Tactile Accessibility: Does Anyone Need a Haptic Glove?
Andrii Soviak, Anatoliy Borodin, Vikas Ashok, Yevgen Borodin, Yury Puzis, I.V. Ramakrishnan · 2016 · Proceedings of the 18th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS '16) · doi:10.1145/2982142.2982175
Summary
This paper presents FeelX, a low-cost haptic glove system designed to enable blind users to tactilely explore graphical user interfaces (GUIs) by moving their hands on any flat surface such as a desk or table. The fundamental problem addressed is that screen readers reduce two-dimensional GUI layouts to one-dimensional sequential lists, losing the spatial information that makes GUIs efficient for sighted users. Existing haptic displays — electrostatic friction screens, pin-matrix devices like BrailleDis (120x60 pins, costing tens of thousands of dollars), and electrovibration surfaces — are either too expensive, too low-resolution, limited to single-finger input, or not portable. FeelX takes a different approach: it places a 2x4 pin Braille cell on each of four fingertips (using standard Metec Braille cells with piezoelectric actuators), tracks finger positions via infrared LEDs and a camera, and maps the finger positions to a virtual representation of the GUI. As the user moves their hand across the table, the pins on each finger raise and lower to represent the geometric elements of the interface beneath that finger's position. The system architecture includes a camera for finger tracking (a modified Logitech webcam with infrared filter removed), a finger tracker using OpenCV blob detection, an Interface Manager that maps finger coordinates to GUI elements, and a controller that drives the Braille cells. The working area is 80x60cm on the desk, mapped to a 640x480 pixel field resolution.
Key findings
Twenty blind participants (16 analyzed after excluding 4 due to various issues) evaluated the prototype by identifying five geometric shapes — horizontal lines, vertical lines, rectangles, triangles, and circles — under four conditions (1 finger vs. 4 fingers, crossed with finger-aligned vs. field-aligned feedback). With 4 fingers, recognition accuracy exceeded 80% for every shape regardless of alignment type, validating the prototype's basic feasibility (H1 confirmed). Four-finger use was significantly faster than single-finger use (p < 0.0001), with average task completion times of 0.54 seconds vs. 1.06 seconds for finger-aligned conditions. The accuracy difference between 1 and 4 fingers was also statistically significant (p = 0.026). Participants preferred 4-finger use (50% vs. 37.5%), confirming that multi-finger tactile exploration is desirable and differentiates the glove from single-finger haptic devices. The overall SUS score was 62.2 (below the 75 threshold for "good" usability), primarily due to ergonomic concerns with the early prototype — fit issues for different hand sizes, insufficient finger flexibility, and general bulk. The most requested future feature was the ability to click on the tactile screen (rated 4.31/5), followed by the ability to spread fingers to cover more area and zoom in/out the field. Participants expressed enthusiasm about potential applications including maps, charts, graphs, spreadsheets, and maze games. One participant noted: "Receiving the information from a hard copy diagram is too slow. Haptic glove is a faster and more efficient solution."
Relevance
FeelX addresses a fundamental limitation in blind computer access: the loss of spatial layout information when GUIs are linearized by screen readers. For accessibility practitioners, this highlights an often-overlooked dimension of screen reader accessibility — even perfectly coded, WCAG-compliant interfaces lose significant usability when spatial relationships between elements cannot be perceived. The glove approach is notable for its cost advantage over alternatives (approximately for the prototype vs. tens of thousands for pin-matrix displays), its portability, and its use of any flat surface as the interaction area (avoiding the "fat finger" problem on small touchscreens). The finding that blind users strongly prefer multi-finger exploration — consistent with how people naturally feel objects — challenges assumptions embedded in single-finger haptic devices. While the prototype is far from a consumer product, the study establishes both technical feasibility and user desire for tactile GUI interaction, providing a foundation for future development. The work also demonstrates that affordable, consumer-grade components (modified webcams, standard Braille cells) can be combined to create novel accessibility solutions.
Tags: blindness · haptic technology · tactile accessibility · screen reader · GUI accessibility · braille · assistive technology · wearable technology