Biologically Based Mobility Models

Examples of simulations that contain individuals operating through space are more common in the ecological literature. Computer simulation has been a very heavily used tool in investigations of ecosystems for several decades. Of particular interest to this study are simulations that feature entities that are able to find their way through a landscape.

One metaphor of entities finding their way through their environment is that of foraging behavior. A mathematical expression of foraging behavior is presented in Stephens and Krebs (1986). Simulations using foraging behavior along with pheremone cues in ants are described in Langton (1986) and Resnick (1996), and examples of agents navigating using vision fields are presented in Booth (1997). This describes behavior where the entity is searching for something within a specified search radius, and goes wherever it may to find the object of its search. It is not as helpful in describing movements by an entity such as a recreational boat that already has some notion of where it wants to go when it starts.

An increasing number of studies are addressing the mobility of agents or entities through computer-generated landscapes or environments. One of the most frequently cited examples of simulations demonstrating emergent behaviors was written by Craig Reynolds as a computer graphics exercise (Reynolds, 1987). In the simulation, titled BOIDS, Reynolds programmed a number of entities to follow a few distinct rules; mainly fly at a reasonable momentum, maintain a certain maximum distance from each other, and avoid collisions; this was able to reproduce very complex flocking behavior that has since come to the attention of animal behaviorists. The algorithm used in this simulation has since become a standard for special-effects technicians who wish to show flocks or herds, and has been implemented in several major motion pictures including Batman Returns and The Lion King (Levy, 1992). The successful application of reproducing complex aggregate behavior using these techniques for automobile traffic has already been mentioned on page .

Boekhorst and Hogeweg (1994) was able to reproduce some of the complex interactions of chimpanzee social organization by concentrating on a few simple rules of individual behavior. An important aspect of their model, which could have important implications if applied to recreational boat traffic as described at the beginning of this chapter, is that self-organization in the simple search for food and mates recreated the complex behavior and required none of the usual sociobiological assumptions of such models.

There is a large body of literature about finding static paths through landscapes which will not be reviewed here. Some are well-established, such as the iterative approach described in the GRASS programmers manual (Shapiro et al., 1993). Similar algorithms are available in most of the major raster-based GIS packages which are both robust and reliable. Other research that has gone into finding optimum paths across terrains has involved finding alternate routes (Lombard and Church, 1993) and applications of genetic algorithms to the problem (Michalewicz, 1992).

The problem of a single ``semi-intelligent'' agent navigating itself through a gridded landscape is presented as a textbook exercise in Pinson and Wiener (1991), in an application called Swamp Runner. A description of several such agents navigating a simplified environment is described in the RARS simulation on page . Studies that involve such agents navigating a gridded environment represented by a GIS data set are Westervelt and Hopkins (1996) using GIS in a complex adaptive environment, and Briggs et al. (1996) integrating GRASS GIS and the STELLA simulation environment. Both are models simulating movement of tortoises around the desert.