AI for road safety and sustainable mobility
Center adviser
Research connects deep learning for crash-severity prediction and automated traffic surveys with transport planning, travel behaviour and sustainable mobility. Selected work spans Thai highway safety, electric vehicles and public transport.
Published research, with a short guide to the question and method.
12 selected publications
Results in Engineering, 27, 106155
A VGG-based convolutional model predicts crash severity from Thai highway records for 2011–2023. Categorical features become grayscale images, with class balancing and Bayesian optimisation; evaluation uses F1-score and AUC.
Proceedings of the 14th Asia-Pacific Conference on Transportation and the Environment (APTE 2025), 188–198
A camera-based traffic survey device detects, counts and classifies vehicles into 13 Department of Highways categories. Image processing and machine learning support automated traffic-volume collection.
Proceedings of the 14th Asia-Pacific Conference on Transportation and the Environment (APTE 2025), 116–126
Statistical association measures and a convolutional deep-learning model examine factors linked to fatalities in Thai highway crash records. The study highlights safety equipment, road-user type and lighting conditions.
Proceedings of the 14th Asia-Pacific Conference on Transportation and the Environment (APTE 2025), 127–136
A survey of 234 participants at Chulalongkorn University combines factor analysis and logistic regression to examine how safety awareness relates to recommending rental e-scooter services.
IATSS Research, 48(1), 55–67
Uses questionnaire responses from Bangkok and Sapporo with structural equation modelling to compare factors associated with teenage road-user violations and identify context-specific safety measures.
Kasetsart Journal of Social Sciences, 45(1), 235–246
Surveys across five Silor routes in Bangkok compare travel patterns and service perceptions among users and non-users. Statistical classification identified four user groups with different needs and expectations.
Sustainability, 15(6), 5273
Structural equation models compare how service quality, passenger confidence and health measures relate to bus-use intentions in Bangkok and Jakarta during the pandemic.
Sustainability, 14(14), 8482
An online survey of 403 participants in Bangkok and nearby areas was analysed using structural equation modelling. The study examined how technology acceptance and environmental concern relate to battery-electric vehicle purchase intentions and use.
Engineering Journal, 23(6), 1–18
Uses factor and cluster analysis to identify four groups of Silor passengers in Bangkok, linking service preferences to reliability, the in-vehicle environment, comfort and convenience, and environmental impact.
Engineering Journal, 22(3), 109–122
Ordered probabilistic models analyse 483 traveller surveys to study bus and informal-transport use in Phnom Penh. Costs, traveller characteristics, comfort and availability help explain differences in usage frequency.
Journal of the Eastern Asia Society for Transportation Studies, 10, 1258–1269
Analyses responses from 661 MRT passengers using factor analysis and structural equation modelling. Travel convenience was the strongest factor associated with overall satisfaction, with differences across passenger groups.
IEEE Transactions on Intelligent Transportation Systems, 14(1), 388–397
Examines how variable-message-sign design relates to motorists’ stated route diversion in Bangkok, using choice models and message-policy experiments.
Try another keyword, researcher or year.
Selected research outputs follow the original sources, including work at different institutions. Not every item is a current center project.
← All researcher portfolios