Design and Test of a Hybrid Foot Force Sensing and GPS System for Richer User Mobility Activity Recognition

Zelun Zhang,Stefan Poslad

doi:10.3390/s131114918

Abstract

Wearable and accompanied sensors and devices are increasingly being used for user activity recognition. However, typical GPS-based and accelerometer-based (ACC) methods face three main challenges: a low recognition accuracy; a coarse recognition capability, i.e., they cannot recognise both human posture (during travelling) and transportation mode simultaneously, and a relatively high computational complexity. Here, a new GPS and Foot-Force (GPS + FF) sensor method is proposed to overcome these challenges that leverages a set of wearable FF sensors in combination with GPS, e.g., in a mobile phone. User mobility activities that can be recognised include both daily user postures and common transportation modes: sitting, standing, walking, cycling, bus passenger, car passenger (including private cars and taxis) and car driver. The novelty of this work is that our approach provides a more comprehensive recognition capability in terms of reliably recognising both human posture and transportation mode simultaneously during travel. In addition, by comparing the new GPS + FF method with both an ACC method (62% accuracy) and a GPS + ACC based method (70% accuracy) as baseline methods, it obtains a higher accuracy (95%) with less computational complexity, when tested on a dataset obtained from ten individuals.

Highlights

User mobility or activity is an important type of user context that can be used as a knowledge source to better tailor and adapt a raft of rich applications to users’ needs in different mobility-related situations
GPS and foot force (FF) sensor data clusters differently with respect to different transportation or mobility modes compared to typical accelerometer data
The potential benefits of using mobile phone GPS in combination with a set foot force sensors to improve daily mobility activity recognition have been examined for the first time

Summary

Introduction

User mobility or activity is an important type of user context that can be used as a knowledge source to better tailor and adapt a raft of rich applications to users’ needs in different mobility-related situations. The increasing use of wearable and accompanied device body sensors networked as body area networks adds a new type of sensor data to help promote an Internet of Things. These sensors can act as an enabler for the hidden computer part of Weiser’s ubiquitous computing vision to increase the implicit human computer interaction (iHCI) with systems and services through reducing users’ cognitive load, distractions and informational overload when users respond to the myriad of intelligent devices and sensors in their immediate environment [1].

Objectives

Methods

Results

Discussion

Conclusion