Fingertip Skin Bioimpedance Identification and Tactile Correlation: Experimental Studies towards a Smart Electro-Braille System
StatisticsView Usage Statistics
With the advances of the last twenty years, information has become readily available to most Americans. Databases of scholarly papers and news articles are now available at the touch of a button. For blind and visually impaired (BVI) individuals, however, reading is still a cumbersome task that requires either huge Braille books or, still fairly large, mechanical Braille systems that often times must be linked to a computer. Electrotactile based Braille systems that utilize electronic stimulation to the reading finger can be made much more portable, but can also prick or shock the user if skin conditions like moisture or contact area change rapidly. To solve this problem, a load-aware adaptive control loop that can adjust current stimulation level based on fingertip condition is needed. As a first step, we detail the preliminary implementation, calibration, and testing of a developed electronic Braille (e-Braille) system that can perform on-line identification of the bioimpedance parameters of a user's fingertip skin. The identified individual bioimpedance was used as a control reference for a load-aware adaptive control system in the next stage. The output of the constant voltage driver for the system was calibrated to a certain voltage output based on user input and data acquisition device (DAQ) output. Next the identification algorithm was developed and verified by observing that plots of the mean square error approached zero over time. Finally, the system was tested using a first order linear model of fingertip skin and a number of adjustments were made to reduce the identification error to an acceptable range. Additionally preliminary data were collected with an EEG headset to determine the tactile correlation between mechanical and electronic Braille. This work shows that with some adjustments, the identification system is capable of performing with errors as low as 10% over a comprehensive range of fingertip impedance values and errors as low as 5% when tuned to a specific set of bioimpedance parameters. The preliminary EEG data collected also, provide a point of comparison with which to test future studies on the brain's response to e-Braille. Thus, this portable, adaptive e-Braille system has the potential to expand the access of BVI individuals to books, news, and social technology.