Engineers from Drexel University have devised a bacteria-powered microrobot that can be steered through fluids with applied electric fields. Imagine a tiny, tiny robotic system covered over by even tinier biomolecular arms (flagella) working cooperatively as a "bacterial surface" to move the biobot from place to place, and then, just as importantly, imagine an algorithm that can steer the thing. The group's work is described in the current IEEE Transactions on Robotics.
The bacteria-powered microrobot (BPM) idea is in itself not new. Researchers are chasing after the technology from many different angles, while biomedical engineers from the same Drexel lab had demonstrated an earlier version of this same concept in 2014, albeit in a much simpler environment.
The idea behind the bot is intuitive enough, at least in theory. Take an inorganic substrate—in this case an epoxy known as SU-8 that's useful for building tiny machine components—and blot it with bacteria swarming on an agar plate. The bacteria adhere naturally the the SU-8 and that's basically it.
The motion comes from the twisting swimming motions of the bacteria's wispy appendages, or, in the paper's words: "The hydrodynamic force is generated by the helical motion created by the bundling of many rotating flagella." This is how bacteria get around generally, via a collection of tiny biological propellers. It's pretty neat.
To harness this bacterial carpet of propellers, the researchers take advantage of the innate negative charge of gram negative bacteria. In electrically neutral surroundings, the bacteria-bot just kind of swims around in whatever direction it starts in, but apply some charge and it becomes possible to (very carefully) steer it.
Here's a much earlier version of the idea, from the same lab:
This steering is the tricky thing, and this is where the current paper makes its advance: a control algorithm. In particular, the Drexel researchers present an algorithm that can optimally avoid obstacles. The challenge is that there are several factors trying to get the bot to go this or that direction in addition to electric charge. For one thing, the assembled bacteria give the bot a natural or preferred direction. Drop a motorboat into some water and it's going to move in whatever direction it's pointed, which will be some function of its hull shape, propeller orientation, and also the current of the water (analogous to the applied charge). The boat's gonna be fairly predictable, but a slime of bacteria on a microscale robot, a bit less so. What's more, the fluid mechanics of the BPM are such that it can change direction almost instantaneously.
"The problem is ultimately an optimization problem to find the velocity that allows the maximum amount of control using electric charge given all of these factors," the authors write. "Through several experiments, our proposed method succeeded in translating the BPM to its goal position while avoiding static obstacles. In terms of the control input, our algorithm steadily maintained the maximum magnitude of input, which was 20 V in our system. This enables the BPM to steer with high velocity due to a strong power voltage."
Yep: microbial drones, at your service.