Robots can be programmed to respond to changes in sunlight, temperature, or even the levels of certain particles in the air. But, what about a robot that drives around controlled entirely by bacteria?
Bacteria-controlled robots aren't here just yet, but a team of intrepid scientists has created a mathematical model to predict how such systems might explore the world around them.
Of course, bacteria are not endowed with appendages suitable for tugging wires and pulling levers. Instead, the hypothetical robot was designed to monitor what color the bacteria were. The bacteria had a choice of green or red, and they switched between the two depending on what they consumed. The theoretical robot peered at them with a miniature microscope and measured the pigment and intensity of each color, which determined where and how fast the robot moved.
This may seem fairly simplistic to start with, but even the basic movements of the robot became more complicated when new bacteria were introduced into the "robot microbiome." The hypothetical robot moved toward a food (fuel) source in a predatory way, involving a sequence known as "stalk, pause, strike." This behavior is seen in predatory animals when they creep up on their prey, halt before going for the attack and subdue their prey if successful.
Warren Ruder, from Virginia Tech, who developed the model, described his reaction to the stalk, pause, strike behavior to IFLScience: "The microbiome can affect animal behavior with a few simple interactions, so I didn't find it particularly alarming. Instead, it was exciting to see that a fundamental set of interactions were enough to cause a complicated behaviour."
The models have already shown fascinating results: the next steps are to start building some real-world robots, Ruder told IFLScience. "We are actively building all of the components that we've actually simulated." Ruder plans on building prototypes that will read E. coli expression levels with miniature fluorescent microscopes. The bacteria will be engineered in his lab.
This approach to robotics could have a wealth of innovative applications in the fields of agriculture, healthcare and environmental cleanup. These fields all rely heavily on the relationships between bacteria and their hosts. Ruder told IFLScience "One area that I think it will be useful in is the area of biocontainment."
He added that they will be building safety nets into the robots' design. "We're taking a weak strain of bacteria which could be modified to do other things but would remain dependent on the robotic life support system that encompasses it. As a result, there's potentially loss of risk of organisms escaping and infiltrating the environment."
Ruder also points out that his experiments can be done with minimal funding. The price opens up this sort of research to a wide range of experimenters, although only people with access to a laboratory will be able engineer the bacteria.
Video with Warren Ruder explaining how his mathematical model works.