UMANS (Unified Microscopic Agent Navigation Simulator) is a crowd-simulation framework that can reproduce many different algorithms for local navigation behavior (such as collision avoidance) in crowds. UMANS translates these algorithms to a single principle, while unifying as many simulation settings and details as possible. This makes it a suitable platform for experimentation and for honest comparisons between algorithms.
For more information on the theory behind UMANS, please see our I3D 2020 publication “Generalized Microscopic Crowd Simulation using Costs in Velocity Space”.
The Git repository of UMANS is public, so everyone can download the latest version of the source code.
We have also published pre-compiled binaries. The latest stable release can be downloaded via our download page.