Method, Design and Implementation of an Indoor Tracking System with Concurrent Fault Localization

Abstract

Thanks to the diffusion of wearable devices there are several indoor tracking systems. Among them, RF-based have been deeply studied for their flexibility and limited costs. These systems can be employed as assistive tools only being dependable, identifying faults. We propose two methods to provide multiuser tracking with concurrent localization of natural hardware and human-made faults. The first method relies on independent measurement systems and on a model-based fault localization apparatus, checking for discrepancies in the subsystems behavior. The second provides an estimation of the fault probability for each device, based on the data collected at runtime. These methods aim to provide dependable tracking for fragile people (such as elderly or people with small impairments). We present examples of Indoor Human Tracking simulations in a large environment, and an implemented case-study. The collected data confirm the validity of both the approaches and highlight their diversity.