Build Your Own Robot Swarm — or — Angular Automatons and Cuckoo Clocks

1,024 little robots got together at Harvard, making the letter "K" and drawing a star. What they do doesn't look as sophisticated as many marching band halftime formations — but it's a good start on collective artificial intelligence.

Meanwhile, Harvard and MIT's angular automatons don't do much except fold themselves into crablike shapes, and scuttle away: today.

1. Harvard's Robot Swarm
   • Rubenstein's Kilobots, Collective Artificial Intelligence, and "Resistance is futile"
   • Build Your Own Robot Swarm
2. MIT's Origami Robots
   • Angular Automatons

---

The Digesting Duck that Didn't, Cuckoo Clocks, and Pygmalion

(From Wikipedia and Boston Dynamics, used w/o permission.)

Automatons have come a long way since Jacques de Vaucanson built his Digesting Duck. It didn't actually digest anything, by the way.

Geneva's singing bird boxes did sing, though: and automatons have performed on clocks from the Cathédrale Notre-Dame of Strasbourg's to cuckoo clocks.

More recently, an animatronic dinosaur appeared in "Jurasic Park," Sony's AIBO played soccer, and Boston Dynamics' Petman robot tests hazmat suits.

As I said last week, tales of automatons are ancient. Ovid's Pygmalion is a bit like today's 'mad scientist,' but Ovid's narrative has Aphrodite giving the statue life. All Pygmalion did was carve a life-like statue: and get infatuated.

Interestingly, in Greek stories about Talos that I've found, the robot guard doesn't turn on its master. The 'mockery of life terrorizes villagers' routine is a fairly recent development.

Victor Frankenstein, The Phantom Creeps, and All That

The Golem of Prague story may have roots in folklore: but didn't show up until the 19th century, after publication of Mary Shelley's 1818 novel, "Frankenstein."

The 1931 "Frankenstein" was, I think, a pretty good film adaptation. Later movies, like "Frankenstein Meets the Wolf Man" and "Abbott and Costello Meet Frankenstein," not so much.

Whether films like "The Phantom Creeps" (1939) and "Robot Monster" (1953) shape popular culture, or reflect it, I think movies give us a pretty good look at attitudes and assumptions from their era.

I could indulge in conventional angst about the evils of technology, collapse of civilization, or whatever — but would much rather take a none-too-serious look at artificial intelligence in the movies:

• "2001: A Space Odyssey (1968)
  • Insane computer kills crewmates
• "Colossus: The Forbin Project" (1970)
  • Massive supercomputer is built
  • Takes over the world
  • Somewhat plausible
    • Given a willing suspension of disbelief
• "Westworld" (1974)
  • Killer robots
    • A whole resort full of killer robots
• "Logan's Run" (1976)
  • A nice, neat, orderly society
    • Where life is groovy
      • Until you hit 30
      • Then you die
    • Well, you can't have everything
  • Then a crazed cop kills the master computer
• "Star Wars" (1977)
  • Comedy-relief robots
    • C-3PO, human-cyborg relations
    • R2-D2, astromech droid
• "The Terminator" (1984)
  • Evil computer mastermind
  • Determined killer cyborg
  • Threat of nuclear apocalypse
• "The Matrix" (1999)
  • Humanity makes an artificial intelligence
    • That takes over the world
  (Drifting at the Edge of Time and Space (January 26, 2010))

My guess is that self-aware robots, if we ever manage to develop them, will have more in common with C-3PO than Skynet.

I wasn't the first person to see a humor in technophobic angst. Daniel Wilson had fun with fear in "How to Survive a Robot Uprising," as has XKCD:

"...Here are a few snapshots of what an actual robot apocalypse might look like:

"In labs everywhere, experimental robots would leap up from lab benches in a murderous rage, locate the door, and—with a tremendous crash—plow into it and fall over.

"Those robots lucky enough to have limbs that can operate a doorknob, or to have the door left open for them, would have to contend with deceptively tricky rubber thresholds before they could get into the hallway.

"Hours later, most of them would be found in nearby bathrooms, trying desperately to exterminate what they have identified as a human overlord but is actually a paper towel dispenser...."
("Robot Apocalypse," What If? XKCD.com)

Getting a Grip About Crash Test Dummies

I'm not troubled that we make increasingly lifelike imitations of living creatures. Somehow, I don't think the Almighty is going to be offended by cuckoo clocks or robot dogs playing soccer.

