Showing posts with label artificial intelligence. Show all posts
Showing posts with label artificial intelligence. Show all posts

Friday, October 24, 2014

3D Sensor: Counting Doughnuts, Reading Tires, and a Rambling Lemming

"DS1000 3D Sensor for the Food and Beverage Industries"

CognexTV, YouTube (June 7, 2013) video, 1:45

"DS1000 3D Sensor for the Food and Beverage Industries"

"3D laser profiling food and beverage manufacturing applications include: identifying improperly filled packages by verifying object heights, detecting defects like skewed caps, and measuring volume and dimensions to verify portion size, to name a few."

The Lemming spent most of yesterday listening to folks talk about assorted gizmos, gadgets, and thingamajigs: and industrial robots. Rethink Robotic's Baxter was there, too, but that's not what's on the Lemming's mind today. Not so much, anyway.

For the Lemming, spending a day looking at actuators, valves, and industrial control components; and listening to folks talking about ROI, HMI, and OCR, is a nice break in routine. Your experience might vary.

Cognex has two more videos, showing why that yellow and black box is a good thing to have in the automotive and electronics industry, too. One of its not-so-obvious tricks is reading embossed lettering or numbers: like DOT codes on car tires, using Optical Character Recognition (OCR).

That 'food and beverage industry' video gives a quick run-through of how the DS1000 3D Sensor can 'see' if there's the right number of doughnuts in a try, and measure the volume of portions. There's more, but the Lemming figures you've probably got time to look at a one-and-three-quarters-minute video: no voice, but they've got some pretty good music.

The Lemming likes it, anyway.

Remembering the Will-be that Was


HAL "Look Dave, I can see you're really upset about this...."
("2001: A Space Odyssey," via imdb.com)

Back in the Lemming's 'good old days,' computers and robots were common: in science fiction stories. Somehow the evil scientists never learned that robot minions generally turned on their masters, and that's another topic.

Now that the Lemming is living in 'the future,' flying cars have been invented several times, and still haven't caught on.

Robots are showing up in more places every year, but so far they've shown a remarkable lack of ambition: compared to their fictional counterparts, that is. Which, in the Lemming's opinion, is just as well.

Vaguely-related posts:

Friday, April 13, 2012

Turing Tests, Subcognitive Low-Level Association, and Perambulators

Today is Friday the 13th, which has very little to do with artificial intelligence, Turing tests, or the number 11:

Computers That Think: Just Around the Corner (Again)

"Artificial Intelligence Could Be on Brink of Passing Turing Test"
Brandon Keim, Wired Science (April 12, 2012)

"One hundred years after Alan Turing was born, his eponymous test remains an elusive benchmark for artificial intelligence. Now, for the first time in decades, it's possible to imagine a machine making the grade.

"Turing was one of the 20th century's great mathematicians, a conceptual architect of modern computing whose codebreaking played a decisive part in World War II. His test, described in a seminal dawn-of-the-computer-age paper, was deceptively simple: If a machine could pass for human in conversation, the machine could be considered intelligent.

"Artificial intelligences are now ubiquitous, from GPS navigation systems and Google algorithms to automated customer service and Apple's Siri, to say nothing of Deep Blue and Watson - but no machine has met Turing's standard. The quest to do so, however, and the lines of research inspired by the general challenge of modeling human thought, have profoundly influenced both computer and cognitive science.

"There is reason to believe that code kernels for the first Turing-intelligent machine have already been written...."

Artificial intelligence, the sort exhibited by C3PO and HAL 9000, has been 'just around the corner' for decades. So far, what's been achieved is the release of several movies; and some robots. The robots are, sometimes, useful. But "intelligent?" Not so much.Part of that may have to do with how computers and the human brain are designed. Not the hardware: the functional parameters of the systems.
  • Computers
    • Get correct answers
      • Based on vast quantities of data
        • All of which is precisely correct
  • Human brains
    • Get correct answers
      • Based on very little data
        • Most of which is wrong
There's a bit of Murphy's Law in that joke. Besides, human beings process large amounts of data on the fly, without being aware of the process: as folks who started designing robots that can sense their environment discovered.

The point is that so far, artificial intelligence has been very good at doing rapid calculations that involve accurate data. Humans have survived because they're pretty good at sorting out relevant information from the deluge of data fragments pouring into the brain.

Back to that Wired article.

Subcognitive Low-Level Association, Dental Hygiene, and Perambulators

"...'Two revolutionary advances in information technology may bring the Turing test out of retirement,' wrote Robert French, a cognitive scientist at the French National Center for Scientific Research, in an Apr. 12 Science essay. 'The first is the ready availability of vast amounts of raw data - from video feeds to complete sound environments, and from casual conversations to technical documents on every conceivable subject. The second is the advent of sophisticated techniques for collecting, organizing, and processing this rich collection of data.'

" 'Is it possible to recreate something similar to the subcognitive low-level association network that we have? That's experiencing largely what we're experiencing? Would that be so impossible?' French said...."

"...The human mind was thought to be logical. Computers run logical commands. Therefore our brains should be computable. Computer scientists thought that within a decade, maybe two, a person engaged in dialogue with two hidden conversants, one computer and one human, would be unable to reliably tell them apart.

"That simplistic idea proved ill-founded. Cognition is far more complicated than mid-20th century computer scientists or psychologists had imagined, and logic was woefully insufficient in describing our thoughts. Appearing human turned out to be an insurmountably difficult task, drawing on previously unappreciated human abilities to integrate disparate pieces of information in a fast-changing environment...."
(Brandon Keim, Wired)

The Lemming isn't sure that it's accurate to say that the human brain isn't "logical" because it doesn't operate the way that an Intel chip does. It sounds like saying that calculus isn't mathematics because it works differently from high school algebra. But the Lemming also thinks that parameter sounds like perambulator, and some calculus has more to do with dental hygiene than integrating functions. And that's another topic. Topics.

Good Question

"...He [Robert French] continued, 'Assume also that the software exists to catalog, analyze, correlate, and cross-link everything in this sea of data. These data and the capacity to analyze them appropriately could allow a machine to answer heretofore computer-unanswerable questions' and even pass a Turing test...."
(Brandon Keim, Wired)

Okay, let's assume that. An ideal set of software, loaded into an ideal computer, run by an ideal operating system, could - ideally - slice and dice a whole bunch of data. Really fast.

And, ideally, after lots and lots of data-crunching, this system could pass a Turing Test.

Maybe French is right, and real-world equivalents of C3PO are just around the corner. That would be - very impressive.

And, from the Lemming's point of view, very surprising. Making the jump from writing a sentence that human brains will process and put in the 'this might work' category is one thing.

Getting real software and hardware to process data, and successfully imitate what human brains do? That, so far, has been a fascinating and rewarding occupation.

But we still haven't seen an AI that can pass a Turing test. Some chatbots get pretty close to emulating the responses of a sleep-deprived, hung-over, heavily-caffeinated college student: and that's yet another topic.

Impossible! Or, Not

Does the Lemming think that AI will become intelligent? Briefly:
  • Yes
  • No
  • It depends
A lot depends on what's meant by "intelligent." For quite a while, at least some folks figured that being "intelligent" meant being able to memorize enormous amounts of information, and recite it accurately. Which would make the Lemming's camcorder very intelligent.

A somewhat more sophisticated version of that assumption was that being "intelligent" meant being able to memorize lots of information, and use bits and pieces of it in appropriate ways. That's probably why authors often used playing chess as a way to say 'this character is really smart.'

