OTIV

OTIV is a company that focuses on creating assistance systems within the rail industry (trains and trams).

In this project we’ve build a virtual world in which tram AI algorithms can train their models. The goal was to replicate a real city layout. Based upon data from Open Street Map the students used Houdini to procedurally generate the layout of a piece of Ghent, which then got integrated in an Unreal 5 level in which vehicles, trafic lights and more were added.

A dynamic wheather system made it possible to drive the tram in various lighting settings and weather conditions.

With this world, several virtual camera feeds could be created which serve as input for that training program.