Tightly-wound folks of a grimly pious bent might have qualms about mechanical birds, music boxes, and other frivolities. I'm convinced that gloominess is not next to Godliness, and that's another topic. (May 5, 2011)

Besides, many automata help make this a safer world for humans.

Crash test dummies have replaced volunteers, human cadavers, and animals, for vehicle testing: and medical simulators help folks learn medical procedures and resuscitation techniques without putting patients or volunteers at risk.

Today's crash test dummies and medical simulators like the Harvey mannequin and Resusci Anne don't move on their own, at least not much, but I see them as automatons, at least in some senses of the word.

Oxford Dictionaries sees "automaton" as meaning a robot that mimics an human. Merriam-Webster defines it as a machine that moves by itself, and the Wikipedia page's definition is pretty broad, too. In English, "automaton" and "robot" mean almost the same thing.

It's no wonder that Artificial Intelligence still has trouble understanding human languages, and that's almost another topic.

Being Creative, Being Human

I don't see the human impulse to create as a problem. It'd be surprising if some of us didn't try to create, since we're made "in the image of God." (Genesis 1:26-27; Catechism of the Catholic Church, 1701-1709)

What we do with our creative impulses is where ethics come in. Since we have free will, we can decide to act wisely — or not. The problem is not in our tech, it's in our decisions. (Catechism, 2493-2499, 2501, 2513)

We have freedom, "the power, rooted in reason and will, to act or not to act...." (Catechism, 1731)

Deciding that we'd rather act against reason, truth, and right conscience is an option: and a very bad idea. (Catechism, 311, 396, 1704, 1730, 1739, 1849)

Studying this universe, and developing new technology with what we learn, is part of being human. It's what we're designed to do. (Catechism, 2293-2296)

More of my take on who we are and what we do:

• "Science, Technology, and Being Human"
  (January 27, 2013)
• "Sin, Freedom, and Toy Monkeys"
  (May 23, 2012)
• "Humans are Animals: But Not Just Animals"
  (August 31, 2011)

Automatons: Not Entirely Lifelike

I've never run into 'religious objections' to crash test dummies.

On the other hand, I wouldn't be terribly surprised if someone took Acts 17:28-29 and decided that all moving machines are Satanic: automobiles, bicycles, and the entire Audio-Animatronic cast of Walt Disney's Enchanted Tiki Room. Folks can get — imaginative — when faith and fear collide. My opinion.

I would, however, be astonished if we learned that a crash dummy refused to get into a test vehicle. They're nowhere near that lifelike.

---

1. Harvard's Robot Swarm

(From Reuters, used w/o permission.)
("The Kilobots, a swarm of 1,000 simple but collaborative robots...."
(Reuters))

"Rise of the machines? Tiny robot horde swarms to form shapes"
Will Dunham, Reuters (August 14, 2014)

"They look vaguely like miniature hockey pucks skittering along on three pin-like metal legs, but a swarm of small robots called Kilobots at a laboratory at Harvard University is making a little bit of history for automatons everywhere.

"Researchers who created a battalion of 1,024 of these robots said on Thursday the mini-machines are able to communicate with one another and organize themselves into two-dimensional shapes like letters of the alphabet.

"Much smaller groups of robots have been able to carry out similar tasks, but never a group this size...."

I don't blame Mr. Dunham for putting part of a movie title in his headline: "Terminator 3: Rise of the Machines." Headlines are supposed to grab attention: and that one gets the job done.

I'm impressed that Kilobots can receive instructions to do a particular job, and take care of the rest on their own. It sounds a bit like what assembly languages do with instructions written in comparatively user-friendly languages like BASIC or Vala, after they're run through a compiler, if memory serves, and that's yet another topic.

If the Reuters article got the price correctly, researchers spent $14,336 on parts for their Kilobots: only $14 for each of the bots. That's not bad for a two-inch-high robot: and roughly half the cost of a paperweight that doesn't do much except look cool.

