INTELLIGENT AGENT BASED SOFTWARE ARCHITECTURE OF A WEARABLE ELECTROTACTILE FEEDBACK SYSTEM FOR BALANCE IMPROVEMENT

Jan Walter Schroeder, Venketesh N. Dubey, Tamas Hickish, Jonathan Cole

Abstract: An electrotactile feedback system (EFS) is a specific form of biofeedback using electrocutaneous stimulation and can help people with problems of pressure estimation caused by a sensation loss in their feet to maintain their balance and reduce the risk of falls. In this research project embedded software has been designed and tested on a wearable EFS that tackles the problem of improving balance control for impaired individuals. The program has several inventive features. An intelligent agent was implemented to process information from the sensor system and taking an action in the form of electrocutaneous stimulation. In a training routine the force sensor system is calibrated and COP information obtained. Furthermore the software detects individual parameters of the wearer by testing the sensation threshold for electrocutaneous stimulation. The system uses artificial intelligence (AI), in form of a simple reflex agent (SRA) to prevent harmful or unwanted feedback. A preliminary test was performed to test the repeatability of the sensory threshold detection and showed a high repeatability. The developed software has several novel components and can be integrated in innovative wearable EFS that can be used by people with sensory impairment to improve their ability of controlling their balance and thus preventing falls.

Keywords: Biomedical Engineering, Electrotactile Feedback, Intelligent Agents

DOI: https://doi.org/10.15623/ijret.2014.0329006