Principal Investigators and Members:

Nguyen Ngoc Doanh, Ph.D., College of Engineering and Computer Science and Center for Environmental Intelligence, VinUniversity.
Le Duy Dung, Ph.D., College of Engineering and Computer Science, VinUniveristy.
Laurent El Ghaoui, Ph.D., College of Engineering and Computer Science, VinUniveristy.
Doan Dang Khoa, Ph.D., College of Engineering and Computer Science, VinUniveristy.

Project summary and Key Research Aims

Urban development in Vietnam and other developing countries has led to a large increase in greenhouse gas emissions and pollution of the surrounding environment. Smart transportation can significantly reduce emissions of air pollutants, thus building resilience against climate change and its horrific effects. For example, a previous study has shown that congestion charge to vehicles entering a restricted zone in London has measurably reduced emissions of CO2, particulate matter PM, and nitric oxide (NOx). With AI-enabled learning of emission conditions measured by a sensor network and user behaviors, enhanced real-time dynamic congestion pricing can also be implemented in SEA cities for greater environmental benefits. SEA cities are selected because these cities are notorious for congestion, resulting in emissions of greenhouse gasses.  

Significance: This research will enable the potential for developing a suite of transportation policies and incentives to encourage shifts to sustainable transportation modes, such as electric vehicles, transit, and ridesharing, to reduce emissions of air pollutants and traffic congestion.  

Goals: We will develop digital twins of transportation and air quality for SEA cities by 1) Combining transportation pattern measurement and the measured concentration of key air pollutants (CO2, and PM2.5 levels); 2) Designing and evaluating economic mechanisms such as congestion pricing for reducing emissions of these pollutants by motivating public transportation, reducing idling, promoting higher fuel efficiency, and transitioning into electric vehicles. 3) Simulating future transportation scenarios for a transportation policy suite that encourages long-term sustainable mobility preferences in personal travel and freight operations.  

Innovation: While congestion charges have been implemented successfully in London and Singapore to reduce traffic and greenhouse gas emissions, these schemes are not designed based on real time data. We aim to design congestion pricing (where, when, and which types of vehicles) for minimizing the emission level, using AI/ML-enabled sensor data inputs to predict the impacts of pricing on driving behavior and greenhouse gas emission. Natural language processing and AI tools can eventually help drivers decide what mode of transportation will be most efficient and cost-effective. AI and ML tools will also elicit public and other stakeholder opinions on the long-term barriers to shifting to sustainable transportation modes for personal travel and freight deliveries.