Constraint Systems Laboratory

1
Herds Moving Over 3D Terrain
Joel Gompert
Constraint Systems Laboratory Department of
Computer Science and Engineering, University of
Nebraska-Lincoln jgompert_at_cse.unl.edu
Abstract Techniques are explored for simulating
and animating life-like herds living on
three-dimensional terrain, while being efficient
enough to run in real-time. The motion of the
animals is controlled using simple modifications
to Reynold's agent-based flocking algorithm. The
modifications use only local properties of the
terrain, and thus have very low complexity. The
flocking algorithm with these modifications
produces naturally behaving herds that follow the
terrain. They will swerve around hills and
attempt to follow paths that reduce energy
expenditure. The terrain-following rule added to
the flocking algorithm has a constant parameter
that can be adjusted to produce different
behaviors. Empirical analysis shows how this
parameter affects energy expenditure of the
animals while traveling over the terrain.

• Craig Reynolds Flocking Algorithm each animal
  follows a set of simple behavioral rules which
  are combined. All of the animals each following
  these simple rules results in an emergent flock
  behavior.
• Flocking Rules
• Collision Avoidance accelerating away from
  nearby members of the flock
• Velocity Matching adjusting an animal's velocity
  to match the average velocity of its neighbors
• Flock Centering accelerating towards the center
  of mass of the flock

To the left, you can see a top-down view of the
animals moving towards the lower-right of the
picture. In this case they are dividing into two
groups to go around this tall hill rather than
going over the top. They will regroup on the
other side of the hill.
Below, the animals are traveling among obstacles.
They use simple, local rules to navigate past
obstacles to their goal.
Several parameters emerged from this herd
algorithm. This table lists these parameters and
shows their allowable ranges and relationships
between some of them.
Landmark values Landmark values
Parameter Minimum Maximum
Animal size 0 8
Max speed Cruising speed 8
Cruising speed 0 Max speed
Max acceleration 0 8
Comfort zone Animal size 8
Danger zone Animal size Comfort Zone
Visible flockmate range Comfort Zone Distance to farthest flockmate
Visible obstacle range Animal size 8
Viewing angle p/2 p
Drag factor 0 55

• Future Work
• Plans for future study include higher-level
  behaviors which may include a more sophisticated
  analysis of the terrain.
• Internal world models for navigation
• Higer-level behaviors such as the herd choosing
  its own travel destinations (to find food and
  water or avoid predators)

This work is supported by NSF CAREER Award
0133568.