Problems involving path planning, exploration, and related activities abound in the world of computing. This paper develops a novel model for realizing ant-inspired algorithms that coordinate robots within fixed, geographically constrained environments ("factory/laboratory/warehouse floors") and studies the model via sample tasks that illustrate various aspects of path planning and exploration. The model, dubbed a Cellular ANTomaton (C-ANTomaton, for short), inverts the relationship between ant-robots and the environments that they navigate: "Intelligence" now resides in the "floor" rather than in the ant-robots. The C-ANTomaton model is illustrated via three proof-of-concept problems: (1) Parking requires ant-robots to congregate in their nearest corners of the "floor," in a maximally compact formation. (2) Food-seeking (both with and without impenetrable obstacles) requires each ant-robot to find its own "food" item, until either "foodless" ant-robots or unclaimed "food" items run out. (3) Maze-threading requires a single ant-robot to find the unique exit to a maze. "Unintelligent" C-ANTomaton-based robots accomplish all of these goals provably more efficiently than traditional "intelligent" ant-robots can; indeed, "intelligent" ant-robots cannot always park at all! All of the presented algorithms are scalable, in that they provably work for any number of ant-robots, within any finite-size "floor." In other words, the computers that collectively supply the system with "intelligence" never exploit any information about either the number of ant-robots or the size of the "floor."
