Illustrating traffic prediction and control algorithms with 3D visualization
Jeroen van Riel (TUE)
It remains a challenge to inform non-expert audiences about research on traffic prediction and control algorithms. We developed a visualization tool that provides an intuitive presentation of various traffic models and their performance. The existing open-source sumo-web3d project provides a 3D visualization for live SUMO simulations. Our extension enables researchers to provide arbitrary vehicle motions that are either recorded using SUMO or generated with custom mathematical models. Furthermore, the tool provides a way to show arbitrary probability distributions as color gradient overlays on the lanes of the network. The tool was used to illustrate platoon forming algorithms for autonomous vehicles and efficient traffic light control algorithms.