How traffic was modeled before the AI era

This Habr article explores the history of mathematical traffic flow modeling, dating back to the 1950s. The author examines how scientists attempted to describe driver behavior and the causes of traffic jams long before the advent of modern neural networks and big data analysis. The material covers classical approaches, such as hydrodynamic analogies and the introduction of driver 'personality' parameters into mathematical models. The article offers a look at the evolution of transport science, emphasizing that many fundamental traffic management principles were established decades ago using analytical methods. It serves as a historical overview of a discipline that has become the foundation for modern intelligent traffic management systems, demonstrating that complex algorithms have deep scientific roots that extend far beyond machine learning.