Flocks, herds and schools: A distributed behavioral model
SIGGRAPH '87 Proceedings of the 14th annual conference on Computer graphics and interactive techniques
Introduction to AI Robotics
| Hi-index | 0.00 |
Craig Reynolds introduced the concepts of simulated flocking behavior via the use of simple aggregate Boid behaviors which make the Boids appear to exercise independent autonomous actions. The term Boid is used to identify the actors within the flock without attaching any particular animal characteristics with the actor (e.g., fish, bird, mammal). Artificial potential-fields (APF) are often used to implement collision avoidance for robotic and multi-agent systems, making them well suited for implementing flocking behaviors. AFP can also be used to define constraints that direct the motion of a flock while still allowing for simple aggregate Boid behaviors. Flocking behavior is used in computer games or movies where groups of Boids are to move together without appearing to move in lockstep. This paper explores the approach used for directing the flocking behavior of Boids via APF constraints. The APF constraints are intended for an end user (e.g., animator or director) inexperienced with programming or the technical details of implementing flocking behavior or APFs. Our constraint system for controlling the motion of the flock is currently composed of two constraint types: A fly-by waypoint constraint is used to control the direction of the flock in a manner similar to the scripted flocking defined by Reynolds. A user defines 3D waypoints on the screen to map a sequential path that the flock is to follow. A boundary constraint is used to control the outward spread of the flock. A user defines a boundary between waypoints that sets the maximum outward spread of the flock by limiting the flock to move within the channel/conduit defined between the waypoints.