TRAILSENSE: Crowdsensing Risky Mountain Trail Segments

Abstract

Excerpted from “TrailSense: A Crowdsensing System for Detecting Risky Mountain Trail Segments with Walking Pattern Analysis,” in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT), with permission. https://dl.acm.org/citation.cfm?id=3131893 © ACM 2017 Mountain trail surface information is critical to prevent mountain accidents, such as falls. This article presents TrailSense, a mobile crowdsensing system that can automatically label risky mountain trail segments. TrailSense analyzes walking patterns of an individual hiker using smartphone sensing. The results from multiple hikers are then aggregated in the cloud for accurate labeling. Our results from two real-world datasets show that TrailSense can identify, fairly accurately, risky trail segments with crowdsensing. Mountain climbing is a popular outdoor leisure activity. In the United States, for Keunseo Kim, Hengameh Zabihi, Heeyoung Kim, Uichin Lee Korea Advanced Institute of Science and Technology (KAIST).