ACD End of Year Report and Roadmap

2025

2025 was officially the inaugural year of the Accessibility Compatibility Data Project. In the spring, I decided to develop an idea that had been slowly worked over in my mind about how we can make accessibility data more accessible to developers. ACD is the result of that and over the last eight months, I have been able to collaborate with great minds and aligned organisations.

We started the project with a list of milestones and KPIs, which can be summed up to:

I’m happy to report that we made significant progress in these areas.

After a series of workshops over the spring and summer I was able to understand where the gaps lay, what was needed and what was fundamentally out of scope. These workshops helped clarify the vision of the project and work out the best way to create the dataset. The Technical Gap Analysis and Scope document outlines what ACD is/is not, the motivations for the project and the scope of the project. This has been pivotal in guiding the communication about the project and deciding the needs of the project.

As I speak to people about ACD, I’m getting the same feedback: this is the right time for this project. Many people have tried in the past to do similar work to what ACD is trying to do, however the timing, societal and industry appetite for accessibility interoperability data hasn’t been there. However, we’re seeing an increase in a desire for more accessibility education, resources and data, as demonstrated in the 2025 State of HTML Survey.

With this in mind, I’ve been successful in courting multiple stakeholders, and seriously engaging relevant groups about the integration of ACD into developer tooling. Our 2025 stakeholders were:

The A11y Project, Open Web Docs and Mozilla have all provided funding to the project, and Open Web Docs, Microsoft, Tetralogical and Cynthia Shelly have all contributed expertise and resources to the project. Thank you stakeholders!

With the funding provided and the contributions to the lola’s lab open collective, I was able to work on a rough implementation of the ACD Collector. This implementation differs from what is outlined in the KPI document, as we decided against writing our own testsuite. Originally the KPI details writing our own test matrix and defining a testing methodology, however Web Platform Tests (WPT) already has most of the browser tests we’d want to write, and the infrastructure to run those tests on multiple platforms and multiple browser versions. So instead we pull test results from WPT and an accessibility dataset and combine those results to give us the accessibility support data. Relying on WPT and an open source accessibility test suite means that we can contribute back to already well maintained open source projects. The collector currently only pulls in data for HTML-AAM. This implementation is written in Ruby because it’s the language I was most comfortable with, however it will be rewritten in JavaScript this year. The idea for this implementation is just to understand how this methodology could work in practice, and it has left us with a number of open questions such as:

We hope to answer these questions this year.

2026

This year along with the questions we’d like to answer, we also have goals we’d like to achieve. We have officially been added to the W3C WebDX Community Group’s roadmap for 2026 which means we’ll be working with the group to actively identify ways we can expand Baseline to include accessibility data.

We will also be working with Open Web Docs and Mozilla to develop integration pathways for MDN.

The first version of ACD uses a static accessibility dataset kindly provided by Tetralogical, however, we want to move to a dynamic dataset such as ARIA-AT which relies on tests in a similar way to WPT. We will also be working with Cynthia Shelly to increase the test coverage of web features in WPT, and potentially create a way to extract subtest data from WPTs.

We will be converting the ACD collector to JavaScript to take advantage of web-based technologies and finalise how we compute test result data to support data

Finally, the most important thing we’ll be doing this year is applying for more funding. I work on ACD full-time (I split this work with web standards work), and the contributors to the project should also get paid for their time and expertise. Open source projects such as these, especially those that address accessibility, are often poorly funded. However, we’ve proven a need for this data and a methodology, and an appetite for this kind of work has been demonstrated. This work cannot happen for free, and nor should it, if you’d like to support this work you can do so by donating here: https://opencollective.com/lolas-lab/projects/acd or by emailing me at lola@lolaslab.co.