Update: There is now a new version of the rovers app which includes a plan editor and better simulator.
Below is a (buggy) simulation of Steels' martian rover experiment (without obstacles) which demonstrates that the subsumption architecture is a good way of quickly generating pretty looking simulations. The reset condition is when the simulation crashes, says a lot really....
Each of the rovers follows a simple set of rules as follows:
| # | Condition | Action |
|---|---|---|
| 1 | holding sample, at base | drop sample |
| 2 | holding sample, not at base | drop radioactive grain,move toward beacon |
| 3 | not holding sample, detect sample on ground | Pick up sample,move toward beacon |
| 4 | not holding sample, detect radioactive grain on ground | Pick up grain, move away from beacon |
| 5 | not holding sample, don't detect radioactive grain | Move in random direction |
The conditions are evaluated in order and if a condition is met then the given action is applied.