Maze solvers demystified and some other thoughts

TL;DR

This paper analyzes physical maze-solving prototypes, revealing they primarily rely on flow currents or chemical gradients, and discusses their limitations through experiments with slime mould, plant roots, and leeches.

Contribution

It clarifies the underlying mechanisms of physical maze solvers and highlights their limitations through experimental evidence.

Findings

Maze solving often relies on flow currents or chemical gradients.

Physical prototypes trace least resistance or steepest gradient paths.

Experiments show limitations of chemical diffusion-based maze solving.

Abstract

There is a growing interest towards implementation of maze solving in spatially-extended physical, chemical and living systems. Several reports of prototypes attracted great publicity, e.g. maze solving with slime mould and epithelial cells, maze navigating droplets. We show that most prototypes utilise one of two phenomena: a shortest path in a maze is a path of the least resistance for fluid and current flow, and a shortest path is a path of the steepest gradient of chemoattractants. We discuss that substrates with so-called maze-solving capabilities simply trace flow currents or chemical diffusion gradients. We illustrate our thoughts with a model of flow and experiments with slime mould. The chapter ends with a discussion of experiments on maze solving with plant roots and leeches which show limitations of the chemical diffusion maze-solving approach.