Click to open a first-person viewpoint window of a randomly chosen target.
A group chase and escape system consists of two kinds of agents, chasers and targets. The chasers (red) and targets (green) play tag, that is, a chaser will chase a target and vice versa.
This model is proposed and investigated by Kamimura (Univ. of Tokyo) and Ohira (Sony CSL), and published as [A. Kamimura and T. Ohira, Group Chase and Escape, New Journal of Physics, DOI:10.1088/1367-2630]. The above flash is the simulator of this system.
The above figure denotes the hopping rule of chasers and targets. The system is the square lattice with the periodic boundary condition. A chaser will chase the closest target, and a target try to back away from the closest chaser. Each target always tries to reduce a distance (a number of hopping steps) to the closest target. If there is more than one choice to move, one of them is chosen randomly.
If a target is cought by a chaser, then it it eaten, that is, it is removed from the system. Therefore, targets will be eventually vanished. Then how long will it take? How does the efficiency of predation depend on the density of targets? Are there any self-organized spatial structures formed by agents? If you get intereted in this model, see the article by Kamimura and Ohira.