Living control systems?

June 1, 2007
Self-reconfiguring and assembling devices may have some surprising unintended consequences. We’re talking the potential for almost unlimited versatility that could change the international balance of trade.

By Jeffrey R. Harrow, Principal Technologist, The Harrow Group

Traditionally, one common way to tell if something is alive is to see if it has the ability to reproduce. It can also, through mutation, improve its lot. But we may have to rethink those definitions as robotic devices take on similar capabilities. Just imagine the implications for future process control system design if the system components can replicate and alter themselves to the task immediately at hand. Science fiction, of course.

Or is it…?

Let’s explore some early work that may place some fascinating new tools into our design toolkit. 

According to the May 2005 issue of National Geographic, Cornell University scientists have created a set of two-inch robotic cubes that each contains a microprocessor (pre-programmed with “building knowledge,” a motor, sensors that note “contact and release” events, as well as the order in which the cubes are assembled, plus several electromagnets. According to the article, “The magnets selectively weaken and strengthen connections between other cubes, thereby determining where the structure breaks and joins.”


These independent cubes, when initially joined into a three-cube “being” and provided with extra cubes at “feeding stations,” clone another three-cube “being” in just under one minute. (See figure, right.)

Although these laboratory beasties are currently limited to making more of themselves, imagine the possibilities once some cubes have cameras and lights, claws, various sensors or scrubbing brushes, etc., to let them reassemble themselves for say, inspecting, cleaning or repairing the interior of reactors from the inside. We’re talking the potential for almost unlimited versatility.

The article goes on to say, “If a new, unforeseen task emerges, a robot might construct a new, more suitable robot from scratch, and then the new robot will dismantle the old robot.”

Inanimate objects that reconfigure themselves are not new. Various crystals placed in an appropriate solution automatically replicate perfect copies of themselves. But these new cubic robots, and especially their potential future macro- and nanospawn, could take this to completely unexplored levels.)

On a Smaller Scale
Perhaps the most interesting aspect is how mature, nano-sized versions of these robots might provide solutions where it isn’t feasible to have replacement parts or every possible tool at hand. Imagine if such self-configurable devices could be placed in toxic systems or on unmanned spacecraft, along with a supply of raw “feedstock” or simply with the ability to cannibalize currently unneeded devices). They might be able to form any suddenly needed tool or replacement part.

Take this another step and imagine such devices wandering throughout your body (or any system) sensing physical or chemical problems or even identifying cancer cells via protein matching and then rendering them ineffective.

Feeling Queasy?
Even if you’re a bit uneasy with the idea of artificial robots using you for a swimming pool, other technologies might come to your rescue. The May 4, 2007, issue of New Scientist reports that researchers at France’s Ecole Supérieure in Paris have added an artificial tail composed of magnetic beads held in shape by special strands of DNA, to red blood cells that allow them to be directed at will! An external magnetic field causes the cells to move where they’re told by waving their tails back and forth.

Fascinating, but what could it be good for? How about loading a cell with a medication and swimming it to just the right place in the body before releasing its payload? Analogous techniques can apply to non-biological systems as well.

The Good, the Bad and the Unexpected
But let’s not get complacent. Like most new technological developments, these nanosystems offer both great potential and great danger.

Consider malfunctioning nanomachines that might clog things up. Or nano-assemblers that run amok. Or Bill Joy’s rogue nano-assemblers (which are also disassemblers) that might attack their environment, leaving a disassembled “grey goo” in their path.

Such devices may have some surprising unintended consequences, such as one described by Mike Treder, Executive Director for The Center for Responsible Nanotechnology—a threat to the global economy and stability. Self-assembly could change the international balance of trade.

If nations can build or reconfigure what they need through inexpensive molecular self-assembly from commonly available resources they may become less dependent on the international trade that currently helps to provide global stability.

How Does This Affect Me?
Which brings us full circle back to process control. It wasn’t too long ago that process control elements, such as flow and level sensors driven by laser, ultrasonic and other technologies, seemed as far-out as self-replicating and self-configuring components.

Each of us probably knows folks who were reluctant to embrace new process control devices—and who then found their solutions at a considerable competitive disadvantage. As self-replicating and configuring devices join the panoply of choices available for future process control projects, it will again be those who embrace the new technologies that will likely prosper. Which side will you be on?

  About the Author
Jeffrey R. Harrowis Principal Technologist at The Harrow Group. He can be reached at[email protected].