Shepherding via Deformable Shapes
Joseph F. Harrison, Christopher Vo, Jyh-Ming Lien
Overview
This work investigates a new motion planning strategy for shepherding in which the shepherds views the flock as a deformable shape. This abstraction allows us to more accurately represent the configuration space, yet do so in an efficient way. The new algorithm, called DEFORM, is more robust than our previous approach and it scales more effectively to larger teams of shepherds and larger flocks.Video
The following videos show the shepherding simulation in action. The black agents are shepherds, the brown agents are sheep. The green shape shows the abstracted flock configuration and the red shape shows the intermediate target configuration. The blue squares indicate useful places for shepherds to occupy in order to guide the green shape toward the red shape.48 sheep and 5 shepherds on the "S" environment
more sheep and more shepherds on the other environment
Publication
Scalable and Robust Shepherding via Deformable Shapes, Joseph F. Harrison and Christopher Vo and Jyh-Ming Lien, George Mason University, (Technical Report), Jan. 2010
Web Site / Tech Report (pdf) / BibTeX
Web Site / Tech Report (pdf) / BibTeX
Scalable and Robust Shepherding via Deformable Shapes, Joseph F. Harrison and Christopher Vo and Jyh-Ming Lien, Proceedings of the Third International Conference on Motion in Games, Springer, 2010
Web Site / Paper (pdf) / BibTeX
Web Site / Paper (pdf) / BibTeX
See Also
Behavior-Based Motion Planning for Group ControlInteractive Shepherding
List of MASC Research Pages