Assistive-Service Accessibility Testing for App Use Cases
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated accessibility analysis techniques fail to accurately evaluate the accessibility of software applications for users with disabilities, as they primarily focus on predefined guidelines and do not consider how assistive technologies interact with the user interface, leading to numerous irrelevant warnings and missed accessibility issues.
Innovation Solution
An automated accessibility testing environment, Latte, that uses assistive services like TalkBack and SwitchAccess to simulate user interactions, extracting use-case specifications from GUI tests and executing them to identify accessibility issues that affect main functionalities, providing a high-fidelity accessibility report.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If conventional automated accessibility analysis tools are used, then accessibility compliance checks can be performed automatically, but they fail to accurately evaluate true access at the use case level and produce many irrelevant warnings
Solution Approach 1:
The patent introduces an intermediary layer (the automated testing system with screen reader integration) that mediates between the accessibility guidelines and the actual app usage. This intermediary simulates how assistive technologies interact with the app, providing more accurate feedback than direct guideline checking.
Solution Approach 2:
The system creates a virtual copy of the user's interaction with the app through screen reader simulation. By copying the behavior of actual assistive technology usage, the system can evaluate accessibility without requiring real disabled users, maintaining automation while improving accuracy.
2Measurement precision
If manual user evaluations with disabled users are conducted, then accurate accessibility feedback can be obtained, but the process is costly and difficult to perform
Solution Approach 1:
Instead of directly engaging real disabled users, the system creates a virtual model that copies their interaction patterns through screen reader simulation. This maintains the accuracy of real user feedback while eliminating the complexity and cost of recruiting and coordinating actual users.
Solution Approach 2:
The system enables the app itself to be evaluated through automated testing without requiring external human evaluators. The testing framework performs self-evaluation by simulating assistive technology interactions, making the process independent and scalable.
3Productivity
If existing automated accessibility tools are used, then development time is reduced, but they report too many superficial issues that are irrelevant in practice
Solution Approach 1:
The system dynamically adjusts the testing approach based on the app's actual behavior and the simulated screen reader's responses. Rather than applying static checklist-based checks, the testing adapts to the specific app context, reducing false positives while maintaining comprehensive coverage.
Solution Approach 2:
The patent applies different testing strategies to different parts of the app based on their accessibility characteristics. By focusing testing efforts on critical use cases and areas with higher accessibility risk, the system reduces noise while maintaining thorough evaluation where it matters most.
Data Source
AI summary
Assistive service driven automated accessibility testing environments in accordance with embodiments of the invention are disclosed. In one embodiment, an automated environment for testing an application (“app”) using an assistive service is provided, the automated environment being embodied in code executable by a processor of a testing device, which when executed causes the testing device to generate at least one use-case specification for an app under test, execute at least part of the at least one use-case specification on the app using at least one assistive service, and generate an accessibility report for the app.