Preliminary results

Figure 10: First run. All agents have the same destination, and at this stage, the same route.

Figure 11: After 50 iterations

Figure 12: This is a picture of our 3-dimensional visualizer, showing the same view as figure 1.

A small proof-of-principle run is documented in Fig. 10 and 11. It is assumed that all agents leave in the morning from the hotel and hike to the same mountain peak. They however want to avoid each other because they want to hike in solitude. Fig. 10 shows the first run, where no hiker knew about the other hikers' intentions. Fig. 11 shows the situation after 50 iterations, where hikers have learned to spread out and avoid each other.

Fig. 12 shows a situation similar to Fig. 11, but in the 3D viewer. The hikers are visible as dark red figures in the background - to make them visible at all, they are drawn at ten times their natural size. In addition, Fig. 12 can be used to obtain an impression of the quality of the computer rendering of the scene, since it corresponds to the scene phographed in Fig. 1.

Considerably more work will be necessary to fill these elements with true real-world meaning. Progress will be reported in future papers.