At long last, the robotic revolution is upon us. Any day now we'll have robot cars driving our roads, robot bees pollinating our flowers, robot plants exploring our soil—wait, what? The lonely teenage years I spent reading Asimov short stories had me fully prepared to see robots one day wading into the seas of Jupiter, but plants, with their signature boring motionlessness, hardly seemed a likely source for new, exciting advances in robotics. As it turns out, however, the sessile nature of plants has an unanticipated consequence: When plants do move, it tends to be in an unusually graceful and efficient manner. The result is that, in some cases, plants are pointing the way toward far more elegant solutions to robotics problems than researchers in the field could have otherwise come up with.
Although their habits are distinctly sedentary when compared to those of most animals, a close study of plants illuminates movements that are at once subtle and astonishing. Some of the most striking examples can be found in carnivorous plants, bladderworts and Venus flytraps and the like, whom evolution has provided with the ability to sense and react swiftly to the presence of the small animals, which they trap and digest inside themselves. That the effect is produced mechanically, without the aid of anything so crude as a central nervous system, only makes it more amazing when we learn that flytraps, for instance, can snap shut in less than a second when a fly brushes the tiny hairs the plant deploys to sense a snack. Other notable cases of rapid movement include plants that can eject pollen or react to touch with comparable swiftness, but all of these are outliers—the overwhelming majority of plant activities occur more slowly. They shift gradually to face the sun, lazily unfurl their petals, and grow shoots upward or roots downward.
Whether fast or slow, plant motion generally happens in reaction to specific stimuli—light, warmth, touch, and moisture being the most common examples. Already, osmotic actuators based on the mechanism by which plants use osmotic pressure to drive movement have been developed as a means of supplying power to bio-inspired robots. Another potentially fruitful sources of inspiration include the decentralized, emergent behavior plants exhibit, which is so unlike the centrally orchestrated activity we're used to finding in the animal kingdom and could allow for greater flexibility and resilience in robots that incorporate this sort of behavior into their design.
So far, the most remarkable model for movement in plants may well be the growth patterns of their roots, which can creep through soil with astounding efficiency, avoiding obstacles and seeking water and nutrients in a coordinated manner reminiscent of the active seeking movements of animals. Although we don't usually think of growth as a way of moving through a space, of course it is one. In the case of roots, the motion is accomplished by adding new material to the farthest end of the root, while all the rest of it remains in place and serves it as an anchor. This means that the friction at most points along the structure is reduced to nothing, and is concentrated only at the tip, which is where the growth proceeds from. It’s this reduction in friction that makes growth so efficient when compared to the more standard strategies for movement, in which the entire object must be shifted from one position to another.
Robotic applications based on the close observation of root systems have already begun to be put into practice, at the European Center for Micro-BioRobotics at the Istituto Italiano di Tecnologia. Coordinator Dr. Barbara Mazzolai and her team have developed PLANTOID, a robot prototype that moves through soil by “growing”— anchoring itself in soil and adding new material (which is spooled out from above) onto its furthest end. In this article for the journal Frontiers in Bioengeneering and Biotechnology, Mazzolai describes the vast increases in efficiency provided by the growth-mimicking movement used by PLANTOID (called elongation from the tip or EFT) compared with other soil exploration strategies. (She also provides an overview of the ways in which plant models may become useful in robotics, which I drew on substantially for this article.) Their next step will be to combine the novel, growth-mimicking motion with sensors to allow the robot to detect things like temperature, humidity, pH, salinity, phosphates, and nitrates in the soil. These sensors will help direct the robot's activity beneath the earth, as well as providing information about the composition of the soil around it to its handlers. Once perfected, bots that employed the root growth strategy could have applications in industries which rely on soil testing such as agriculture or mining, and could also be used to seek out and monitor contaminants in places where pollution has occurred or is suspected..
Wherever a robot is called upon to sense and respond to its environment, the manner of plants' movement may point towards unusual or unusually elegant solutions. Only time will tell if these PLANTOIDs will be followed by all manner of other robotic creations modeled after the vegetation around us. Who knows, perhaps we'll have robotic rocks as well—or robots that make up the soil itself. And lo, one day, a robot sun will gently sink behind emergent robot clouds, as weary robot bees return from pollinating robot trees.