Accurate positioning using long range active RFID technology to assist visually impaired people

Abstract

The aim of this paper is to describe a new positioning technique to assist the blind and people with low vision to indicate their location and reach their destinations in both indoor and outdoor environments. The proposed technique is based on a combination of power attenuation and a received signal strength indicator (RSSI) using active radio-frequency identification (RFID) technology. The system uses a mobile reader with a power attenuation feature. RSSI is used as a quantized distance estimator for a short range and in combination with one of eight receiver attenuation level settings for a wider range of up to 70 m. A Global Positioning System (GPS) works efficiently in a similar environment but is only accurate to around 10–20 m and does not work efficiently in indoor environments. This research produced a system that identifies various locations such as offices, laboratories, theaters to assist users in reaching their destination of interest. It was then implemented in an indoor environment as an empirical case study to identify laboratories based on a combined technique with a successful identification rate of around 98%. The reader has eight attenuation settings, and the geographic range of each level using various tags was calculated. Then, to evaluate reliability, 6 experiments with 108 samples were conducted using three tags with distances from 1 m to 25 m, using power settings 1–6. A successful detection rate of 93.5% was achieved, as well as a false positive rate of 1%. Following this, the system was implemented in a park to evaluate its ability to indicate the position of the reader among a grid of tags in an open area. A satisfactory result was achieved.

Introduction

For blind people and those with low vision, determining their current location is a significant challenge. Over the years, various researchers have discussed navigation issues using RFID technology in order to assist visually impaired people. This work can be classified into three categories. The first group of researchers used passive RFID technology for positioning in indoor environments. An advantage of passive RFID technology is that it does not require a power supply because it depends on the power of the probe signal itself. Furthermore, passive RFID is relatively inexpensive. The second group of researchers used active RFID technology in both indoor and outdoor environments, but almost all of their proposed systems did not use active RFID technology by itself, rather they combined it with other technologies such as GPS. The third group of researchers used active RFID exclusively, based on distributing transmitters (readers) in the ceiling for triangulation purpose then uses the Received Signal Strength Indicator (RSSI) technique to estimate the position of the tag that is carried by a user. However all these methods have some disadvantages:

• The disadvantage of passive RFID approach is that it requires a very short distance to communicate between the reader and the tag.

• The disadvantage of the second approach is that when GPS is unavailable, such as in between skyscrapers or inside buildings, their system is disabled or may provide inaccurate positioning information.

• The disadvantage of the third approach is more costly strategy because it is based on distributing readers rather than distributing tags.

In this research which is a development of our previous research (Alghamdi and van Schyndel, 2012), we extend the indoor-only method described there to include an outdoor environment. The improved system is able to inform users of their current position – whether indoors or outdoors - and provide them with useful information to guide them to their destination using active RFID technology.

The novelty of the system is the use of RFID for positioning the blind using antenna gain controls to selectively receive RFID tag responses at different signal strengths. So instead of measuring the signal strength of the RFID signal directly, the signal strength is implied using a set of impedance settings covering a number of ranges. The process is automatic and quantizes the RFID distances. This approach will more easily tolerate small environmental differences.

The direct measurement of signal strength is still used, but only for closer proximity measurement where the uncertainty is reduced. This new technique automatically combines these two methods at various ranges to produce a range estimate for each RFID tag. We have achieved a successful detection rate of 93.5%, as well as a false positive rate of 1%, and it could estimate the position of the reader in outdoor settings with less than 2 m uncertainty in all cases. The proposed system works efficiently in circumstances when GPS fails to work (especially indoors) and provides useful information to blind people with higher accuracy than GPS.

In deployment, the system will appear as shown in Fig. 1, where a user, when entering a building, would receive a floorplan and a list of tags via Wifi, indicating tag positions and labels. The system would then use these to allow positioning indoors. For instance, when users would like to check whether a floor level has an elevator or not, the system will inform them from the downloaded file and estimate the distance to the elevator, indicating the user's position from the destination of interest. In contrast, a passive RFID system cannot provide this kind of service. Also, in the outdoor environment, the system is able to indicate the position of the user on a pedestrian path and inform them of surrounding objects of interest. We do not intend to replace the cane, only to provide an independent positioning system at meter resolutions.

The rest of the paper is structured as follows: related work is outlined in Section 2; the system description is given in Section 3; the results and the case studies are described in Section 4; followed by the conclusion in Section 5.