Accessibility testing for web applications is often time consuming and requires specialized knowledge. For these reasons, if accessibility issues are addressed at all, it is often only after an application or feature is already built and deployed.
The Center for Academic Innovation (CAI) develops and maintains software for use at the University of Michigan as well as other institutions. As a federally-funded institution, we are required to meet certain accessibility requirements laid out in section 508 of the Americans with Disabilities Act. As our portfolio has grown, so too have our accessibility testing needs.
This presentation describes how CAI runs many types of accessibility tests automatically using browser automation within a continuous integration pipeline. Adding accessibility testing into our continuous integration workflow allows us to catch many issues before they even reach our staging environment, enabling more efficient use of manual testing time.
While the examples discussed in this presentation will be Python applications using the Django web framework, our hope is that many of the tools and approaches covered will be useful across many languages and frameworks.