At a Glance
The Ohio State University Libraries is undertaking a major accessibility initiative to convert millions of pages of archived documents into screen-reader-friendly PDFs. Partnering with Arizona State University's Artificial Intelligence Cloud Innovation Center, powered by Amazon Web Services (AWS), the team has developed an AI-powered process that can remediate documents in minutes instead of hours. The project helps the University Libraries meet new federal accessibility requirements and advances Ohio State’s broader AI strategy and commitment to equitable access.
A Mission Rooted in Equitable Access
Cory Tressler and his team at University Libraries have a massive task ahead – converting millions of pages of documents stored throughout the libraries’ repositories and archives into easily accessible PDFs for vision-impaired users. It’s an ambitious undertaking with benefits that go well beyond the PDFs themselves.
Tressler, assistant dean for technology and digital programs, and his team are drawing from and Amazon Web Services (AWS) to meet a federally-mandated deadline that’s less than one year away from the date this is published. The new rule requires public entities, including public universities, to ensure their digital content such as documents and websites are accessible.
The mandated deadline is not the leading motivator for Tressler and his team. Access for all is the true motivation.
“One of our main mission-centered goals is expanding equitable access for as many people as we can, within our university community and beyond. We are committed to creating more pathways for more people to discover and learn. Technology makes it possible.”
Tressler explains that existing PDFs may create barriers for people to use and understand their content, especially for individuals using assistive devices like screen readers. “If there are a lot of images, graphs and other ‘noise’ in the document, a screen reader may not describe them accurately,” he says. “PDF remediation provides clear document structure and descriptions, helping provide equal access for users. We are working to create a comparable experience for all.”
The Scale of the Challenge
Producing remediated PDFs of the materials stored at University Libraries, Tressler admits, is a monumental task. “There are hundreds of thousands of pieces, from faculty journal articles, manuscripts, and research reports to photographs and cartoons. The libraries contain hundreds of thousands of PDFs that currently do not meet the accessibility requirements, and each PDF can be hundreds of pages. That’s why finding an automated method is so important. We do not have the human ability or funding to otherwise complete this project within the timeframe.”
The Ohio State University Libraries are trailblazing the effort to tackle the challenge, collaborating with some surprising partners and serving as subject matter experts for other institutions.
Innovating Through Collaboration
Tressler and his team partnered with Arizona State University’s AI Cloud Innovation Center powered by AWS last year. The teams analyzed compliance requirements and accessibility issues that exist in Ohio State’s documents. Working to leverage various AI products, the teams created a prototype of a remediation process.
An AI Breakthrough for Accessibility
The result? An innovative, AI-sourced solution that automates remediation of PDFs and makes them compliant with the new federal standards. “We brought the problem to AWS and we’ve been in a rapid sprint since then – testing it, making changes, looking at different AI models,” Tressler explains. “We’ve made improving accessibility a priority.”
Ohio State’s work is getting noticed, too. “We’re ahead of the game and are in production right now. Other universities are questioning what to do and are interested in replicating what we’re doing,” he says. “I’ve given more than 30 presentations on our work since we’ve started.”
Leveraging AWS and Adobe products, the university’s prototype can create remediated PDFs in a fraction of the time – and at a fraction of the cost – of doing so manually.
According to Tressler, the AI solution can not only reframe a document properly, but it also analyzes any images, deciphers them, generates alt text and imbeds them into the remediated PDF so a screen reader can access them.
“This innovative process takes just seconds to a couple minutes,” Tressler says. “You could feed a 100-page document into the systems, and in very little time, you’d have the remediated PDF. We recently digitized over 20,000 slides of images from the Byrd Polar Institute, images of ice cores, decades of research,” Tressler explains. “We could feed those images into this AI system, and it could generate alt text for all the slides in about an hour.”
The AI-supported method has long-range benefits for empowering learners and researchers, too. “By utilizing AI to clean up documents, we’re creating more data or meta-data of the information,” he explains. “Potentially what we are doing will lead to more data for massive scale research projects that are enabled by AI, and that is exciting.”
Tressler’s project meshes seamlessly with Ohio State’s broader AI strategic vision outlined in its Education for Citizenship 2035 plan. “What we’re doing is not just compliance work — it’s about living our land-grant mission,” Tressler says. “It aligns perfectly with the university’s strategic focus on innovation and impact."
Looking Ahead: Sharing the Solution Broadly
And it won’t stop at Ohio State. Tressler explains, “Once we refine the system, we can share it broadly with open-source code so anyone else at the university – or anyone across the country and around the world – can implement it, too.”
For now, Tressler and his team will continue to refine the prototype remediation process and continue to tackle the accessibility project at hand. They are testing the system behind the scenes and partnering with curators, staff, librarians and catalogers throughout the library.
“This is only the beginning, further demonstrating how technology can be used to serve the Buckeye,” states Tressler. “We have a massive task ahead and these regulations have accelerated our attention to it. But it’s inspiring. Our mission is expanding access to the wealth of resources housed in University Libraries while making the experience comparable for everyone. Accessibility is always the right thing to do.”
About The Ohio State University Libraries
University Libraries aspires to be one of the leading research libraries, advancing the educational, research and engagement missions of a 21st-century public land grant institution.
We support students, scholars and the broader Buckeye community through rich collections, expert services, innovative resources and welcoming spaces that advance effective research, teaching and learning. Guided by our strategic directions, we are dedicated to sharing knowledge while empowering researchers and students for lifelong success. We are committed to supporting access to high-quality, affordable education and sharing knowledge for generations to come. Learn more about University Libraries and our mission on our About Us page.