There are a lot of similarities between the popular home robot and the forest-floor goop you wish you hadn’t stepped in. Following a combination of programmed cues, an iRobot Roomba responds to its environment, detecting dirt and vacuuming it up. Slime mold, or Physarum polycephalum, functions in a similar way, behaving in accordance with a list of genetically inherited rules. It’s a simple, single-celled organism that digests detritus and leaves behind a trail of slime wherever it goes. What makes slime mold remarkable is that when threatened, all the individual cells in the area combine, creating a larger organism—a collective entity capable of solving problems that a Roomba would find challenging.
So, is slime mold smarter than a Roomba? We’ll lay out the arguments. You be the judge.
Slime mold has a memory for its external environment that rivals those of some robots, researchers at the University of Sydney found this summer. As it oozes around, Physarum consumes food, so covering the same ground twice would be a waste of energy. By reading the chemical cues in its own slime, or slime left by other slime mold, it can avoid areas that have already been depleted of tasty rotten leaves. This skill is comparable to what’s known as externalized spatial memory in a robot.
Slime mold can design a transportation network. In two different experiments, it re-created massive transit systems in miniature. Researchers set out tasty treats to represent population centers, and slime mold spread out to find them, duplicating the Tokyo rail system and the Canadian highway system.
Slime mold doesn’t have a brain, but it can sense and react to its surroundings without one.
Slime mold has been inspiring computer scientists since the 1960s. Some of the math underlying the workings of video games, the Internet, and robots stems from studies of slime mold.
Slime mold evolves.
Slime mold can control a robot. In 2005, Japanese researchers engineered an interface that let a slime cell control a hexapod robot.
Slime mold makes things slimy.
Roomba has no memory of what your living room looks like. Every time it starts up, it’s starting over. Of course, this is exactly what it was designed to do. It would be possible to build a Roomba that remembers, but that would require more internal memory and processing power.
Roomba can’t compete at urban planning, but it can identify particularly gritty spots on a floor to do some focused cleaning.
A Roomba’s brain is made up of an ARM processor core and thousands of lines of software code. The CPU is coupled to infrared, piezoelectric, and RF sensors.
Roomba helped inspire the creation of a household robotics industry worth hundreds of millions of dollars.
Roomba gets updated, with a new generation of robots arriving every year or so.
Roombas control themselves.
Roomba makes things sparkling clean.
Comments