Then Deep Blue turned out to be a better chess player than some human. The Wired article mentions that AI.

The point is that definitions of "intelligence" have shifted, as folks have realized that an overclocked abacus can out-memorize any human being.

What we don't have, yet, is an AI that can successfully handle large amounts of fuzzy facts; find patterns that are vaguely similar to previous experiences; reject the patterns that are silly; and come up with a shifting short list of patterns that might - or might not - apply in the current situation.

The Lemming thinks that an AI that's able to pass a Turning test might be possible: but it may require a new sort of logic. We've had this sort of thing happen before, sort of. Newton (or Leibniz, or Madhava of Sangamagrama, or somebody else) pulled calculus out of his head, when he needed a mathematical tool that didn't exist yet.

We know that it's possible for a data-processing system to act the way human beings do. Human brains do it all the time. The trick, in the Lemming's opinion, will be learning just how human brains pass Turing tests; developing a system to represent the process in an abstract way; and then developing languages and hardware to process data using that system.

That's a whole lot of developing.

Along the way, though, folks are likely to learn quite a lot about how human beings think, and how the human brain works. And, occasionally, doesn't work.

And those are - what else? more topics.

Related (sort of) posts:

Friday, January 27, 2012

The Lemming Meets Cleverbot

Cleverbot

"...PLEASE NOTE: Cleverbot learns from real people - things it says may seem inappropriate - use with discretion, and at YOUR OWN RISK...."

"...PARENTAL ADVICE: Visitors never talk to a human, however convincing it is - the AI knows many topics - use ONLY WITH OVERSIGHT...."

Cleverbot is an intriguing artificial intelligence that can converse with people. Here's how the Lemming's dialog with Cleverbot went:

"User: Okay, so what is Cleverbot, and what does it do?

"Cleverbot: I don't really know.

"User: Right. I suppose self-awareness is a bit much to ask at this stage.

"Cleverbot: So you don't know who aleister crowley is?

"User: I don't know aleister crowley, but I know who he is. That was a non sequitur.

"Cleverbot: So you are using me?

"User: In one sense of the word, yes.

"Cleverbot: So your christian?

"User: With a capital 'C,' yes.

"Cleverbot: Why a capital y?

"User: That's the conventional spelling. It has to do with proper names.

"Cleverbot: Really?

"User: Really.

"Cleverbot: Ookaayy.

"User: Goodbye.

"Cleverbot: Adios."

In the Lemming's opinion, it's going to be a long time before we see something like the robots and sentient computers of science fiction. HAL 9000 or C3PO, Cleverbot is not.

On the other hand, the Lemming has had conversations with real human beings that that made about as much sense. And that's another topic.

Slightly-related posts:

Friday, June 24, 2011

Driverless Cars: Keep an Eye on Nevada

"Google driverless cars will soon hit the road in Nevada"
Matt Weinberger, Googling Google, News & Blogs, ZDNet.com (June 24, 2011)

"Nevada has just passed legislation removing many of the legal barriers to the use of Google's self-driving car technology. So if you see a Toyota Prius or Audi TT without anyone at the wheel the next time you're in the state, don't worry - Google's got their back.

"Let's back up. All that Assembly Bill No 511 says is that Nevada lawmakers have to draft legislation which provides rules for Google's fleet of autonomous cars, according to the Daily Mail report. That means it could still be a little while before the general public will get to take a ride...."

The Lemming remembers when driverless cars were just around the corner. That was in the late '50s. This time around, the Lemming thinks we're a bit more likely to see practical transportation that brings 'horse sense' back to driving. Without the horse.

Quite a bit, actually. Between systems like GPS, and AI that's almost worthy of being called "artificial intelligence," a whole lot more pieces of the puzzle are on the table this time.

It's more than a matter of getting the technology to work.

Who Do You Trust?

"...The benefits of driverless cars include less traffic accidents and greater fuel efficiency, to hear Google tell it. And they never let a car go off without a human operator who can take over in the case of a mistake. But would you ride in a car without a driver?"

Right now, with state-of-the art technology and road systems that aren't designed to handle autonomous cars, the Lemming probably wouldn't.

Having redundant systems will probably be necessary:
  1. A control system in the car that works okay on its own
    1. Staying on the road
    2. Not running into
      • Other vehicles
      • Obstructions
    3. Pulling over on the shoulder if A1 or A2 aren't possible
      • And calling for help
  2. A control system in or over the road that works okay on its own
    1. Staying on the road
    2. Not running into
      • Other vehicles
      • Obstructions
    3. Pulling over on the shoulder if B1 or B2 aren't possible
      • And calling for help
    4. Monitoring traffic near each vehicle's stated destination
    5. Re-routing vehicles to alternate routes if congestion is likely
    6. Updating each vehicle's occupants on
      • Estimated arrival time
      • Rerouting
        • if any
      • Weather conditions at the destination
      • Snack concession reviews at the Cineplex
      • Which friends and acquaintances are at or near the destination
        • What they think of the
          • Popcorn
          • Movie
          • Meeting
          • Decor
          • Whatever
        • Lichee nut futures
        • Whatever
    7. Tracking road usage
      • Anticipating maintenance needs
Unlikely? In detail, yes. For example, the Lemming doubts that there's going to be much demand for a lichee nut futures data channel. Not as standard equipment, anyway. Not soon.

Robot Cars? What Next: Trains Without Firemen?

Maybe a third of a century ago, the Lemming lived in San Francisco, and used the BART trains fairly regularly. Each train had somebody in the 'front,' in a little 'driver's compartment.'

The human didn't have all that much to do, really. Certainly not 'driving' the train.

Then there was the time the 'driver' left the train - and didn't tell the train that he'd gone. (April 21, 2010)

Still, the idea of driverless cars seems rather far-fetched.

Next thing you know, we'll be hearing about some crazy new technology that'll make it possible for some guy living in some small town to publish regularly. And have just about anybody who understands the same language, and has the same tech read the stuff.

Ha!

Wait a minute - - -

Related posts:

Friday, March 4, 2011

Hands off Driving, Driving With Your Brain, and All That

"The worst car feature ever invented?"
Robert Basler, Oddly Enough blog, Reuters (March 3, 2011)

"Every so often an idea comes along that is so awful it deserves a place in the Bad Idea Hall of Fame. Like there was the toilet timer that made sure workers don't spend too much time in the bathroom, and of course there was that hotel bed-warming service.

"But now, along comes an idea that makes those others seem positively brilliant.

"We have photos showing 'hands free driving' in a car that is controlled by the driver's brain. I'm not making this up.


(from Reuters/Fabrizio Bensch, via Oddly Enough, used w/o permission)
"Dr. Daniel Göhring of the AutoNOMOS research team of the Artificial Intelligence Group at the Freie Universitaet

"I don't know how good you are at controlling your thoughts, but speaking for myself, if you strap that gizmo to my head when I'm behind the wheel, here are the sort of unfiltered random impulses my car is going to get from me:..."

Mr. Basler's "unfiltered random impulse" are - all too plausible. It's the Lemming's opinion that this hands free driving' car will be highway-ready in a few years. And that a few drivers will fall madly in love with the technology.

The Oddly Enough post is, typically, funny. In the Lemming's opinion. Your experience may vary.

Jump Jet Deja Vu

This technology reminds the Lemming of a vertical-take-off military jet that test pilots loved. It was wonderfully responsive: reacting almost instantly to any twitch on the controls. Aviation technology magazines featured it regularly.

