Author: Wenlong Jin
Institute of Transportation Studies, University of California, Irvine
At ITS-Irvine, researchers are advancing the science of automated vehicle safety to ensure a smoother transition toward mixed human-autonomous traffic. In this study, Dr. Wenlong Jin addresses a critical gap in how automated vehicles (AVs) interpret and respond to real-world driving conditions. Current car-following models often fail to mimic the nuanced braking and spacing behaviors of human drivers, leading to potential safety and traffic-flow issues—especially at intersections designed for human decision-making.
To overcome this challenge, Dr. Wenlong Jin developed a multi-phase projection-based model that allows AVs to anticipate future braking scenarios rather than simply react to present conditions. This approach produces driving patterns that are both safer and more human-like, reducing the likelihood of “dilemma zones” and improving coordination between AVs and traditional vehicles. The work provides an essential step toward ensuring that automated vehicles can coexist harmoniously with human drivers and the infrastructure shaping California’s transportation future.
Key Research Findings
- Standard car-following models have significant limitations in maintaining safe distances between vehicles in everyday driving situations
- Human drivers are able to anticipate when other vehicles may brake, but existing car-following models fail to adequately do this
- Our car-following model mathematically guarantees safety while exhibiting human-like behavior
- The model can be readily installed in AV systems
Stay connected with ITS-Irvine for more insights on mobility innovation. To learn more...
Read about this project Read the Policy Brief