TactBack: VibroTactile braille output using smartphone and smartwatch for visually impaired
Aritra Dhar, Aditya Nittala, Kuldeep Yadav · 2016 · Proceedings of the 13th International Web for All Conference (W4A) · doi:10.1145/2899475.2899514
Summary
This extended abstract presents TactBack, a system that uses vibration patterns on off-the-shelf smartphones and smartwatches to convey braille characters haptically. The system was motivated by the limitations of audio-based screen reader feedback (TalkBack) for visually impaired users—audio feedback fails in noisy environments, compromises privacy for sensitive information like passwords and PINs, and using headphones blocks ambient sounds that blind people rely on for navigation and safety. TactBack encodes each braille character as a 6-bit binary pattern, splitting the representation across two devices: the first three bits of the braille cell vibrate on the smartphone while the remaining three bits vibrate on a paired smartwatch. Different vibration intensities distinguish embossed dots (140ms) from flat dots (40ms). The system was developed as an Android prototype using Google's Wear API and supports training mode (learning braille characters and their vibration equivalents) and testing mode (simulating real-world scenarios like password entry and phone dialing). The research was informed by a qualitative study at a blind school in Bangalore, India, where semi-structured interviews with 20 participants revealed that most students had no prior smartphone experience, frequent typing errors with TalkBack, and difficulty decoding audio feedback in noisy classrooms.
Key findings
The system was evaluated with 15 participants (10 visually impaired, 5 sighted) using two vibrational interval settings: slow (700ms) and fast (550ms). All blind participants were proficient braille users; no sighted participants had prior braille experience. Results showed that participants could recognize individual braille characters with minimal training. Average recognition times were 4.5 seconds per character at the slow setting and 3 seconds at the fast setting—even users with no prior braille knowledge could recognize characters under 4 seconds. Accuracy varied by character and speed: reducing the vibrational interval from 700ms to 550ms decreased accuracy for some characters, revealing a clear speed-accuracy trade-off. Interestingly, sighted participants sometimes achieved higher recognition accuracy than blind participants, which the authors attribute to sighted users' greater familiarity with smartphones and haptic feedback rather than braille knowledge differences. Prior approaches to vibrotactile braille on mobile devices required users to explore each dot position on screen, taking 4-27 seconds per character—TactBack's split-device approach significantly improves on this speed.
Relevance
TactBack addresses a real gap in mobile accessibility: the over-reliance on audio output for blind and deafblind users. The system is particularly relevant for deafblind individuals who cannot use audio-based assistive technology at all, offering a potential communication channel using consumer hardware. The use of standard off-the-shelf devices rather than specialized equipment is a practical advantage for adoption. For accessibility practitioners, the work highlights the importance of multimodal feedback—designing systems that do not depend solely on one sensory channel. However, as an extended abstract with a preliminary study of 15 participants, the results are exploratory. The 3-4.5 second per character recognition speed, while faster than prior approaches, is still too slow for practical text reading. The system would benefit from longitudinal studies to assess whether recognition speed and accuracy improve with extended practice, and from evaluation in the real-world scenarios it targets, such as password entry and navigation.
Tags: braille · haptic technology · wearable technology · visual impairment · deafblindness · mobile accessibility · vibrotactile feedback · assistive technology