Comfort zone prediction around commuter for personal mobility device : pedestrian

• Main Author: Murugesan, Jeyakaran
• Other Authors: Justin Dauwels
• Format: Theses and Dissertations
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/78626
• Institution: Nanyang Technological University

For an eco-friendly mode of transportation Personal Mobility Devices (PMDs) are being used for minimal distance travel inside the city environment. But before allowing Personal Mobility Devices (PMDs) in shared paths, there is a need for analysis in terms of safety factors and the impact of PMD with the users in shared paths. Mainly this study focuses on predicting the comfort zones for four different PMDs (Go Cycle, Inokim, Schaeffler, Zero). Inferences for comfort zone are being done in two ways, data visualization and analysis using machine learning classifier model. This data visualization is purely based on the data extracted from the videos and the responses from participants. Machine Learning analyzing is done using two classifier algorithm (Random forest and SVM), developed based on diverging distance and passing distance for predicting the comfort state ( comfortable or uncomfortable). The prediction model helps the rider to know whether they are causing discomfort to the pedestrians they meet. Weights of the features affecting the comfort zone are also analyzed by training the models with a different combination of features (age of the participant, type of PMD, the gender of the participant, diverging distance and passing distance).