Information transfer of tactile signals on mobile devices

Hsiang-Yu Chen, Purdue University

Abstract

This thesis was motivated by the need for a systematic design of tactile signals on mobile devices. Three studies were conducted to quantify human users' ability to (1) localize multiple tactors on the forearm, (2) identify tactile signals varying in amplitude, frequency and number of pulses, and (3) identify tactile signals with the aid of audio signals that conveyed, redundantly, amplitude, frequency or both. Study 1 used two 3-by-3 tactor arrays placed on the volar and dorsal sides near the wrist. Participants were asked to localize the tactors. It was found that up to 4 tactor locations could be identified without error. Study 2 used a piezoelectric actuator embedded in a structure similar to a mobile phone. Computer-generated waveforms were used to drive the piezo to create click-like sensations on the finger. The results indicated that up to 5 click signals could be identified without error through touch. This study also employed several absolute identification paradigms to verify that information transfer for multi-attribute signals can be predicted from multiple experiments focusing on one attribute at a time with the other attributes varied randomly. Study 3 augmented the tactile signals used in Study 2 with audio signals providing redundant information. It was concluded that the audio signal aided identification performance only when it provided information that was not transmitted perfectly through the tactile signal alone, and the improvement was sometimes gained at the cost of increased processing time. These findings provide important design guidelines for mobile devices.

Degree

M.S.E.C.E.

Advisors

Tan, Purdue University.

Subject Area

Behavioral psychology|Electrical engineering|Experimental psychology

COinS