Officially known as Physarum polycephalus, Slime mold is not a plant, animal or fungus, but a single-celled organism older than dinosaurs. In search of food, it extends tentacle-like projections in several directions at once. It then redoubles its efforts on the most efficient routes leading to food, abandoning less productive routes. This process creates optimized networks that balance efficiency and resilience—a sought-after quality in transportation and infrastructure systems.
The organism's ability to find the shortest path between multiple points while maintaining redundant connections has made it a favorite among researchers studying network design. Most famously, in 2010, researchers from Hokkaido University reported the results of an experiment in which they dropped a drop of slime mold onto a detailed map of Tokyo's train system, marking the main stations with oatmeal. First, the mindless organism absorbed the entire map. A few days later, he cut himself down, leaving behind only the most effective paths. The result closely matched Tokyo's real-life railway network.
Since then, researchers around the world have used slime mold to solve mazes and even map the dark matter that holds the universe together. Experts from Mexico, Britain and the Iberian Peninsula have tasked the body with redesigning its roads, although few of these experiments have resulted in real improvements.
Historically, researchers working with an organism would print out a physical map and add slime mold to it. But Kay believes Mirete's approach, which replicates the construction of slime mold pathways without using real organisms, could help solve more complex problems. Slime molds are visible to the naked eye, so Kay's team studied the behavior of the clumps in the laboratory, focusing on the key behaviors that allow these organisms to create efficient networks so well. They then translated this behavior into a set of rules that became an algorithm.
Some experts are not convinced. According to Jeff Boeing, assistant professor of urban planning and spatial analysis at the University of Southern California, such algorithms don't address “the messy realities of walking into a room with a group of stakeholders and sharing a vision of the future for your community.” The problems of modern urban planning, he says, are not purely technical problems: “It’s not that we don’t know how to make infrastructure networks efficient, resilient and connected – it’s that doing so is politically difficult.”