Rubenstein's Kilobots, Collective Artificial Intelligence, and "Resistance is Futile"

(From Reuters, used w/o permission.)
("A combination handout photo ... demonstrates the Kilobots ability, given a two-dimensional image, to follow simple rules to form the same shape."
(Reuters))

"...In a study published in the journal Science, they formed themselves on a large tabletop into the shapes of the letter 'K,' a star, a solid square and a wrench.

"It may be a step forward for collective artificial intelligence, although the researchers acknowledge the Kilobots are not exactly thinking deep thoughts.

" 'This is a "collective" of robots - a group of robots that work together to complete a common goal,' said Harvard computer scientist Michael Rubenstein, who led the study. 'If you call collective artificial intelligence the ability of a "collective" to start to behave as a single entity, you could call this collective artificial intelligence.'..."
(Will Dunham, Reuters)

"Collective artificial intelligence" reminded me of Star Trek's Borg, and speculative fiction's robots.

I'm no Trekkie, but I enjoy watching reruns of the original series now and then: and recently saw "Star Trek: First Contact" again. As entertainment, it's not the worst way to spend nearly two hours.

But — as Mr. Dunham's article points out — Rubenstein's Kilobots are "not exactly thinking deep thoughts," and most certainly are not a threat. I suppose they might, given time, spell out "RESISTANCE IS FUTILE," but that's about it.

Build Your Own Robot Swarm

"...The Kilobot name is a play on the word kilobit, meaning 1,024 bits of digital information. But to some it might sound menacing - as in 'killer robot' - as if it belongs in a movie like 'Terminator 3: Rise of the Machines.'

" 'I tell people that these robots are not very dangerous. The only way that they could hurt you is if you try to eat one. They can't even go over a piece of paper. So they're kind of stuck where they are,' Rubenstein said."
(Will Dunham, Reuters)

"Where they are" is an eight-by-eight-foot surface: basically, a big whiteboard. As I said, the Kilobots' offensive capabilities are pretty much zilch.

Harvard's SoSRG — a nearly-unpronounceable acronym, not nearly as cool a name as as Skynet — isn't as optimistic as the Reuters article, where it comes to costs. The Harvard folks say that it'll take between $20 and $50 dollars per unit to build your own Kilobots: depending on how many you make at a time.

For the convenience of anyone who might want to build an army of Kilobots, SoSRG thoughtfully provided this assistance:

"...If you would like to build your own Kilobots, all the software and hardware details are available under a Creative Commons attribution-NonCommercial-ShareAlike 3.0 Unported (CC BY-NC-SA 3.0) license. The documents can be found here. The design is fairly simple for any lab that is used to getting electronics made, and this is a great and affordable option if you plan to make a large number of robots (We estimate approximately $20/robot for thousand, up to $50/robot for a hundred). If you decide to make your own robots, feel free to contact Mike Rubenstein to let him know and if you need help...."
(Self-organizing Systems Research Group, Harvard)

More:

• Mike Rubenstein
  Harvard University
  • Includes videos of kilobots at work
• The Kilobot Project
  Self-organizing Systems Research Group, Harvard

(From Kilobot User Guide, Beta Release, Self-Organizing Systems Research Group, Harvard University; used w/o permission.)
(Kilobot circuit board diagram.)

---

2. MIT's Origami Robots

(From Harvard's Wyss Institute, via Wired, used w/o permission.)
("The approach represented by these origami-bots leads to designs that are less powerful, but far more flexible and allows designers to experiment and iterate more quickly and cost effectively."
(Harvard's Wyss Institute, via Wired))

"The Deep Design Thinking Behind MIT's Incredible Origami Robot"
Joseph Flaherty, Wired (August 12, 2014)

"Imagine if Dunder-Mifflin made robots and you'd have a mental image of the latest breakthrough in robotic technology.

"A partnership between Harvard's Wyss Institute and MIT, with funding from the National Science Foundation, has yielded a new breed of angular automatons can transform themselves from a 2-D sheet of plastic into a fully functional cyber crab with the aid of a mere coin cell battery...."

Don't bother looking for Dunder Mifflin in the phone directory. It's a fictional paper sales company featured in the United States sitcom "The Office."

A few years from now, we might see novelty greeting cards that fold themselves into little dancing robots: sort of like today's musical and pop-up cards, and that's yet again another topic.