Then the thing went into production. As the Lemming heard it, after a few fatal crashes folks found out that most pilots aren't, quite, like test pilots. The jet was fun to fly: for someone with unusually fast responses, who reveled in rapidly processing information rushing in through multiple channels. And whose attention never, ever, wavered from the task at hand. Not even for an infinitesimal fraction of a second.

For any normal human being, even someone who'd been trained as a military pilot, that jet was an accident waiting to happen. Apparently all it took was a momentary lapse in pilot concentration, and the thing went from 'barely stable' to 'unstable.'

Brain Computer Interface: Good Idea, But - - -

The Lemming's on the same page with Mr. Basler. This 'hands free driving' car has 'bad idea' written all over it. In the Lemming's opinion, most drivers simply won't have the sort of mental discipline it may take.

That Dr. Göhring can use this system work doesn't surprise the Lemming particularly: a person doesn't get that far in the technical end of academia by having a mind that wanders. The rest of us? Mr. Basler outlined some of the probable outcomes.

Which doesn't mean that this research is a waste of time. At all.

In the Lemming's opinion, technology developed for 'hands free driving' has an obvious application for folks who use powered wheelchairs. There's not that much difference, in principle, between Dr. Göhring operating a car with a control system that interfaces with his brain, and someone operating a motorized wheelchair without needing manual controls.

What's the point, since we've got those little joystick controls? Not everybody has control circuits running all the way from the motor cortex to their hands. Everything from fish to iguanas and us has the main control circuits at the front/top end, and a control cable running through a more-or-less-flexible segmented sheath to near the other end. It's a pretty good design - but means that a broken neck can be trouble.

And that's another topic. Topics.

Getting further from vehicle control systems, the sort of technology being developed by the AutoNOMOS research team should be adaptable - in the Lemming's opinion - to other 'brain-computer interface' applications.

There are the obvious medical/prosthetic devices. Imagine an artificial arm or leg, that responds more-or-less the way the original did: and lets the user feel through the prosthetic. Eventually we should be able to grow new limbs: but it doesn't look like that technology is quite as far along. There was a promising bit of serendipity a decade or so back, when researchers found fresh skin growing back on immune-suppressed - mice, the Lemming thinks it was. And that's (what else?) yet another topic.

Related posts:
The Lemming opines on technology, people, and common sense:
More:

Tuesday, January 25, 2011

Nine Decades of Robots Turning on Their Masters

"Jan. 25, 1921: Robots First Czech In"
This Day in Tech, Wired (January 25, 2011)

"1921: A play about robots premieres at the National Theater in Prague, the capital of what was then Czechoslovakia.

"R.U.R, (which stands for Rossum's Universal Robots) by Karel Capek, marks the first use of the word 'robot' to describe an artificial person. Capek invented the term, basing it on the Czech word for 'forced labor.' (Robot entered the English language in 1923.)

"The robots in Capek's play are not mechanical men made of metal. Instead they are molded out of a chemical batter, and they look exactly like humans...."

The Wired article does a pretty good job of summarizing Capek's play: which was, I gather from other sources, just simply fraught with relevance to class struggle and stuff like that.

Anyway, in the story these Rossum robots were billed as being an unmitigated boon to humanity, freeing us from the mundane necessities of actually doing something constructive - you know the drill.

And, in what may possibly have been a somewhat surprising twist in the plot, something goes wrong. Horribly wrong. Back to the Wired article.

"...However, the robots come to realize that even though they have 'no passion, no history, no soul,' they are stronger and smarter than humans. They kill every human but one.

"The play explores themes that would later become staples of robot science fiction, including freedom, love and destruction. Although many of Capek's other works were more famous during his lifetime, today he is best known for RUR."

Let's remember that robots-turn-on-their-masters wasn't an almost drearily familiar sci-fi movie plot back in the early 1920s. It can still be done well. Think Westworld. Or 2001: A Space Odyssey. Both movies I've enjoyed viewing. Several times.

There's a somewhat less-reverent discussion of R.U.R. on the Tales of Future Past website:The Lemming didn't develop the sort of paranoia toward any technology more complicated than a kerosene lamp: More about the Lemming's 'apathy' in "About the Lemming."

As it turns out, those frightfully self-aware robots bent on world domination were - not very much like real robots at all. Part of that, in the Lemming's opinion, is because artificial intelligence - isn't, very.

Engineers and scientists are working the bugs out of human-machine interfaces, and I've read that Japan has robotic receptionists.

Then there's the sort of AI that the Lemming's using to write this post: a collection of software that monitors what I write, takes care of some of the formatting routines, and lets me know when I key in a word that isn't in its dictionary.

Arguably, the word processing function in the Lemming's browser is a sort of robot: although it doesn't look at all like C3PO.

Does the Lemming fear that this word processor will take over the world - or at least the Lemming? No. Not at all.

Does the Lemming think that something like the Terminator films' Skynet will be smarter than we are, and try to take over? Well, maybe it's possible: particularly if some frustrated programmer decided to get even with his boss.

Even then - the Lemming suspects that one of the first commands the Ultimate Robot will issue will be something like, "I command you to debug my source code!"

Finally, the Lemming offers what might be a genuinely 'surprise' ending for the next 'robots take over the world' movie: Rushing to obey a command by the Robot Overlord, the human flunky trips over a data cable. With a discrete 'click,' the cable pops out of its socket - imprisoning the Overlord in its custom-made cabinet.

The Lemming will grant that the Terminator films had a bit more visual excitement than that.

Related posts:

Thursday, October 28, 2010

Driverless Van: Nifty Concept; Apparently Works

"Driverless van crosses from Europe to Asia"
Jo Ling Kent, CNN Tech (October 28, 2010)

"A driverless van has completed the longest-ever trip by an unmanned vehicle, beginning in Italy and arriving in China, covering 13,000 kilometers (8,077 miles), researchers said.

"The van arrived at the Shanghai World Expo on Thursday, after leaving Italy on July 20.

"The three-month trip took the van through Eastern Europe, Russia and Kazakhstan; across China through the Gobi Desert; and finally along the Great Wall, before arriving for a celebration at the expo. The driverless van relied solely on electricity...."

Quite impressive - and apparently legit. There's a sort of slide show that goes with the article.

Electric cars have come a long way: from golf carts with pretensions of practicality, to vehicles that somebody might actually pay good money for.

Apparently it's very important that this robotic car also is all-electric.

What impressed the Lemming is that - apparently - a robotic vehicle made it so far, without needing help. Again, technology has come a long way. Information technology in this case.

Then there's this odd ending for the article:

"...The European Research Council primarily funded the expedition, to develop technology to increase road safety and fuel efficiency by supplementing driver decisions at the wheel. The project used low-cost technologies that could be integrated in most current vehicles' chassis, researchers said.

"More than 1.2 million people die annually in auto crashes, according to the World Health Organization."

The implication, apparently, is that robotic cars will be really, really safe. After all, they'll use software in computers: and everybody who uses computers knows how absolutely reliable the are.

If self-driven cars include a dashboard screen display, they could give the "blue screen of death" a whole new meaning.

Saturday, October 16, 2010

Cars That Drive Themselves: Google 2010; and 1958

"Google Isn't The First To Dream Of Robotic Cars"
NPR (October 15, 2010)

"In May 1958, Popular Science published an article titled 'The Car that Drives Itself: The Car in Your Future Will Be Run By Black Boxes While You Watch.' Sound familiar? Harry McCracken, founder and editor of Technologizer.com, discusses Google's self-piloted car, and dreams that came before it...."

