Growing up, nothing fascinated Jidong Yang more than how things were built.
It began when he learned about the Great Pyramid of Giza. He marveled at how such a large structure was built with so little technology.
“I was so curious about the process, and it sort of became my motivation,” said Yang, associate professor in the University of Georgia’s College of Engineering. “I began to look at structures and wonder how they were built.”
From buildings to bridges, and nearly every human-made structure throughout civilization, Yang wondered how they were designed, constructed, and integrated into the natural world. From that moment, he knew he wanted to study engineering.
Yang received his doctorate in civil engineering from the University of South Florida. In the years following, he worked in both private and public sectors, conducting design and operational analysis of transportation networks and investigating safety for roads and bridges.
“The transportation side of engineering isn’t just infrastructure,” he said. “It requires a systems-level approach: optimizing traffic flow across the network under both normal conditions and disruptions, such as hurricanes, while assessing and mitigating the risk of accidents.”
For the past six years, Yang has researched at the intersection of artificial intelligence (AI) and transportation. For example, an ongoing challenge in autonomous vehicle development is ensuring this evolving technology can adapt to diverse weather conditions.
Autonomous vehicles perceive their surroundings through cameras and sensors that continuously scan the road, identifying pedestrians, other vehicles, traffic signs, and potential hazards. Advanced AI algorithms process this visual data in real time, enabling split-second driving decisions—just like a human but with faster reflexes and no distractions.
However, the technology isn’t foolproof.
Self-driving cars often struggle in rainy conditions because precipitation obstructs and impairs their “vision.” To address this challenge, Yang’s team has developed deep-learning-based vision models that process live images from the vehicle’s cameras, removing rain-induced visual hindrances and generating images that resemble clear, rain-free scenes.
To train this system, researchers used the Car Learning to Act (CARLA) simulation, a virtual driving environment, to collect images of roads in both clear and rainy conditions. The AI model was designed to function like an image translator, removing rain-related noise and restoring clear visuals. A unique training method helped the model distinguish between raindrops, which change quickly, and the actual road scene, which remains more stable as vehicles navigate.
The system could enhance the safety and reliability of autonomous vehicles overall, improving the likelihood of future widespread deployment and adoption.
In addition to enhancing the safety of autonomous driving, Yang’s team has leveraged foundation models, such as Multimodal Large Language Models, to improve safety reasoning and enable causal analysis and inference.
With the growing popularity of electric vehicles (EVs), Yang’s team has also studied network optimization for charging stations. When your gas tank runs low, it’s usually simple to find a nearby filling station. But if you’re driving an electric car, finding a charging station nearby can prove more difficult.
Yang’s team collects data on charging station locations and explores smart system concepts to facilitate seamless cooperation between charging stations and EVs . This could help local governments and businesses identify ideal sites for new charging stations, ensuring efficient infrastructure deployment.
“I strive to weave my industry experiences into my teaching. I try to design my material in such a way that a student can easily absorb it and relate it to their own experience. Teaching and research are inseparable to me.”
– Jidong Yang, associate professor in the College of Engineering
EV infrastructure is rapidly growing nationwide, with Georgia leading the Southeast in both number of registered EVs and EV charging stations. For more than 30,000 EV drivers on the road, there are currently 2,144 stations housing 5,749 charging ports in the state, according to Georgia.org.
While research and development continue to be a deep passion for Yang, he’s found a profound sense of purpose educating students and shaping the College of Engineering’s curriculum and strategic direction. Looking back on his years in industry, he recognizes the value of that experience in the classroom.
“I strive to weave my industry experiences into my teaching,” he said. “I try to design my material in such a way that a student can easily absorb it and relate it to their own experience.
“Teaching and research are inseparable to me.”
Yang constantly updates his teaching material with his own research, as well as the latest developments in the field. While his expertise has deepened over time, the rapid advancements in AI and emerging technologies have reshaped his perspective, fueling a renewed curiosity about their transformative impacts on engineering.
“Today’s AI, driven by large foundation models, is incredibly powerful, particularly their general knowledge and reasoning abilities,” he said. “It has laid a new foundation for how we approach problems. So now, in both my courses and research, I’m looking at how we can build upon that foundation and push the boundaries of AI-Human collaboration and synergy.”
