A Framework for Robust, Fast, Efficient and Configurable Navigation of Video Game Characters

We will create new software technology for generating intelligent-looking movement behavior of virtual characters in video games. When characters move around in some game worlds, they typically plan their path around fixed obstacles and use some simple rules to avoid dynamic obstacles, such as other characters. Recently a family of methods based on the concept of Velocity Obstacles (VO) has been developed at the University of North Carolina. These algorithms have been shown to produce collision free navigation in theory and in practice. We aim to extend these methods in two ways: (1) find approximate methods with very fast and efficient implementations for real time games; and (2) make the character behavior “tweakable”, i.e. provide good behaviour over a range of parameter settings so that different virtual characters have perceptively different behaviour while still navigating sensibly

Faculty Supervisor:

Dr. Richard Vaughan

Student:

Seyed Abbas Sadat

Partner:

Koolhaus Games

Discipline:

Computer science

Sector:

Digital media

University:

Simon Fraser University