Mr. Flaherty's article compares electron beam melting and 3-D printing to "mundane and time-tested 2-D printers and laser cutters." My guess is that by the time he's my age, electron beam melting and other rapid prototyping technologies will be "mundane," too.

He makes a good point, though.

Tech we use to print kidneys, for example, is still experimental. (May 3, 2013)

But what was "science fiction" tech in my youth, like laser cutters, is the sort of thing that machine shop managers get retail or wholesale: depending on the scale of their operation. They also need to keep up with what's changed since their last purchase.

So do folks who can spend a few thousand dollars on equipment for a hobby:

• "A Technical and Commercial Comparison of Fiber Laser and CO2 Laser Cutting"
  John Powell, Alexander Kaplan; Fiber Lasers, Laser Cutting; Laser Insights; Continuous Wave (September 22, 2013)
• "Review: Full Spectrum Engineering 40W Hobby Laser (Deluxe Model)"
  Kevin N Gunn, (April 25, 2011)

Angular Automatons

(From Harvard's Wyss Institute, via Wired, used w/o permission.)
("A new breed of angular automaton can transform themselves from a 2-D sheet of plastic into a fully functional cyber crab with the aid of a mere coin cell battery."
(Harvard's Wyss Institute, via Wired))

"...Instead of building 'bots by assembling a metal frame, discrete motors, sensors, and power sources like some kind of mechanical Frankenstein, these researchers are promoting a new kind of manufacturing where engineers can elegantly specify a design and watch it spring to life like a seed emerging from the ground.

"This manufacturing method could be critically important in extreme environments like space. Astronauts on a mission to Mars will need robotic helpers, but the proposition of packing them in bulk quantities or bringing bulky manufacturing rigs and explosive gases for acetylene torches to manufacture them on site isn't feasible. Likewise, search and rescue robots are becoming more common, but their bulk and cost make them slow to deploy...."
(Joseph Flaherty, Wired)

There's a difference between tech that's available — and tech that's readily available.

High-end 3-D fast prototyping is something that a few companies and individuals are good at, not something you expect to find at the back of a department store.

But most offices and many homes have good- to high-quality 2-D printers, and if you don't have laser cutting equipment, you can get it at Amazon.com: search for laser cutting machines in their Business, Industrial & Scientific Supplies department. You still need to know what you're doing, of course, and have the right software.

This 'origami robot' tech's portability and low cost should make in very useful for disaster response.

'Ready for printing' sheets, attachments like servos and sensors, a 2-D printer, more-or-less-automated machining equipment, and a power source, should fit into a small truck or cargo container.

Airlifted into a disaster area, a portable 'robot factory' could churn out rescue robots made to order for whatever search crews encounter.

A downside to 'origami robots' is that they'd be difficult or impossible to repair. A flaw anywhere in the robot makes the whole unit unusable. Since it's basically a single flat sheet, swapping out the defective part isn't an option: the whole robot has to be replaced.

That may not be a serious disadvantage, though, as long as there's a 'robot factory' available. The things might not be all that expensive or difficult to replace. With a bit of planning, recycling this sort of robot should be fairly straightforward: still more topics.

More about self-folding robots:

• "Watch this Origami-inspired Robot Fold Up and Walk Away"
  Spectrum, MIT (August 13th, 2014)
• " Origami robot folds itself up, crawls away"
  "Prototype made almost entirely of printable parts demonstrates crucial capabilities of reconfigurable robots."
  Larry Hardesty, MIT News Office (August 7, 2014)
• "Rus, Demaine help create first self-folding, fully-operational origami robot"
  News, MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) (August 7, 2014)

This isn't the world I grew up in, and I'm okay with that:

• "Neurosynaptic Cores and Retinal Implants: Getting a Grip About Tech"
  (August 15, 2014)
  Particularly
  • IBM's TrueNorth Chip and Cockroach Brains
• "The 'Fast Web' Impairs Thinking — or — Fear and Phaedrus Ride Again"
  (February 20, 2014)
• "Spaceships, Robots, and Being Catholic"
  (April 12, 2013)
• "Prayer, Technology, and Looking Ahead"
  (February 24, 2013)
  Particularly
  • Looking Forward, Not Back
  • Being Human, Being Catholic
• "HAL 9000, Skynet, and C-3PO"
  Drifting at the Edge of Time and Space (January 26, 2010)