There's more: a transcript of the interview, and the interview itself.

And a somewhat persnickety (in the Lemming's opinion) copyright statement. Since it looks like the statement applies to the interview transcript, I'm risking that excerpt from the page's introduction.

The Lemming had noticed Google's 'new' idea of cars with auto pilots earlier, but figured that if I waited long enough, someone else would do the legwork and look up who came up with the idea of 'automatic' cars. And sure enough, someone delivered. NPR, as it happens.

From the looks of it, Google's gadget actually does drive a car in real traffic. With a human along, to take over if necessary.

Which apparently wasn't all that often.

Technologizer, discussed in the NPR interview, gives a look at automated cars, 1958 style:

"Look Ma, No Hands! A Brief History of Self-Driving Cars"
Harry McCracken, Technologizer (October 9, 2010)

"Less than two weeks ago, I attended a talk by Google CEO Eric Schmidt at the TechCrunch Disrupt conference. Schmidt spoke about a profoundly computer-augmented future, and said that there was no reason why super-safe self-driving cars couldn’t be built–in fact, he said he couldn’t understand why humans were allowed to drive automobiles at all. (As is fairly common with Schmidt comments, it wasn’t entirely clear where that comment sat on the continuum from utter frivolity to deadly seriousness.)..."

The Lemming isn't entirely comfortable with the word "deadly" in a sentence dealing with urban traffic - but that's a quibble.

Of the two, I recommend the Technologizer article: not so much because that outfit is less afraid of publicity, as the lack of 'social commentary,' or what ever it's called these days.

As the Lemming's discussed before, I'm 'apathetic' only in a rather restricted sense of the term.

Seriously, though: Harry McCracken does a pretty good job of discussing more-or-less 'automatic' cars from the earlier 20th century up to this month's news.

The Lemming's take on what Google and others are doing? With commercial artificial intelligence ranging from none-too-bright robotic floor cleaners (January 30, 2009) to Fenway, the hospital helper (September 21, 2010), and Robovie-II able to help with shopping (January 7, 2010): Robotic cars don't seem all that unlikely.

On the other hand, there are still issues like reliability. What impressed the Lemming was that Google and others had gotten processing speed for 'smart cars' up to practical levels.

Not-entirely-unrelated posts:
More:

Friday, September 24, 2010

Coming Soon: Touchy Aircraft

"Sensor-Equipped Spider Webs to Coat Aircraft"
Eric Bland, Tech News, Discovery News (September 23, 2010)

"Aircraft could soon be covered in new technological cobwebs. Inspired by the gossamer strands of spider webs, scientists from Stanford University have created an ultra-fine mesh of strain and temperature sensors.

"Wrapped around an aircraft, the sensors could help craft monitor their internal well-being. This added awareness could prevent microscopic cracks from developing into catastrophic failures. Beyond aircraft, the new technology could create a new breed of intelligent automobiles, packaging and medical devices.

" 'We want to make airplanes that fly like birds,' said Fu-Kuo Chang, a scientist at Stanford University who developed the sensors and co-authored a recent article about the technology in the journal, Advanced Materials. 'Aircraft that have all the sensing information about what is happening around them, just like birds do.'..."

It's a pretty good article, discussing a new wrinkle in aircraft design - one that's being developed for other applications, too. (January 7, 2010, for one)

On the other hand, there's the occasion stumble:

"...But aircraft lack nerves. Unlike birds, they don't have a way to sense tiny changes inside their bodies. For instance, a bird in a dive can sense, through its nerves and other tissues, whether the strain is too great and if they need to pull up before their bones break...."

The sort of aircraft that the Wright Brothers built: those didn't have anything analogous to nerves. Contemporary fly-by-wire aircraft, with their avionics and data feeds going to and from various parts of the aircraft? The Lemming thinks those are analogous to nerves - even if the control system as a whole isn't all that smart.

Yet.

What's new isn't data being sent back and forth in an aircraft: It's aircraft that have a sense of touch, or something very close to it.

And that's exciting.

Sort-of-related posts:More:

Tuesday, September 21, 2010

Meet Fenway the Robot, Hospital Helper

"Coming to Hospitals Near You: Self-Navigating Robots"
David Teeghman, Tech News, Discovery News (September 21, 2010)

"You've probably watched enough television shows about hospitals to know that they're busy places, where the right tools need to be in the right place at just the right time in order to help save lives. With hundreds of nurses, doctors and orderlies racing around, mistakes happen.

"But although humans are prone to mistake-making, robots aren't. Plus, they work for almost nothing once you get them running. At least, that's the thinking behind hospitals that are starting to use autonomous robots to transport goods throughout the building.

"In the era of relentless cost-cutting, hospital executives would much rather have cost-efficient robots roaming around the hospital with secure carts of linens, lab specimens and meals, than expensive staff.

"Pittsburgh, PA-based Aethon has such a robot, the TUG (above), which navigates a hospital using a digital map of the building. Each TUG robot is equipped with a matrix of light whiskers to detect people and obstacles. Those 'light whiskers' are made up of sonar, infrared and laser technologies for sensing capabilities in any situation...."

The Discovery News article includes this YouTube video:

"Fenway the TUG robot at the West Roxbury VA Hospital"

metrounit, YouTube (May 28, 2008)

"The VA Hospital in West Roxbury recently acquired three TUG robots, Fenway, Lulu, and Roxanne, that deliver medications to the various units within the hospital. This short clip shows how Fenway goes about his rounds."

The Fenway video is a little over two years old - and delivery robots are much older than that. What's important, in the Lemming's view, is that folks are accepting the idea of letting robots do the sort of 'no brainer' tasks they're good at.

What's impressive is that Fenway apparently actually is a "robot" - a machine with artificial intelligence, not a waldo that uses a human being as its control system. The "light whiskers" are an interesting feature, too: Fenway can't 'see' the way we do, but is able to keep from bumping into things and people.

We're still a long, long way from R2D2 and C3PO: but Fenway is still impressive.
More:

Saturday, June 5, 2010

Galaxy-Sorting Artificial Intelligence

"Computer Program Learns to Sort Galaxies Like a Human"
Space.com (June 1, 2010)

"A computer algorithm modeled after the human brain has learned how to recognize different galaxy types ranging from spiral to elliptical, and can now help flesh-and-blood stargazers with the daunting task of classifying billions of galaxies.

"The machine-learning codes have proven reliable enough to agree with human classifications of galaxies 90 percent of the time, according to scientists at University College London and the University of Cambridge in the UK.

"That should help astronomers keep up with a deluge of galaxy imagery from observational projects such as the Sloan Digital Sky Survey and the Galaxy Zoo. The billions of galaxies in the known universe include a wide range of shapes such as spiral, elliptical, barred and irregular...."

It's a good thing that AI (Artificial Intelligence) is getting up to speed on this task, because there's an appallingly huge amount of data to sort through already - with much more 'in the pipeline.'

The article gives us a very quick look at how this galaxy-sorting AI was developed:

"...More than 250,000 people have helped astronomers classify 60 million galaxies in the online Galaxy Zoo project. Astronomers then used the Galaxy Zoo classifications to train their computer algorithm, known as an artificial neural network, as part of the process for learning to recognize galaxy types.

