Voice Controlled Navigational Aid With RFID-based Indoor Positioning System for the Visually Impaired

This paper presents the development of a voice-controlled navigational system that could assist the visually impaired people in travelling indoors independently. The system provides audible directions to the users in navigating from a specific location going to the desired destination. This system uses passive radio frequency identification (RFID) tags strategically located along the hallways of the facilities. These tags contain audible directions and they are detected by an RFID reader attached at the tip of a customized cane. The reader is connected to the Raspberry Pi 3 that serves as the brain of the system. It processes the data from the reader, validates the voice input of the user and provides the audio response. The adjustable customized cane measures from 88 cm to 130 cm with an approximate weight of 165 grams, which is lighter than commercially available canes that weigh around 300 grams. The speech recognition feature of the system is evaluated and an accuracy rate of 75.76% is obtained. Its response time in reading the tags is 1.38 seconds for a maximum distance of 1 inch. Overall, the system runs autonomously which helps the visually impaired people in navigating independently inside a specific establishment.