AV Accessibility

This project was a UX research-based case study focused on increasing the comfort of autonomous vehicles for people with Autism Spectrum Disorder.

Team Members: Lydia Burns, Dipti Gupte, Lakshmi Seelam

My Role: Lead Researcher

Duration: August 2022 - December 2022

Overview

The availability of autonomous vehicles is a growing industry with great promise to aid individuals who are unable to drive or access public transportation. Only one in three teens with Autism Spectrum Disorder (ASD) obtain a driver’s license at a rate of 9.2 months later than their typically developing peers. As transportation presents a major barrier to independence for individuals with ASD, autonomous vehicles afford a unique opportunity to alleviate this pain point for people with ASD. However, many individuals with ASD experience anxiety in regular vehicles, let alone autonomous ones. Because of this, we strived to design a system that would make the experience inside of autonomous vehicles more comfortable for individuals with ASD. After conducting interviews receiving feedback, we designed a conversational voice assistant and customizable notification system that can be integrated into any autonomous vehicle dashboard.

Skip to the Solution

User Research

Surveys: People with ASD & Parents/Caregivers

4 User Interviews: People with ASD & Parents/Caregivers

2 Stakeholder Interviews: Behavior Analyst and Cruise Staff Analyst

After conducting the above research methods, we took away three main topics for our research findings:

Problems with Current Transportation Methods
Current Attitudes on AVs
Preferences in Comfortability

Affinity Maps

User afffinity map based on insights collected from interviews with people who have ASD.

Stakeholder affinity map based on insights collected from interviews with parents of people with ASD.

From our research, we collected insights through affinity mapping in order to come up with the following design requirements:

User needs a way to calm down or have a distraction from anxiety induced by being in a car/AV
User wants a sense of ownership
User likes to follow a set routine
Parent/Caregiver would like a way to know user chose the correct route or made it to their destination

Design Concept: Communication Assistant Dashboard UI

We created multiple digital sketches of functions that would address our design themes. From there, we held user feedback sessions to hear the parts of our ideas that the users liked and what could be improved. Based on their answers, we updated our design concepts through the creation of a wireframe.

The wireframe (right) outlines the experience that the user will have while they are inside of the autonomous vehicle. On the left side of the screen is the map view. Here the user is able to see the progress of their ride along with the navigation details that disclose information about the vehicle’s upcoming turns and movements. On the right side of the screen is the communication center. This side always allows the user to see a transcription of the voice assistant and the conversation, trivia, or other calming activities will appear on this side of the screen. After conducting more user interviews to gain feedback on these wireframes, we designed our prototype based on the users’ preferences.

Design Solution

Our prototype solution is an autonomous vehicle dashboard interface for a communication simulator and customizable notification feature. During all of the interactions, a voice assistant, named Sandy, will speak to and guide the user through their journey within the app. Our vision is that our product would function as assistive technology in that it could be overlaid on top of any car dashboard display. Throughout the ride, the user is having a “chat” or answering questions prompted by the voice assistant; they can also monitor their route, ride silently, receive notifications, and change the settings of their notifications. All of these features work to create the most comfortable and customizable experience as possible to people with Autism riding inside of an autonomous vehicle. Our app lives on the left side of the screen and has pop-up cards with directions and ride information that overlay atop the map.

Communication Simulator

The user is greeted with a welcome message from the voice assistant.
The user can say whether or not they would like to have a chat with Sandy, so that they maintain control over the system.
Here the user is being prompted with their first chat question. If the user does not like the question, but wants to continue the chat, they can skip the question.
Sandy, the voice assistant, responds accordingly to the answer the user gave and then asks if they would like to continue the chat.
Here, Sandy is prompting the user with another question, but in this instance the user has the opportunity to freely respond to the question with their voice.
When Sandy is finished with the prompts, the microphone will appear in place of her photo, so that the user can press on it to start recording.
As with every other screen, Sandy will voice aloud the text on the screen so that the user can feel as though someone is genuinely engaged in the happenings of their day.

Customizable Notifications Feature

Here the user would press the settings button in the bottom left corner to open this popup with the customization options for notifications.
Now that the user has set their notification settings to receive updates about the weather, they are met with a popup that alerts the user to the bad weather up ahead on their route.
At the conclusion of the ride, the user receives a farewell message from Sandy to verbally and visually tell the user that the ride has come to an end.

Prototype Walkthrough

Here is a video walkthrough of our prototype, conducted by my teammate Lydia Burns.

Prototype Evaluation

In order to evaluate the effectiveness and ease of use of our prototype, we utilized the following evaluation methods:

Cognitive Walkthrough

Heuristic Evaluations

Usability Testing Sessions

Evaluation Results

84.5%

task completion rate across usability testing participants.

We asked users to rate a variety of statements on their feelings when using the voice assistant feature on a scale of 1 to 5, 1 being Not Likely at all and 5 being Very Likely. The results are displayed in the chart to the left.