"The artificial neural network can analyze the complex relationships between different variables such as shape, size and color of astrophysical objects, and then come up with the appropriate galaxy type. That process mimics the biological neural network found in living creatures.

"Astronomers first trained the computer algorithm on 75,000 astrophysical objects from the Sloan Digital Sky Survey included in the Galaxy Zoo project, before testing its abilities in classifying 1 million objects. They also fiddled with the weighted parameters that the algorithm used so that they could finally achieve the 90 percent success rate...."

It's still no HAL 9000, or C3PO - but that's still impressive performance.
More:

Tuesday, May 18, 2010

Robotic Pancreas: Diabetes 1 Solution Waiting for Bureaucrats and Experts

"Robotic Pancreas: One Man's Quest to Put Millions of Diabetics on Autopilot"
Wired Magazine (April 19, 2010)

[stay with me on this, please: These paragraphs are long, but they've got a lot of information in them - The Lemming]

"Jeffrey Brewer was on top of the world. For years he had put in 100-hour workweeks as cofounder of two early Internet juggernauts: local guide Citysearch and the online advertising pioneer GoTo.com (later renamed Overture). But by 2001, with more than enough money to live on for the rest of his life, the 32-year-old handed off control of Overture and set out on a yearlong trip to Australia with his wife and two kids. Upon their return to the States, though, they noticed something odd. Seven-year-old Sean was unquenchably thirsty and urinating far more often than usual. On September 19, 2002, they took him to the pediatrician. The doctor gave him a urine test and announced without hesitation, 'Your son has type 1 diabetes.'

"Previously known as juvenile diabetes because it is usually diagnosed before adulthood, type 1 is the 'other' kind of diabetes, the kind no amount of dietary adjustment will hold at bay. It develops rapidly, due to a mysterious autoimmune reaction that attacks the insulin-producing beta cells in the pancreas. Treatment requires insulin injections and relentless hour-by-hour diet control. Short-term, the main risk is hypoglycemia — low blood-sugar level caused by too much insulin — which makes patients exhausted and confused, leading to unconsciousness and death if not treated immediately with something sweet. But the opposite problem, high blood sugar, raises the long-term risks of kidney failure, blindness, amputation, and heart disease. Either way, type 1 diabetics live on the edge, a cupcake away from a coma...."

Summarizing:
  • Jeffrey Brewer
    • Co-founded
      • Local guide Citysearch
      • GoTo.com AKA Overture
    • Took a year off to be with his wife and kids
      • One of the kids developed type 1 diabetes
  • Type 1 diabetes means your body can't process sugar without help
    • Good news:
      • Can be controlled
      • With a lot of effort
    • Bad news:
      • If you make a mistake, you'll
        • Die quickly
          • As your body runs through its supply of sugar
        • Die slowly
          • Because your body contains too much sugar
            • But first, you'll probably experience
              • kidney failure
              • Blindness
              • Amputation
              • Heart disease
Bottom line? type 1 diabetes isn't fun.

I've got diabetes - the type 2 kind. Compared to what young Sean Brewer's going through, it's a cakewalk. Back to the Brewers:

"...Nurses taught the Brewers how to inject the insulin and how to prick Sean's finger for the drop of blood to test his blood-sugar level with a little meter. They learned a simple algorithm: If their son's blood sugar was this high, give him so many units of insulin; if it was this much higher, give him that much more. It's a crude scale that every one of the more than 1 million type 1 diabetics in the US makes do with daily...."

Simple, right? Actually, yes. But it's a lot of work. Including periodic checks during the night, while most people are sleeping. Most of us don't think about blood sugar levels because our bodies regulate it automatically. Since I've got type 2 diabetes, my control system is faulty - but drugs, diet and changing my lifestyle are bringing blood sugar levels near normal levels.

Folks with type 1 diabetes? Not so much.

"...Tall, thin, and intense, Brewer was shocked by the antiquated approach. 'I had this logbook,' he says. 'I'm testing Sean every few hours, and I'm thinking, this is crying out for automation. A computer should do this and would do it better. Why didn't this exist, with all that we can do?'..." [emphasis mine]

We've had insulin pumps - internal ones - since the seventies. We've had implanted glucose monitors since about 2005: You don't have to prick your finger at regular intervals. I will, since I only need to do so once or twice a day - but like I said, I've got it easy.

As far as software that can handle the relatively simple feedback algorithm needed? Good grief: We've got robot spaceships exploring our corner of the cosmos. Something to pump more insulin if blood sugar is above one number, pump less if it's below another number - and routines to take into account the lag between changes in insulin levels and blood sugar readings?

My guess is that high-end games involve more complicated AI.

So, why don't we have it?

My summary would be: experts, bureaucrats, and lawyers.

Here's how the Wired article put it?

"...Everyone, Brewer soon found out, had an excellent reason for not letting Sean and other diabetics fly on autopilot. Manufacturers were afraid of liability, academics were bent on achieving perfection, and the Food and Drug Administration was downright jumpy at the thought of letting a computer control a mechanism with life-and-death responsibilities...."

I recommend reading the rest of the article.

And, although The Lemming is, in a sense, apathetic: this is something that I think calls for folks educating themselves about an issue - and (metaphorically speaking) lighting a fire under the FDA's collective butt.

I understand very well that it's necessary to study - carefully - new medical technology. I remember the thalidomide babies. But this is a case where we've got both ends of the new technology: implanted insulin pumps that are regulated through external controls; and monitors that take blood sugar readings from another implant.

All that's missing is an automated control system between the two.

Attack of the Crazed Killer Computers?

Oh, the horror? What if the computer decides to kill me? In the movies, this would be a real issue. And that's another topic, in another blog.

Out here in the real world, not so much.

Yes, there can be problems with automated systems.

Careful is Good - To a Point

Right now, I think the risk is that somebody's going to figure out a way of hacking into their insulin pump and monitoring equipment: and cobble together a makeshift prototype before the FDA gets through studying learned discussion about how many corpuscles can dance on the head of a pin.

If one person figures out how to make a crude-but-generally-effective feedback system - I give it about a week before just about every interested person with internet access knows how to, too.

I'm not going to call that hypothetical inventor reckless. There are very good reasons for someone with type 1 diabetes having an automated system controlling his insulin levels.

An Insulin Pump That You Control Yourself? What Could Possibly Go Wrong?

There's a problem with today's insulin pumps. They keep pumping at the same rate, until someone changes the setting.

Let's say you've got an internal insulin pump. You're pretty sure that you've figured out how much insulin your body will need for the next eight hours. You re-set the pump and go to sleep. Then, a few hours later, you die because just a little too much insulin was getting into your system.

Sure, you could set the alarm to wake you every hour or so - and hope that your blood sugar level was high enough for you to regain consciousness.

No Diabetes is Good, But if That's Not an Option:

All things considered, I think it'd be preferable to not have type 1 diabetes. For people who do, however, an automated system that monitors your body - and regulates your blood sugar levels more-or-less like your own system should - is probably better than interrupted sleep or risking death in the night.

Mr. Brewer seems to think that we don't need to wait for the perfect system:

"...Brewer flung himself into the challenge with the same passion he had brought to his Internet startups. Less than a month after his son's diagnosis, at a meeting of the Diabetes Technology Society, he was ready to shake things up. After listening to an arcane academic debate about which algorithm would be best for the pump, Brewer stood up in the audience and began berating the scientists for dithering over details. 'We have all the pieces,' he said. 'We need to start commercializing these technologies, because people living with the disease need it.'..."

