Automatic Ability-Based Design (AutoABD): A Novel Framework for Accessible Technology Evaluated on Visual Abilities

Abstract

Humans automatically adapt to their interlocutors, making human-human interaction universally accessible. In contrast, technology’s accessibility remains overlooked, delayed, or incomplete. To improve accessibility, many design frameworks, like Ability-Based Design (ABD), have emerged. However, to mimic human adaptation, a system must first accurately assess the interlocutors’ abilities. Thus, we propose autoABD, a novel framework rendering ABD automatic via an autonomous assessment of user abilities. We test autoABD with a study including blind and sighted participants. The model’s scores, based on participants’ eye gaze behavior, differed significantly between both groups (t(14)=5.15, p<.001), allowing the correct classification of 15/16 participants. Equivalent accessibility was demonstrated (W=5.0, p=0.042). Interviews confirmed technology’s lack of accessibility, highlighting the need for such a framework. autoABD’s feasibility was established for visual abilities, providing a foundation for broader application. Extending it to other abilities will enable systems to adapt to any user, thus preventing the exclusion of users with impairments.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.