How Knowledge about the Environment Influences the Navigation of a Robot through a Crowd

Thumbnail Image
Issue Date
Journal Title
Journal ISSN
Volume Title
Robots are integrating more into society and some of them have to navigate through crowds. For those robots it is important to know how much knowledge they need about the environment to efficiently navigate. This Bachelor’s thesis examines how the amount of knowledge about the environment influences the navigation of a robot through a crowd. The influence has been researched using a simulation created in Unity. The knowledge has been provided by the Navigation Mesh (NavMesh) which the robot used to navigate. The size of the NavMesh has been varied to change the amount of knowledge. The influence of the knowledge has been researched in different environments with different properties, such as static and dynamic obstacles and narrow and broad passages. The size of the crowd has also been manipulated to see if that influences the effect of knowledge. The results show that the knowledge has no systematic influence on the number of collisions. It does affect the time a navigation takes in broad environments, narrow scenes show no effect. However, the additional benefit of more knowledge is quite minimal. This concludes that a robot’s navigation through a crowd can do relatively well with relatively little knowledge.
Faculteit der Sociale Wetenschappen