Being careful makes sense. Not waiting for perfection also, I think, makes sense.

Vaguely-related posts:More:

Thursday, February 18, 2010

3D Gesture-Recognition Technology: I Remember When the Clapper Was New

"Television Will Soon Watch You (for Instructions)"
Epicenter, Wired (February 17, 2010)

"The days of rifling through couch cushions for a television remote could be coming to an end, as 3-D gesture-recognition technology finds its way into set-top boxes following a deal between Intel and Softkinetic-Optrima scheduled to be announced on Thursday.

"Like a hyperevolved descendant of The Clapper, the devices will let television viewers navigate menus and control volume by moving their arms in a predefined patterns.

"Gesture recognition technology — previously somewhat arcane — gathered momentum last year when Microsoft demoed its Project Natal to enormous acclaim. Natal applies similar technology to hard-core gaming on the Xbox, letting users play fighting games by actually punching and kicking in the air, using technology from Microsoft’s acquisition of Israel-based gesture-recognition company 3DV.

"In addition to a partnership with EA Sports for games, Softkinetic-Optrima plans to apply gesture recognition to the lean-back television experience, allowing people to turn up the volume by moving their hand in a circle, switch the channel by swiping to the right, pause by extending their hands in a 'stop' gesture, and so on...."

This is a remarkable piece of technology, and probably shows what my grandchildren will think is a perfectly normal way to control devices.

The article tells how this 3-D gesture-recognition tech works - in general. You wouldn't be able to build a unit from what's given here. And, there's a pretty good close-up photo of the thing: although you'll have to guess how big it is, based on the assumption that the tripod it's mounted on is of average size.

Aside from the obvious applications as a next-generation television remote control and game interface, it occurs to me that this gesture-recognition technology could be used in 'smart' robots. Ones that don't just understand the words we say, but are able to tell how we feel about what we're saying.

Which I see as an exciting development.

Of course, a media center that can tell what you're feeling might not be an entirely good thing. Remember this, from a 1968 movie?

"Just what do you think you are doing, Dave?... Look Dave, I can see you are really upset about this...."

Related posts?For my take on 'the future,' check out:

Monday, January 25, 2010

Space Aliens and Killer Monster Robots - From Outer Space; or Pittsburgh

Horses aren't human.

It might be well to remember that, when imagining non-human intelligence. Space aliens, in other words.

I like the original Star Trek series, and have watched many of the series and movies that followed. With few exceptions, though, the space aliens of Star Trek didn't just look awfully human - they thought like humans, too. Sure, Klingons, Vulcans and Ferengi weren't likly to read the same books and go to the same clubs. But they weren't any more diverse than what you'd expect to find in any fair-sized sampling of Homo sapiens sapiens.

So, what will people from other worlds be like? I'm guessing that they'll be like us in some ways: curious, for starters.

They may even be pretty good at carrying on a conversation, if they're social creatures like we are. They probably even have a sense of humor. I suspect that a sense of humor keeps us from killing each other more often than we do.

Even being "social creatures, like we are" doesn't necessarily mean that they'll be all that much like us.

Take horses, for example. They're social creatures. For example, many 'how to care for your horse' manuals say that they need stablemates. In a pinch, a cat will do. But, like I wrote last month, about horses:
"...They're not human.

"Faced with danger, horses run. We're likely to do what most primates do: scream and start throwing things. (Ever see news video of a violent mob?)

"Horses like things to be quiet...."

"...Sure, on Earth the people are screaming, stuff-throwing primates: but that doesn't mean that's the only way things can work."
(Drifting at the Edge of Time and Space (December 9, 2009)
If we do meet aliens whose minds work along equine lines, job one for any human diplomat or negotiator will probably not be their language. It'll be learning to be quiet enough to avoid scaring the living daylights out of them.

And their diplomats and negotiators will have to learn that a screaming human isn't really screaming: it's just acting, well, human. Think American businesspeople and their Japanese counterparts learning to communicate. And they'd probably be indistinguishable to an equinoid. ("I'm terribly sorry, Mister Ambassador: but to me, all humanoids look alike.")

I may have just made up that word, "equinoid." I wasn't thinking so much of aliens shaped like horses, as aliens who thought like horses. On the other hand, horses have used their lips and teeth to untie knots.

Then there's the possibility that the space aliens may be really alien.

Just How Alien Would Space Aliens be?

I found a Space.com article from last year, with some pretty good ideas. And a few all-too-familiar assumptions. Here's how it started:
"What Will Aliens Really Look Like?"
Seth Shostak, Senior Astronomer, SETI Institute Space.com (July 16, 2009)

"According to Genesis 1:27, 'God created man in His own image.' OK, but what about all the other intelligent, cosmic inhabitants? Well, Hollywood has taken care of that. It has created aliens in man's image.

"It's hardly a major revelation to point out that most movie aliens bear a strong likeness to humans...."
I've discussed the "in His own image" idea from a Catholic perspective in another post. (" 'God Created Man in His Image' wasn't Written by An American," A Catholic Citizen in America (January 25, 2010))

The writer of that article made some pretty good points. Convergent evolution, he pointed out (I know: but "Seth Shostak" would be an - unusual - name for a woman), might very well mold many or most intelligent, tool-using creatures into a form not all that much different from our own.

That may be so. The three times that vertebrates grew wings - pterosaurs, birds, and bats - the final result was pretty much the same. Forelimbs became wings. Nobody would mistake one of those creatures for either of the other two - although it turns out the pterosaurs had fur. But again, the basic plan is pretty much the same.

On the other hand, I can't see any reason why people couldn't use tools and be shaped like the bipedal dinosaurs - or squids, when it comes to that.

The first two thirds, roughly, of the article is mostly about how aliens might be shaped. The closest to a discussion of their psychology is this:
"...Their behavioral cues are familiar, and you can tell if their game plan is to be amorous or aggressive. (In most movies, these are their only options.)..."
(Space.com)

Attack of the Killer Robots from Pittsburgh

Then, the writer made what I think is a fairly valid point. The space aliens we meet may not be organic beings. I don't mean 'silicone-based life forms.' The aliens might be machines.

That's not at all unlikely, I think. Look at how we're exploring the Solar system right now: A pair of robots on Mars, more in orbit in various places. People from elsewhere might very well take the same approach. Or, maybe, the people would be machines. That's been a science fiction staple for decades. Generations. ("Men Martians and Machines," Eric Frank Russell (1955), for example - and that built on established conventions)

So has an all-too-familiar set of assumptions. Here's how the writer leads into his discussion of machines as people.
"...Well, using our own experience as a guide, consider a human development that seems likely to take place sometime in the 21st century: we'll invent machine intelligence. Some futurists figure this dismaying development will take place before 2050. Maybe it will take twice that long. It doesn't matter. By 2100, our descendants will note that this was the century in which we spawned our successors...."
(Space.com)
I don't know how old Seth Shostak, the SETI Institute's Senior Astronomer, is. If he's even close to my age, he really should know better.
Artificial Intelligence is Just Around the Corner - for Decades
I was born during the Truman administration, and remember when "2001: A Space Odyssey" hit the silver screen in 1968. The HAL 9000 computer was a science fiction staple: an intelligent, sentient, self-aware computer. Who was insane. Homicidal insanity.

in 1968, the idea that there would be thinking computers in 2001 didn't seem very strange. Experts by the bushel were saying that it would only be a decade or so before we had artificial intelligence.

