Swimming Snakebot

Swimming Snakebot

TechEBlog just released their list of the 10 coolest robots. The one in the video below really caught my eye.


(ACM-R5 website)

I am plenty partial to these serpentine-type robots; I spent a good portion of my life (as in a third)developing autonomous robots very similar to the radio-controlled one in the video above. The video shows the robot slithering around sinusoidally a little bit on a flat surface and then swimming like a snake does, but this class of robots is capable of a lot of other types of movement: from the anchor/bend/anchor/extend movement of an inchworm, to the coil/spring movement of a striking cobra, to the vertical climb of a boa constrictor, the sideways motion of a sidewinder, or even the "hand over hand" movement of the Canadarm II on the international space station. All these modes of motion mean that a serpentine robot can handle a very wide variety of terrains, far more than a wheeled or even legged robot can, from big rocks and steep crater walls to a truss in a weightless environment to underwater.

Serpentine robots have other advantages besides their many modes of motion. The joints can be sealed off from their operating environment (as is the case with the robot in the video), thus keeping dust and other debris from interfering with the mechanical motion. They have robust failure modes: complete failure of one or even several joints merely leads to slightly stiffer movement - by comparison, a wheeled robot that loses the operation of a wheel faces huge difficulty in movement. The snakebot body segments are identical, so they can be mass-produced on an assembly line. And, such robots can fit into much smaller spaces than other types of robots.

These advantages and versatility mean that serpentine robots can be used for a wide variety of missions, from search operations in collapsed buildings or mines, to exploration of other planets, to work outside a space station. Indeed, serpentine robots have been used in space for decades, in the form of the Canadarm in each space shuttle; any serpentine robot that has one end in a fixed position is simply a robotic arm.

The robots that I designed had something the robot in the video does not: grippers at each end. These grippers could grab and use specially-designed tools to carry out a wide variety of functions, such as welding or drilling or sampling or spectroscopy or damn near anything. The grippers I came up with are also capable of grabbing onto each other, so that two or more snakebots could connect to each other end-to-end to form one long serpentine robot, capable of climbing out of even very deep craters. (These grippers are illustrated on the lunar robot in my video.)

Because these robots are so flexible, they can be coiled up, so lots of them could be packed together in a small volume; a dozen or more robots the size of the one in the video could be packed into the same lander that carried just one Mars Exploration Rover. With that kind of redundancy, it doesn't matter if one or several of the robots sent on a mission are destroyed.

All of this makes snakebots a wonderful choice for future space missions. I'd love to see a mission to an asteroid carrying a dozen snakebots, all working together to explore and begin setting up industrial operations.

Finally, the serpentine design is scalable down to the nanotech scale. The same swimming motion that is demonstrated in the video above would also work for a nanorobot a few thousand atoms long, swimming along in the bloodstream. Try doing that with a wheeled robot!

Technorati Tags: Robots, Space, Technology, Nanotechnology