[This article by Raynard Churchwell first appeared on the Kennesaw State University website, republished with permission]
Much like ants working together in a colony, Kennesaw State University researchers envision a future where driverless cars coordinate seamlessly with one another to enhance traffic flow, reduce congestion, and minimize emissions.
Mahyar Amirgholy, an assistant professor in the Southern Polytechnic College of Engineering and Engineering Technology and research affiliate in Lawrence Berkeley National Laboratory’s Energy Analysis and Environmental Impacts Division, is engaged in a broad spectrum of projects aimed at improving transportation efficiency and sustainability through advanced technologies. His projects include cooperative control of connected automated vehicles (CAVs), the use of advanced machine learning techniques for traffic analysis, and predictive analysis of electric consumption and emissions from the increasing demand for electric vehicle (EV) charging.
Amirgholy’s group in the Futra Lab is developing a control system that enables CAVs to communicate with both traffic lights and each other through a cloud computing network. By synchronizing their speeds while integrating with human-driven vehicles, CAVs can harmonize mixed traffic flows and align them with green lights. At the same time, CAV speed and location data is used to optimize traffic light timing for vehicles approaching intersections from conflicting directions. This coordinated interaction between vehicles and traffic signals helps reduce congestion and delays, while minimizing fuel consumption and emissions by cutting down on unnecessary idling and stop-start driving patterns.
Transportation electrification is another key area of interest for Amirgholy’s group. Although EVs are often perceived as emission-free, the growing demand for electricity to charge EV batteries places additional pressure on the power grid, potentially increasing emissions from power plants.
“The impact of rising EV energy consumption on power plant emissions largely depends on the energy mix of the grid,” Amirgholy said. “In our recent study, we compared California, Georgia, New York, and Washington. Our findings show that increased EV adoption in states like Georgia, which heavily rely on fossil fuels, will significantly raise emissions from power plants. So, transportation electrification only effectively reduces emissions if the electricity is primarily sourced from renewable energy, as it is in California and Washington, and soon will be in New York.”
Amirgholy’s group is currently developing an interactive tool that uses machine learning to predict electricity consumption from EVs and the resulting impact on grid emissions across the United States.
In partnership with Berkeley Lab, and with funding from the Federal Highway Administration, Amirgholy is also engaged in developing a national-scale multimodal transportation model utilizing machine learning. Their goal is to develop an analytical tool, called GEMS, designed to support policy analysis and investment decision-making at the national level.
Leveraging insights from GEMS, Amirgholy and his group are developing an advanced machine learning model for analyzing urban networks. He is also collaborating on two interdisciplinary projects with the University of Georgia, funded by the Georgia Department of Transportation (GDOT). These projects focus on integrating artificial intelligence and wireless sensor technologies to develop real-time monitoring systems for GDOT.
Amirgholy’s dedication to advancing transportation technologies is driven by his belief that technology can be a powerful force for positive change. He advocates for policymakers to embrace emerging technologies to enhance the efficiency and sustainability of transportation systems.
“The choice is clear,” he said. “We must either embrace these technologies to upgrade our outdated infrastructure or face a future of escalating traffic and pollution.”
Duleep Prasanna Rathgamage Don, a Ph.D. candidate in Data Science and Analytics, works as a graduate research assistant in the FUTRA Lab. His participation on the project has significantly deepened his understanding of smart transportation system modeling.
“This research has completely transformed my understanding of transportation technology,” he said. “Developing models with machine learning algorithms has provided me with valuable insights into leveraging AI for analyzing urban networks.”
Chrisley Licona-Hernandez, a First-Year Scholar in civil engineering, said her experience in the lab challenged her to step outside of her comfort zone.
“One of my biggest challenges was my lack of coding experience, which made it difficult to keep pace with my graduate peers,” she said. “However, it motivated me to learn quickly and sharpen my research skills. This experience encouraged me to seek support and build connections with my colleagues, reminding me of the value of collaboration in overcoming obstacles.”
Amirgholy’s research leverages emerging technologies to develop effective solutions for complex transportation challenges, focusing on enhancing mobility, efficiency, and sustainability for all.
“My goal is to use new technologies to develop solutions that improve quality of life for everyone, particularly underserved communities with limited access to the new technologies,” Amirgholy said.
— Story by
Be the first to comment on "Kennesaw State’s FUTRA Lab tackles traffic congestion and emissions with AI solutions"