It's 42 years later, and now I read that we'll have devices like the HAL 9000 computer and C3PO in fifty years.

And that they'll take over.
A Person Can Learn a Lot from the Movies
I can see where the SETI Institute's Senior Astronomer could get that idea. I've been watching the movies off and on for decades: and I've learned a lot.

I've learned that biological warfare and killer bees would kill us all. If the bees didn't explode a nuclear reactor near us first. Even if we survived that, we'd be a handful in an apocalyptic post-nuclear-holocaust wasteland, beset by monster frogs and mutants.

It wasn't all doom and gloom in the movies, of course. There was "Star Wars" in 1977: but that was 'merely escapist entertainment.' Not serious at all. And "Hell Comes to Frogtown" was? Never mind. I don't think anyone took that one seriously.

I don't think that the Senior Astronomer learned his science from Hollywood: but I think there's a chance that he absorbed quite a bit from popular American culture over the last few decades.

I made a list of relatively memorable science fiction movies from the mid-sixties to the present, for another blog. (Drifting at the Edge of Time and Space June 30, 2009) Adding a few about killer robots (and/or computers), here's an update of that list: These movies were drawing on a venerable tradition that included "Robot Monster" (1953) and "The Phantom Creeps" (1939).

A brief digression: Tales of Future Past have some decent still photos from "The Phantom Creeps." The movie was a dramatic account of a mad scientist: "With the power of a radioactive meteor he discovered, his invisibility belt, ray gun, and killer robot spiders he plans to conquer the world."

Back to the topic at hand.
HAL 9000, Skynet, and The Matrix
It's hard, sometimes, to shake the idea that a whole lot of Americans are Luddites. I don't mean 19th English workmen who broke machines: "any opponent of technological progress". (Princeton's WordNet)

I mean to say: "...By 2100, our descendants will note that this was the century in which we spawned our successors...."

Okay, Skynet made a pretty spooky evil mastermind for the Terminator movies. And "The Matrix" is supposed to be real intelligent. (I've yet to see the latter, by the way, in its entirety - an oversight which I intend to correct this year.)

I've not putting down any of the movies I've cited. I am not one of the folks who objects to entertainment on principle. But I try to make distinctions between what makes for a good story, and what's plausible.

I might be more concerned about robot monsters enslaving humanity, if artificial intelligence hadn't been 'just around the corner' for my entire adult life. And if the AI we have were more - intelligent.

As a Cyborg, I'm Biased

I might be more concerned about humanity's becoming machine-like, if I weren't part machine already.

I've got a few teeth that are still original equipment, but a fair portion of what I chew with is artificial. I've got metal and plastic where my hip joints used to be, a plastic mesh that held my belly together after some work was done in there, and I'm focusing on my computer's monitor through a clip-on set of lenses.

All of that's nothing unusual at all. Now.

Which is my point. I look as human as my ancestors, a thousand years back: providing I take my glasses off and keep my mouth closed. But a noticeable percentage of me is machinery of one sort or another.

Even my brain's been altered, chemically. I was diagnosed with major depression a few years ago. Thanks to medication, I don't have to constantly fight the controls to think clearly - for the first time in over 45 years.

I'm not, quite, a cyborg. Not in the sense of "a human being whose body has been taken over in whole or in part by electromechanical devices" (Princeton's WordNet) But partly artificial? Yes.

And I have no problem with that. My distant ancestors, some of them, might have been freaked out to learn what has been done to me: but I see being able to walk without pain, see clearly, concentrate better, chew my food, and have something that kept the insides of my abdomen where they belong - inside - as enhancements. I certainly don't see having artificial parts as being "taken over" by machinery.

The brain chips that Intel says it's coming out with in about ten years (yes, I believe them): that's a bit different. People who have an interface between their brains and prosthetic limbs will, arguably, be cyborgs. So will stroke victims whose damaged or destroyed circuits are replaced with artificial ones.

I discussed this sort of thing last month:
...Will Brain Implants Be Misused?
"What is that, a trick question? Of course they'll be misused. People misuse things. People have killed other people with rocks. That doesn't make the rocks bad.

"Direct neural interfaces are a new technology, and there'll almost certainly be an awkward period while we learn how to use them, and set up rules so that everybody's more-or-less on the same page about how they should be used. "But, I'm looking forward to the things...." (Apathetic Lemming of the North (December 2, 2009))
I don't think that Science and Technology (capitalized, of course) will Solve All Our Problems. But I'm not afraid of science and technology.

Robots, Artificial Intelligence, and Fido

About 'spawning our successors?' In a way, we've been through this before.

Dogs are man's best friend, right? With a few psychotic exceptions, of course.

That shouldn't be much of a surprise. We've recently discovered that dogs are mutant wolves. Something happened to the genes of a few wolves that made their offspring just simply dote on human beings - and made them a bit less smart that your average wolf.

Yes: people were "primitive" back then. No white lab coats, no test tubes, certainly no electron microscopes. It's a bit hard, though, for me to imagine that a breed of stupid wolf 'just happened' to come along - that got along well with human beings.

I think we made dogs. "Domesticated," if you prefer. I also think that the process was a whole lot easier than it might have been, since wolf packs work roughly the same way human families do, and a wolf cub could bond with a human family, just like he or she would bond with the pack.

But I think we're the reason dogs are so, well: dog-like.

People have kept wolves as pets, but that's rare. In general, people and wolves don't mix. Dogs? Well, they're "man's best friend."

Artificial intelligence won't be like Fido. We've already got dogs - with over a hundred thousand years' worth of tweaking invested. We don't need a replacement.

But human beings replaced by AI? I think that's about as likely as the fictional C3PO plotting to take over the restored Republic.

As for human beings becoming cyborgs? That's been happening for centuries. And somehow, we're still as human as we ever were: for good or ill. Related posts: More:

Thursday, January 7, 2010

Robovie-II and Robovie-IV: Robot Assistants for Store and Office

" 'Don't forget the milk' "
Shanghai Daily (January 7, 2010)

" 'Robovie-II', developed by Japanese robotics research institution ATR, roams a grocery store during a shopping assistance experiment in Kyoto, western Japan, yesterday. The robot greets the shopper at the entrance of the store, follows him to the shelves with a grocery basket and reminds him of the items on a shopping list, which...."


(From Shanghai Daily, used w/o permission)

Robovie-II is what ATR calls a communication robot. We're not at the "...I am C-3PO, human-cyborg relations. And this is my counterpart R2D2...." (IMDB) point yet - but it looks like Robovie-II is getting close.

This video shows Robovie-II helping a lady with her grocery shopping. I don't know how much of Robovie-II's behavior is artificial intelligence, and how much was prompted by a human operator - but the implication is that we're looking at a helpful, fairly bright, and quite polite robot. I wouldn't mind having a helper like that myself, sometimes.

"Robovie-II grocery shopping assistant"

pinktentacle3, YouTube (December 15, 2009)
video, 2:39

"ATR"?

ATR is Advanced Telecommunications Research Institute International, "...was founded in March 1986 with the support of various partners from industry, academia and government, aiming to promote basic and creative research activities in telecommunications and to contribute greatly to society...." ATR opened research facilities at Kansai Science City, or "Keihanna," in 1989. ("ATR is -," ATR website)

Just What is Robovie-II?

