Transportation Mode Detection by Embedded Sensors Based on Ensemble Learning

Abstract

Context-aware computing has become a certainty due to the widespread use of smartphone devices equipped with sensors. A wide range of services, such as vehicular traffic monitoring and smart parking, can be accomplished with the help of awareness of user mobility. Transportation mode detection (TMD) using machine learning algorithms and the data captured from smartphone embedded sensors have attracted research community attention. In this research, ensemble learning is utilized to differentiate between transportation modes, including walking, standing, riding a train, driving a car, and riding a bus. The ensemble learning consists of three classifiers; each classifier votes independently on the instances, and the majority vote is applied for robust generalization. The proposed method was validated using three datasets; the samples included in these datasets were gathered by smartphone sensors (belonging to heterogeneous users), such as rotation vector sensors, accelerometers, uncalibrated gyroscopes, linear acceleration, orientation, speed, game rotation vector, sound, and gyroscopes. The proposed ensemble learning method achieves an accuracy of 89%, 93%, and 95% on the first, second, and third datasets, respectively, when 10% and 90% of the data are used for testing and training, respectively. On the other set of experiments, in which 30% and 70% of the data are used for testing and training, respectively, the proposed method yields accuracies of 86.8%, 92.1%, and 94.9% on the first, second, and third datasets, respectively. The proposed method shows promising results compared to existing human activity recognition (HAR) methods.

Context-Aware Edge-Based AI Models for Wireless Sensor Networks—An Overview

“…Monitoring is the second most studied issue, namely in of the included papers. Not surprisingly, in the context of monitoring, various applications have been identified, e.g., health [ 60 , 89 , 115 , 158 , 159 ], smart buildings [ 90 , 116 , 160 ], agriculture [ 161 , 162 ], stress [ 61 , 117 , 136 , 163 ], transportation [ 91 ], military defense [ 164 ], etc. Other challenges comparatively highly studied in the included papers are QoS [ 92 , 93 , 118 , 128 , 137 , 138 , 139 , 152 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 ] with , and energy saving [ 62 , 63 , 94 , 95 , 96 , 114 , 119 , 142 , 143 , 144 , 175 ] with …”

Accuracy Improvement of Vehicle Recognition by Using Smart Device Sensors

“…Using rotation vector sensors, accelerometers, uncalibrated gyroscopes, linear acceleration, orientation, speed, game rotation vector, sound, and gyroscopes, Ref. [18] recorded activity data when subjects were standing still, walking, driving in a car, riding a bus, and taking a train. With an ensemble method of machine learning algorithms, KNNs, and random subspace, all bundled into stacked learning, they ultimately used a neural network architecture to successfully determine the transportation mode with an impressive 90% accuracy, though still decidedly lower than our 98% obtained with only an accelerometer and a gyroscope.…”

“…Measures linear acceleration, directional movement, and three-dimensional object orientation or stationing, as well as changes in the ambient environment Xia et al, 2014 [3]; Alotaibi, 2020 [18]; Badii et al, 2021 [20];…”

Data Fusion for Deep Learning on Transport Mode Detection: A Case Study

“…Transport Mode Detection (TMD) is a classification problem whose goal is to detect the transport mode of a human subject, e.g., walking, cycling, using any sensor. With inertial sensors, multiple methods coexist: some use handcrafted features with traditional learning methods (see [6,9,1], for instance), while others use directly use Deep Neural networks (Convolutional [8] or Recurrent Neural Networks [15]). For inertial sensors, the most important dataset is the SHL dataset, which is used to organize a yearly challenge [20,18,19].…”