[Gallery picture]
Mock-up of base
[Gallery picture]
Mock-up of clay movement
[Gallery picture]
Illustration of base
[Gallery picture]
Picture of project Demo
Tangible User Interface Final Project
Description
The standard keyboard has barely changed since the 19th century: press a key, get the same character, every time, with no trace of the body behind it. Tactile Language reconfigures typing of a conventional keyboard for character entry, paired with a continuous, pressure-sensitive "emotion ball" that lets users shape a word's size, rhythm, and distortion as they type. Rather than inferring emotion from behavior, as most affective computing systems do, it gives users direct, embodied control over how affect enters the written record — drawing on Gibson's affordances, Norman's mappings, and McLuhan's "the medium is the message" to ask: what happens to writing when the medium can register tension or hesitation directly in its visual form?
Technical Build
Two Arduino-based controllers with force-sensitive resistors beneath a compliant rubber ball, paired with real-time hand-pose classification via MediaPipe landmarks. Detection features (curl, grip, orientation) are computed as relative rather than absolute values, so the system adapts to different hand sizes and grip styles without recalibration.
Findings
In an in-class demo, participants described the letter-selection gesture as "kneading toward the letter" and said it felt like "creating a new language" though the learning curve of a pressure-based, two-channel system also surfaced usability tradeoffs that need to be assessed in future work.
The standard keyboard has barely changed since the 19th century: press a key, get the same character, every time, with no trace of the body behind it. Tactile Language reconfigures typing of a conventional keyboard for character entry, paired with a continuous, pressure-sensitive "emotion ball" that lets users shape a word's size, rhythm, and distortion as they type. Rather than inferring emotion from behavior, as most affective computing systems do, it gives users direct, embodied control over how affect enters the written record — drawing on Gibson's affordances, Norman's mappings, and McLuhan's "the medium is the message" to ask: what happens to writing when the medium can register tension or hesitation directly in its visual form?
Technical Build
Two Arduino-based controllers with force-sensitive resistors beneath a compliant rubber ball, paired with real-time hand-pose classification via MediaPipe landmarks. Detection features (curl, grip, orientation) are computed as relative rather than absolute values, so the system adapts to different hand sizes and grip styles without recalibration.
Findings
In an in-class demo, participants described the letter-selection gesture as "kneading toward the letter" and said it felt like "creating a new language" though the learning curve of a pressure-based, two-channel system also surfaced usability tradeoffs that need to be assessed in future work.