Detecting Single-Hand Riding with Integrated Accelerometer and Gyroscope of Smartphone Xuefu Dong, Zengyi Han, Yuuki Nishiyama and Kaoru Sezaki The University of Tokyo Meguro, Tokyo, Japan {dongxuefu,hzy}@mcl.iis.u-tokyo.ac.jp,{yuukin,sezaki}@iis.u-tokyo.ac.jp ABSTRACT Single-hand cycling poses a safety threat with the decrement of riders’ response capacity. Recognizing risky behavior by preva- lently used smartphones could lead to enhanced riding safety. In this work, we propose a single-hand cycling recognition method based on motion data acquired from the three-axis accelerometer and gyroscope integrated into a handlebar-installed smartphone. We conducted a 4-person experiment. The data result demonstrates that motion data of double-hand cycling clearly distinguishes from that of single-hand, revealing the chance to materialize a robust de- tection tool in smartphones to enable safer biking. For future work, we prepare to redesign the experiment under more sophisticated circumstances with an improved platform, thus scaling this sensing method for real-life usage. CCS CONCEPTS · Human-centered computing → Ubiquitous and mobile com- puting. KEYWORDS Human Activity Recognition, Accelerometer, Gyroscope ACM Reference Format: Xuefu Dong, Zengyi Han, Yuuki Nishiyama and Kaoru Sezaki. 2021. De- tecting Single-Hand Riding with Integrated Accelerometer and Gyroscope of Smartphone. In Adjunct Proceedings of the 2021 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2021 ACM International Symposium on Wearable Computers (UbiComp-ISWC ’21 Adjunct), September 21–26, 2021, Virtual, USA. ACM, New York, NY, USA, 2 pages. https://doi.org/10.1145/3460418.3479294 1 INTRODUCTION The safety of riding trips needs to be paid added attention. Globally, 86% of bicycle travels take place on roads with extremely restricted conditions such as poor surface, lack of safe crossings, and mixed usage with trafc of velocity limitation above 60km/h 1 . Due to violation of handlebar operation stipulation, more than 1544 bicycle- related accidents happened on average annually in Japan alone over 1 WHO, https://www.who.int/publications/i/item/9789241565684/ Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for proft or commercial advantage and that copies bear this notice and the full citation on the frst page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s). UbiComp-ISWC ’21 Adjunct, September 21–26, 2021, Virtual, USA © 2021 Copyright held by the owner/author(s). ACM ISBN 978-1-4503-8461-2/21/09. https://doi.org/10.1145/3460418.3479294 the past decade 2 . Hence it is essential to recognize jeopardizing single-hand riding to improve the safety concern of bicyclists and other road users. It would be straightforward to place touch sensors in the handlebar to acknowledge the holding situation and even the force applied to it, but the high installation cost often makes it unpractical for individual users and uneconomical for bike-sharing providers. The prevalence of smartphones installed in the handlebar for navigation enables versatile data collecting, as well as security alert delivering. Previous work has investigated sensing methods for mobile safety[3][2]. However, these schemes do not consider handlebar gripping. Therefore, to fll this gap, we propose a single-hand riding de- tection method. Our work aims to predict the handlebar gripping scenario from the acceleration and gyro sensor integrated into the smartphone. A data-collecting platform was installed in the test bike and the collected riding behavior data help trained the machine learning model. Evaluation demonstrates the high validity of our model. 2 RELATED WORK This work follows the category of Cycling Behavior Recognition Schemes with Smartphones. The rise of smartphones facilitates an increasing number of researches aimed at smartphone-based human activity behavior sensing. Wu et al.[1] developed BikeMate, a multi-function system that is able to study individual-level char- acteristics of dangerous riding behavior by a smartphone in the pocket, including retrograde riding and lane weaving. Usami et al.[4]utilize the embedded motion sensor of a handlebar-installed smartphone to build a recognition method of basic bicycle behavior, through noise cancellation and random forest classifer. This paper is inspired by these previous works and complemen- tary to them in detecting extra jeopardizing behavior of single-hand cycling with the smartphone. We use the handlebar-installed cell- phone and plan to develop a detect-in-pocket version in the future. 3 SYSTEM DESIGN This section briefs the steps to estimate the rider’s handlebar grip- ping scenario. 3.1 Collection Platform Our test bike is as shown in Fig.1(A). We employed a Raspberry-Pi- based capacitive touch sensor to monitor the ground truth of the gripping situation of conductive-paint-covered handlebars. A Pixel 3a smartphone is utilized to measure the acceleration and gyro data. The installation of the smartphone is marginally inclined for better 2 National Police Agency, https://www.npa.go.jp/news/release/2021/20210218jiko.html 19