Research Area
Keywords: Simulation, Smart Mobility, Automated Mobility-on-Demand (AMOD), Intelligent Transportation Systems (ITS), Predictive Analytics
My research motivation is to deal with the future uncertainty in the transportation system by leveraging the recent advancements of ICT technology from both data-driven and simulation-based approach. To improve the transportation system, I explain the traffic issues by analyzing large-scale traffic data, and develop prediction/control mechanisms through machine learning models. And, finally, I evaluate the future mobility scenario through simulations.
Behavioral Modeling and Simulation
My research interests include understanding travel behaviors and developing simulation models based on travel behavioral models and traffic flow theories. To make the models consistent with the real world, I incorporates microscopic characteristics in the macroscopic domain, contributing to the advances of traffic monitoring and control strategy. A fundamental to this research is developing simulations that are used to replicate/predict human behavioral changes, and a significant part is focused on improving the model systems by i) enhancing the models in a more realistic way, ii) understanding the interaction between demand-supply of transportation, and iii) calibrating/validating the model against the real-world.
Design and Optimization of Smart Mobility System
One of major application of simulation model is to investigate the effects of a portfolio of emerging technology, policy and investment options under alternative future scenarios. My emphasis here is to reduce inequality, increase accessibility of travelers to the public transit, and impact car-ownership behavior through Smart Mobility solutions such as Automated Mobility-on-Demand (AMOD) system. For the successful integration, I conduct the researches to address the implication of this new mobility service under the local context for both demand (the individuals’ preference and travel behavior towards new service) and supply (from the service operational configuration but also from the intrinsic city level infrastructure setting and usage, including congestion effects). It features i) detailed interactions between the agents of demand (travelers) and supply (infrastructure and the transportation operations) and ii) multiscale evaluations in time and space, comprising long-, mid-, and short-term prediction modules in which we consider different decision-making levels of an urban system.
Big Data-driven Prediction & Control
Another research focus of mine is predictive modeling, where I have an expertise in pattern recognition and travel behavior. My general approach is to increase the applicability of pattern search algorithm in a big-data environment, without sacrificing predictive accuracy, whilst accelerating inference algorithms through optimizing search structures. The final goal is to leverage both predictive information and future travel behavior changes by establishing a feedback loop between the predictive models and route choices.
For more information: