2 of 14

Agenda

What is ARIA-AT?

Overview
Example

Status Update

What’s Next?

Expanded coverage
AT Automation
Continuously-updated reports

ARIA spec process

Integrating ARIA-AT into ARIA process

3 of 14

What is ARIA-AT?

A project to test the interoperability of screen readers and other assistive tech (ATs) using the ARIA Practices Guide design patterns as examples

Tests Evaluation criteria for AT rendering	Automation Headless driving of ATs to collect results	Infrastructure Run tests regularly on cloud servers	Reporting Share results with ATs & developers

4 of 14

The ARIA-AT homepage at aria-at.w3.org.

5 of 14

Why is AT interoperability important?

Ensures AT users can reliably and consistently browse the web and closes the gap between those who use ATs and those who don’t
Helps AT and browser vendors find, prioritize, and fix bugs
Simplifies AX testing for web developers and helps them make informed decisions about support with interop data

6 of 14

Running an ARIA-AT test plan and recording results.

7 of 14

ARIA-AT Example

Test Plan: Modal Dialog

Test: “Close a modal dialog”
Command: “Press the ESCAPE key”
Expected Behavior: The dialog should close and the AT should announce the role and name of the button that toggles the modal dialog
Results: VoiceOver for Safari, JAWS / Chrome, NVDA / Chrome all have FULL support

8 of 14

Status Update

9 of 14

ARIA-AT in 2022

Test Suite: Coverage of most impactful APG patterns

Draft test plans for 45 of 65 APG examples for NVDA, JAWS,and macOS VoiceOver

Test Running: Manual results collection and review

First 8 design patterns have been tested across VoiceOver, JAWS, and NVDA, with 2 or more independent manual testers confirming output and results

Reporting: First round of reports

Candidate reports at aria-at.w3.org/reports

Automation:

Built generic AT Automation driver & drafted AT Automation API standard

10 of 14

What’s Next?

11 of 14

ARIA-AT in 2023 and beyond…

Test Suite: Coverage of APG patterns, ARIA Roles, ARIA States & Properties, HTML
Test Running: Automated test running with manual review
Reporting: Continuously-updated report, tracks new AT versions and new browser versions
Syndication: Embedded tests results in ARIA Practices Guide and elsewhere, e.g. MDN Web Docs
Standards: ARIA spec-writing process includes ARIA-AT tests and AT Automation API is standardized and implemented
Automation: Multiple ATs can be programatically automated out-of-the-box

12 of 14

ARIA Spec Process

13 of 14