Is it Possible to Detect Mobile Phone User’s Attention Based on Accelerometer Measurment of Gait Pattern? Josip Musić, Ivo Stančić, Vlasta Zanchi Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture University of Split Split, Croatia jmusic@fesb.hr , istancic@fesb.hr , vzanchi@fesb.hr Abstract— Mobile phones have become ubiquitous in today’s world. Their ever increasing computational power and sensing capabilities have made them well suited for number of tasks well beyond their original purpose of communication. But mobile phone usage while walking or driving can potentially be dangerous leading to serious injury or even death. In the paper we answer the question is it possible using only mobile phone’s embedded accelerometer to detect changes in gait pattern caused by changed attention level due to interaction with mobile device like reading on-screen text. Experimental measurements were conducted on 8 test subjects in indoor environment with each test subject performing 6 trials. Two different approaches based on gait phase and gait velocity were tested on recorded data in batch mode with more promising one implemented in real-time manner. Obtained results are presented and discussed and possible future research directions outlined. Keywords-human gait; accelerometer; smartphone; phase; attention detection; safety I. INTRODUCTION Number of cell phones in the world is constantly increasing making them ubiquitous. In the process, with inclusion of additional hardware capabilities, they have outgrown they initial (and only) purpose of communication. Smart phones especially are getting more and more sensors built into them with increasing computational power making them ideal for certain applications which in the past required expensive and cumbersome laboratory equipment. These applications are mainly in the domain of measuring, recording and helping humans in situations of everyday life. Some of interesting applications include biometrics [1], activity classification [2], gait anomaly detection [3], dengue fewer detection [4] and so on. At the same time some undesired side effects of mobile phone usage in everyday life have surfaced when number of mobile phones had reach a critical mass so that their life taking capacity has overcame their life saving potential [5]. Most significant is injury and even death because of divided attention while simultaneously performing some other task, most notably driving [6, 7, 8]. But serious injuries can also occur while using the mobile phone during walking due to decreased situational awareness [9, 10]. Ohio State University research [11] showed that in 2008 over 1000 people visited emergency rooms because of injuries sustained while walking and using mobile phone which caused subjects to fall or hit obstacles in the way. Similar research conducted by Consumer Reports National Research center [12] on USA representative adult sample showed that 85% of pedestrians used mobile device in busy area with 42% of them bumping into something/someone in the way and 34% of them stepping in front of moving vehicle. It is believed that many additional injuries occurred but haven’t been reported. This phenomenon has becomed so severe that certain USA states/cities are considering to implement or have implemented (e.g. Fort Lee, New Jersey) fines for “careless walking” and are launching public awareness campaigns. Thus, systems are being developed to detect undesired user behaviors while using mobile devices and alert users so that serious consequences can be avoided. For example in [13] mobile phone usage by the driver can be detected based on specially generated sound tones via car speakers (using Bluetooth) and measurement of time-of-flight. This information can then be used to limit or completely block mobile phone usage while car is moving [14]. Another system aimed at pedestrian safety is WalkSafe [15]. This system employs mobile phone’s camera to detect incoming vehicles while user is talking on the phone and crossing roads. Recent research [16] has shown that gait style changes when person is using his or hers mobile phone. If one was able to detect these changes in gait pattern by means of available on board sensors (accelerometers, gyroscopes, magnetometers or even camera) then timely and appropriate warnings could be issued to the user so that dangerous situations could be avoided. As a side effect, this information could be used to conserve battery power since screen could be turned on and off as needed. In the paper accelerometer measurements are used to that end since they are available in most mobile phones. The hypothesis is that when a person walks and reads text on mobile phone screen it needs to adjust gait (e.g. steady it) so that reading task can be achieved, and when he/she walks freely that constraint is not present. To that goal we employ two approaches: one based on gait phase characteristic and other based on gait velocity profiles. The article is structured as follows. In next section materials and methods used are presented, followed by section three in which obtained results are presented and