and the car control screens were tailored based on the premise that users will seat on their wheelchairs.
Having an accessibility mindset from the beginning, I learned how to have clear communication with users. By putting myself into diverse users' shoes, I was able to understand how different users receive the information. This made me realize that accessibility is not about designing edge cases but delivering values to the users.
