[Paper Review] Experiments of posture estimation on vehicles using wearable acceleration sensors

Yoji Yamato|arXiv (Cornell University)|Jun 6, 2017

Ergonomics and Musculoskeletal Disorders14 references3 citations

TL;DR

This paper proposes a method to estimate drivers' postures in vehicles using acceleration data from a T-shirt-type wearable sensor (hitoe), focusing on detecting dangerous postures like picking up objects. By applying threshold-based analysis on Y-axis acceleration changes to detect body inclination, the method successfully identifies posture shifts without relying on smartphone-based acceleration subtraction, which suffers from sensor accuracy discrepancies. Field tests confirmed the method's feasibility and low false-positive rate even on steep slopes.

ABSTRACT

In this paper, we study methods to estimate drivers' posture in vehicles using acceleration data of wearable sensor and conduct a field test. Recently, sensor technologies have been progressed. Solutions of safety management to analyze vital data acquired from wearable sensor and judge work status are proposed. To prevent huge accidents, demands for safety management of bus and taxi are high. However, acceleration of vehicles is added to wearable sensor in vehicles, and there is no guarantee to estimate drivers' posture accurately. Therefore, in this paper, we study methods to estimate driving posture using acceleration data acquired from T-shirt type wearable sensor hitoe, conduct field tests and implement a sample application. Y. Yamato, "Experiments of Posture Estimation on Vehicles Using Wearable Acceleration Sensors," The 3rd IEEE International Conference on Big Data Security on Cloud (BigDataSecurity 2017), pp.14-17, DOI: 10.1109/BigDataSecurity.2017.8, May 2017. "(c) 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works."

Motivation & Objective

• To address the challenge of accurately estimating drivers' postures in vehicles due to interference from vehicle motion.
• To evaluate the feasibility of using wearable acceleration sensors (hitoe) for real-time posture monitoring in moving vehicles.
• To compare and validate two methods: acceleration subtraction using a smartphone and threshold-based posture detection from hitoe data.
• To develop and test a sample application that visualizes posture changes and enables cloud-based safety management.
• To lay the foundation for future field trials with bus companies using combined posture and fatigue monitoring.

Proposed method

• The study uses a T-shirt-type wearable sensor (hitoe) to collect 3D acceleration data from drivers in real vehicles.
• A threshold-based method detects posture changes by analyzing Y-axis acceleration changes, with a -0.34G threshold corresponding to a 20-degree body inclination.
• Noise in acceleration data is reduced using a low-pass filter to improve detection accuracy.
• The method avoids reliance on smartphone-based acceleration subtraction, which introduces errors due to sensor accuracy differences.
• A sample application processes the filtered data, visualizes posture changes, and transmits data to the cloud for further analysis.
• Cloud-based analysis uses heart rate variability (RRI) and cardiac vagal index (CVI) to estimate fatigue and relaxation levels.

Experimental results

Research questions

• RQ1 Can wearable acceleration sensors accurately estimate driver posture in moving vehicles despite vehicle-induced motion?
• RQ2 Does subtracting smartphone-acceleration from wearable sensor data provide reliable posture estimation?
• RQ3 Can threshold-based detection of Y-axis acceleration changes effectively identify dangerous postures like picking up objects?
• RQ4 How does the method perform under challenging conditions such as steep slopes?
• RQ5 Can a real-time application successfully visualize posture changes and support cloud-based safety management?

Key findings

• The method of subtracting smartphone acceleration from hitoe data was found to be unreliable due to inherent accuracy differences between sensors, even after noise filtering.
• Threshold-based detection using Y-axis acceleration changes (set at -0.34G) successfully identified posture changes such as forward bending during object retrieval.
• No false positives were detected during field testing on a steep Mt. Fuji climbing bus with a 20% gradient, confirming robustness under high vehicle tilt.
• The sample application successfully visualized posture changes in real time during field testing on a regular urban bus.
• The system demonstrated feasibility for real-time posture monitoring and integration with cloud-based fatigue and health status analysis.
• Field trials with bus companies are planned to further validate the combined monitoring of posture and fatigue using hitoe ECG and acceleration data.