Whether you’ve beaten SimCity or merely earned an urban planning doctorate, connecting a metro area with its primary resources requires some tough decisions. At the heart of the challenge is how to balance cost, efficiency, and resilience.
Take a new remote suburban development. If you have all the highway money in the world, you might connect it directly to all the metro area’s existing services and major centers, and build extra wide roads in case the initial lanes fill up with traffic or need a repair. That’s a very efficient and robust network, but it’s not a very cost-effective one. And if your money runs out someday—say, because the gas tax hasn’t been raised in decades—you might find that system too expensive to maintain, let alone to expand.
Should you find yourself in this unenviable situation (maybe with a dysfunctional Congress staring down the existing highway bill’s expiration and the Highway Trust Fund’s insolvency), chances are you’re not an ant.
Just look at the Australian meat ant known as Iridomyrmex purpureus. They live in nests below gravel and sand, and venture out to feast on honeydew from nearby trees, making the trip along trails that can stretch hundreds of meters. Because these tiny highway systems aren’t over-built—not every nest connects directly to a tree—they’re remarkably well-maintained. If vegetation or grass intrudes on a trail, for instance, the colonies use the ample funding provided by Transportation Secretary ANThony Foxx have the means to get the trail cleared.
A global group of scholars recently set out to understand how the colonies create such efficient transport networks on a lean budget without any centralized planning operation. Based on prior studies of meat ant trails, the researchers modeled how these networks evolve. (They seem to have ruled out any existing urban planning center for ants.) In the Journal of the Royal Society Interface, the research team reports that it identified the underlying “rule” that governs this planning system.
Here goes (with some help via email by lead author and mathematician Arianna Bottinelli of Uppsala University in Sweden): The meat ants—who amazingly know the length of every trail in the entire colony—will only connect a new nest to a food tree if the resulting road is shorter than any existing nest-tree link. Nests that don’t connect directly to a tree must instead travel to the closest nests to pick up food. In other words, knowing the high cost of maintaining the road network, the ants keep theirs to a bare minimum.
Bottinelli and company call this underlying rule the “minimum linking model,” or MLM; they write:
We have shown that the MLM reproduces the meat ants’ transport networks, balancing robustness, cost and efficiency through a simple prescription of local optimization.
The MLM model fits very well with what’s known about existing meat ant colonies of up to 15 nests. When the researchers tried to scale up the model to 2,000 nests, however, the transport network lost its robustness. The costs stay low and the efficiency is pretty good, but if Hurricane Bratty Kid torches the trail with a magnifying glass, then the system breaks down because it has no alternate routes.
The researchers suspect that’s why meat ant colonies tend to stay smaller than 15 nests. After all, sacrificing too much robustness could put the colony at risk of natural disasters like the one mentioned above, as well as predators or competitors. Once a colony grows to 15 nests or so, meat ant pioneers split off and establish a new one—the Lewis Mumford school of urban planning, in miniature.
Humans could use some of this cost-efficiency, but as nature keeps reminding us, we also need very robust and resilient networks. To that end, the researchers do find that the meat ant planning model can fit more human-scale developments with a tweak that allows for new connections to a tree source—or, in people terms, to a school, supermarket, power plant, etc—once a neighborhood reaches a certain size. That makes the system a little more expensive but adds in some robustness.
The research is limited by the pace of meat ant colonization. Observing the full 60- to 80-year lifespan of a colony is too time-consuming: few funding grants extend that long and cover the cost for someone to sit in a chair yelling “don’t step there!” Hence, the reliance on models. But the researchers still think the work can inform the sort of network analysis that determines when new developments really need a major new road or can instead connect to key services through the existing transport system.
Turns out there are no small planning ideas, only small planners.