PROJECT OVERVIEW
The Future of Accessibility & Autonomous Ridesharing
Organization
CMU HCII x USDOT Inclusive Design Challenge
My Role
Product Design Lead (Visual Interface Team)

Building the Autonomous Ridesharing Experience for People with Disabilities

Product Design
Accessibility
Mobile UI
In this project, funded by the U.S. Department of Transportation (USDOT), I worked on an interdisciplinary team of CMU HCII designers and engineers to create Unigo: an autonomous ridesharing app, featuring an integrated VUI, that leverages autonomous vehicle technology to support users with disabilities.
Our team was named one of 10 semifinalists in the USDOT's Inclusive Design Challenge, which seeks design solutions to improve passenger-vehicle accessibility with autonomous vehicle technology. Working with collaborators at Propel, we are looking to fully implement Unigo for iOS and Android by December 2021.
Timeline
June 2021 - Present
I was responsible for...
UX/UI
interaction design
prototyping
Design System
User testing
Meet the team
Advisors
Nikolas Martelaro
Patrick Carrington
Visual Interface
Xander Fan
Patricia Yu
Ginny Zhao
Voice Interface
Joseph Han
Yvonne Hou
Kyra Low
Backend
John Chae
Megan Lin
Charmaine Qiu
The Challenge
Autonomous Ridesharing & Accessibility
Developments in autonomous ridesharing have yet to truly include users with disabilities. Existing ridesharing apps are currently lacking in the amenities, accomodations, and features that users with disabilities need, limiting their ability to use ridesharing services.
The Solution
Unigo: Personalized Autonomous Ridesharing
Building on existing ridesharing apps, we designed Unigo to create the autonomous ridesharing experience to supports users with physical, visual, and hearing impairments. The app allows for greater customization and control for all users, as well as tailored settings and features for accessibility.
Settings & Preferences
Customization for Accessibility & Comfort
With Unigo, users can set custom vehicle and app preferences based on their accessibility needs, and control in-ride vehicle settings to maximize their comfort. The ability to personalize vehicle type, amenities, and settings is key in accounting for the diverse range of accessibility needs that vary across different user disabilities.
Pickup & Dropoff
Personalized Guidance & Support
To support users during pickup and dropoff, Unigo incorporates features such as tailored location recommendations and alerts, as well as a guided boarding process.
in-ride features & visuals
Enhanced Situational & Location Awareness
During the ride, Unigo helps users stay informed of their location, trip status, and surroundings. Maintaining this awareness is key in helping users build trust with their autonomous vehicles throughout the ride experience.
Voice Control
Flexible Control Options
Unigo allows for flexible control of the app interface via an integrated voice assistant. Users will be able to switch to voice control at any point through the voice command, "Hey Unigo", and other customizable activation shortcuts.
RESEARCH PROCESS
The Current-state of Ridesharing for Users with Disabilities
User interviews
Understanding User Needs for Accessibility & Trust
SEMI-STRUCTURED INTERVIEWS
FOCUS GROUPS
SECONDARY RESEARCH
To gain a better understanding of salient user needs within the autonomous ridesharing experience, I conducted secondary research and interviewed users with a wide range of physical, visual, and hearing impairments. From this, I identified the three main categories of user needs within the ridesharing process.
User need #1
Customization
Users with different disabilities need ways to tailor the app to their specific needs.
User need #2
Guidance
Users need clear guidance at key touchpoints (e.g. pickup, dropoff).
User need #3
Customization
Users need to be able to maintain awareness of their trip and vehicle status, location, and environment.
Storyboarding
Capturing Key User-Vehicle Scenarios
STORYBOARDING
SPEED DATING
Building on our research, the next step was storyboarding to capture key touchpoints within the different stages of the ridesharing process. Speed dating with these storyboards then surfaced the specific user needs at each of these touchpoints, within different scenarios and interactions.
I need to know where I am, and I need to be able to situate myself. If I can’t be dropped off at my exact destination, I’d want to be at an intersection or a major building that I can easily navigate from.
— Participant, 62, Blind User
A lot of things depend on where I’m going, if there’s traffic or anything that I should watch out for. It’d be nice to know what things are like [at my destination] ahead of time.
— Participant, 35, Wheelchair User
Design Guidelines
Establishing Accessible Design Guidelines
Design Guidelines
User scenarios
Synthesizing our research in conjunction with principles from the Web Content Accessibility Guidelines (WCAG) 2.1, I worked with my team to create a set of accessible design guidelines for Unigo. This would inform our design process and ensure that the visual design, features, and functionality were all in line with user accessibility needs.
Design Process
An Accessible & Fully Autonomous Ridesharing Experience
vehicle requirements
Balancing Customization & Automation
To establish the basis for the app interface design, I needed to determine what aspects of the ride experience users wanted to customize, and to what extent. My first step here was to wireframe and test the different categories of accessibility preferences and amenities that users would want to control.
Automated settings configuration wireframes & flow
Throughout testing, users wanted more manual control over specific vehicle amenities, rather than having them determined automatically by the app. In addition, they needed to be able to understand exactly how the information they provided would change their ride experience.
Integrating this feedback across several iterations, I designed the final Ride Preferences pages to allow for more customization and explainability in these areas, while minimizing the information users would need to provide.
Ride preferences for vehicle accessibility features and amenities
Vehicle controls
Enabling Full User Control over Autonomous Vehicles
One of the main goals of this project was to ensure that users can control the vehicle entirely from the app, whether it was opening the vehicle doors, turning on the AC, or changing seating positions. In designing these vehicle controls, I focused on establishing and optimizing the overall information architecture and specific user flows in user-vehicle interactions.
Boarding & Deboarding
Some of the most complex flows within the app interface were in boarding and deboarding: my main design goals in constructing these flows were to build the optimal sequence of actions, while also allowing for flexibility in the options provided to users.
Control Panel
I designed the in-ride Control Panel to provide users with a way to customize vehicle settings. In designing the Control Panel, I focused primarily on information architecture and interaction design—the main challenges were in determining the optimal layout and grouping of controls, and ensuring that there was enough feedback and feedforward within each control.
Final design
Alternate layout sketches
Throughout testing, users wanted more manual control over specific vehicle amenities, rather than having them determined automatically by the app. In addition, they needed to be able to understand exactly how the information they provided would change their ride experience.
Integrating this feedback across several iterations, I designed the final Ride Preferences pages to allow for more customization and explainability in these areas, while minimizing the information users would need to provide.
Location & Navigation
Building Features for Location Awareness
In interviews and testing, a recurring concern users expressed was in regard to maintaining awareness of their location and surroundings during trips. To address this, I designed and tested in-ride features that leveraged mapping and GPS technology and camera feeds in ways that could be tailored to users with different disability types.
Location Updates
Helps users understand their location via recognizable venues (e.g. landmarks).
GPS Navigation
Provides both user navigation to pickup locations and in-ride vehicle navigation.
360 Live Cam
Allows users to see a real-time 360-degree view of their outside environment.
Other projects
A platform that optimizes team game booking with AI matching
An AI system that combats misinformation on social media
An anomaly interface system for GVSC robots & operators
JUBOX
VERACITY
GVSC RAI