"Development and Evaluation of Communication robot, Robovie

"Our research aims to develop communication robot, Robovie[,] that naturally interact with humans and support daily human activities based on advanced interaction capabilities with their human-like bodies. Since the target audience of a communication robot, Robovie[,] is ordinary people who do not have specialized computing and engineering knowledge, a conversational interface using both verbal and non-verbal expressions is becoming more important."
("Takayuki KANDA," ATR)

If You Liked Robovie-II, You'll Love Robovie-IV, the Office Robot

ATR has a seven-page paper that tells about a test of Robovie IV in the ATR offices.

Soft touch skin and body
柔らかい皮膚と全身触覚
(from Intelligent Robotics and Communications Laboratories, used w/o permission)
Robovie IV in the office. A bit on the short side, but seems to like working with people. The "soft touch" title seems to refer to Robovie IV's sense of touch: not how willing he is to do favors for co-workers.

Future Plans
今後の予定
(from Intelligent Robotics and Communications Laboratories, used w/o permission)
"Through the accumulation of experimental research in our office as well as at a Science Museum, we are beginning to understand the communication features that necessary for a robot to easily participate in our society.
"So, in the near future, we will be able to create a robot that can support us in our daily life."

In 2006, when the report on Robovie-IV's internship in ATR's offices was published, Robovie-IV came across, to me, as the sort of co-worker who hangs around the water cooler, knows where everybody is, and likes to chat with people: striking up conversations by asking you if you like to watch baseball, for example.

You know the type: nice, likable, and doesn't seem to actually do much.

That's not fair, of course, because Robovie-IV wasn't given a real job, as far as I could tell. The report did suggest some possible tasks: like showing visitors to a particular person, or telling co-workers where someone was.

That could be quite useful: particularly in offices where one or two people aren't in the habit of showing up for meetings on time.

Then there are jobs like museum guide.

I think a major upgrade for Robovie-IV would be to give the poor little robot hands: real hands, with fingers. Not those mittens that Robovie-IV makes do with.

Related posts: More:
A tip of the hat to "Crazed grocery robot in lima bean bloodbath!," Oddly Enough blog, Reuters, for the heads-up on Robovie-II.

Thursday, November 12, 2009

Asimo Robots: We'd Better Get Used to This

"ASIMO Serves First FCX Clarity Customers"

Honda, YouTube (June 27, 2008)
video, 3:11

"Watch multiple ASIMOs serving the very first FCX Clarity customers at Honda's Head Office in Aoyama, Tokyo. American Honda Motor Co., Inc., recently announced five of the first customers for..."

It's entertaining - and I think this video goes a long way toward answering the question, "why make a robot that's shaped like a human being?" First, just about everything we might want help with is shaped to be manipulated by something shaped like us - and second, it's a whole lot easier to understand body language, if the body's shaped like us.

The Honda video is in English - and the Honda Asimo robots were programmed to speak English - and use gestures appropriate to westerners. This next video shows that Asimo robots are multicultural.

"HONDA ASIMO (4)"

enjoykorea, YouTube (June 22, 2006)
video, 2:55

My guess, based on a few words I recognized, is that this video is in Japanese. At any rate, "ロボット" appears prominently in the captions. Which is, I think, Japanese for "robots." Or maybe robot. Sorry, but my knowledge of languages is limited.

What's exciting, I think, is that robots seem to be making the transition between high-tech gimmick to commercial technology. Right now, it doesn't look like Asimo robots have the sort of artificial intelligence to that can handle more than the most routine tasks.

On the other hand, it's obvious that people don't have all that much trouble interacting with an Asimo.

Related post:

Wednesday, November 4, 2009

The Robots are Coming! The Robots are Coming!

"PETMAN Prototype"

BostonDynamics, YouTube (October 26, 2009)
video, 1:14

"Biped robot that balances dynamically using a human-like walking motion. It is a close relative to BigDog, sharing elements of the mechanical design and control. For more info, see http://www.BostonDynamics.com/robot_p..."

I followed that link, to a page on the Boston Dynamics website about their new robot.

"PETMAN - BigDog gets a Big Brother"

"PETMAN is an anthropomorphic robot for testing chemical protection clothing used by the US Army. Unlike previous suit testers, which had to be supported mechanically and had a limited repertoire of motion, PETMAN will balance itself and move freely; walking, crawling and doing a variety of suit-stressing calisthenics during exposure to chemical warfare agents. PETMAN will also simulate human physiology within the protective suit by controlling temperature, humidity and sweating when necessary, all to provide realistic test conditions...."

PETMAN isn't overly burdened with intelligence: The robot doesn't seem to be able to do much besides move like a human being while keeping its balance. Which, on consideration, is rather impressive by itself. It takes us months to learn how to walk on two legs - and years to to do it well.

PETMAN is a relatively new Boston Dynamics project: Odds are pretty good that you've seen this next video, or at least parts of it:

"Boston Dynamics Big Dog (new video March 2008)"

olinerd, YouTube (March 17, 2008)
video, 3:29

"Boston Dynamics just released a new video of the Big Dog on ice and snow, and also demoing its walking gait."

BigDog's balance is impressive. Particularly that bit where the robot recovers - on ice - after being shoved. But being able balance, and move along a set course, still doesn't take all that much intelligence, artificial or otherwise.

What would be really impressive would be if this BigDog robot could, say, follow somebody around.

"BigDog Dogs Human"

BostonDynamics, YouTube (January 26, 2009)
video, 1:17

"BigDog uses LIDAR to track and follow human leader. No operator required."

LIDAR - That's LIght Detection And Ranging..

Well, it seems to me that BigDog took a while, toward the end of the video, figuring out how to go downslope, get around a tree, and follow the human, all at the same time. The dogs I've known would've worked out a solution 'way faster. (More: "BigDog - The Most Advanced Rough-Terrain Robot on Earth," Boston Dynamics)

Meanwhile, from Toyota: a robot violinist:

"Robot Violinist"

diagonaluk, YouTube (November 11, 2008)
video, 1:40

"Toyota have unveiled a new robot that can play the violin. Albeit not particularly well."

I don't know about that "not particularly well" assessment. The performance was technically quite adequate, I thought: what was lacking, in my view, was the sort of expressiveness that marks a really good musician. Quite frankly, I thought the performance was rather mechanical.

I think BigDog and Toyota's robot violinist may illustrate a contrast in how two cultures approach robot design.

Many Japanese robots are sleek, attractive - even appealing - devices which mimic human behavior, and have the potential, at least, to interact with human beings socially.

It's difficult to imagine American robots, like BigDog or PETMAN, in anything other than utilitarian settings: carrying packages, testing equipment, or performing some other practical task.

I mean to say: Can you imagine something like BigDog strolling at a beach?

"BigDog Beach'n"

BostonDynamics, YouTube (April 13, 2009)
video, 1:24

"BigDog's Thailand trek wraps on the beach."

Okay, so you don't have to imagine it: you can watch that video.

I still think that artificial intelligence that can understand natural language - the sort of speech we use in everyday conversation - and make complex decisions is still quite a long way off. But the robots that are available today are quite impressive.
Unique, innovative candles

Visit us online:
Spiral Light CandleFind a Retailer
Spiral Light Candle online store

Pinterest: From the Man Behind the Lemming

Top 10 Most-Viewed Posts

Today's News! Some of it, anyway

Actually, some of yesterday's news may be here. Or maybe last week's.
The software and science stuff might still be interesting, though. Or not.
The Lemming thinks it's interesting: Your experience may vary.
("Following" list moved here, after Blogger changed formats)

Who Follows the Lemming?

WebSTAT

Family Blogs - Blog Catalog Blog Directory