Loading ...
Sorry, an error occurred while loading the content.

RE: [SeattleRobotics] Cell phone robotics

Expand Messages
  • Randy M. Dumse
    More proof of utility of cellphones. http://feeds.technologyreview.com/click.phdo?i=6e12fe3437287a9a6 efedaee0c4977cc
    Message 1 of 56 , Sep 1 2:20 PM
    • 0 Attachment
      More proof of utility of cellphones.
      http://feeds.technologyreview.com/click.phdo?i=6e12fe3437287a9a6
      efedaee0c4977cc
    • dpa_io
      Not surprising. Remember all the hoopla surrounding the Aibo? That old folks would prefer them for companionship over a real-live warm-bodied face-licking
      Message 56 of 56 , Dec 14, 2011
      • 0 Attachment
        Not surprising.

        Remember all the hoopla surrounding the Aibo? That old folks would prefer them for companionship over a real-live warm-bodied face-licking puppy? Cause you don't have to feed them, ya know...

        cheers,
        dpa



        --- In SeattleRobotics@yahoogroups.com, "David Buckley" <david@...> wrote:
        >
        > This is a reference to one of the reports that Therapeutic robots don't work as well a people would have us believe.
        > http://www.plasticpals.com/?p=30947
        > "After graduating high school Kentaro took an interest in artificial intelligence. Therapeutic robots like the baby seal Paro were being introduced in hospitals, so he thought maybe a robot could become a kind of artificial friend. But after volunteering in hospitals and interacting with people, he realized that they really want to connect with other people."
        > DAvid
        >
        > ----- Original Message -----
        > From: David Buckley
        > To: SeattleRobotics@yahoogroups.com
        > Sent: Wednesday, October 12, 2011 3:12 PM
        > Subject: Re: [SeattleRobotics] Re: Robots - What are they good for?
        >
        >
        >
        >
        > Those are only the 'yes they work side'. There have been other reportings which indicate everything is not as rosy and ready as the creators would have you believe. They are out there on the web, I have read them, you just need to search for them. You won't find them in any 'official' report or paper which covers all your references. You don't get Ford saying 'actually our cars don't work all that well after 2 years and you ought to buy Toyota' so it is no surprise your references follow the same line.
        >
        > > "In sum, children .......
        > This is just what children do with stuffed toys, so it is not surprising they do the same with robots.
        >
        > > "Interacting with Paro found to improve brain function ......
        > Of course it does, it is a stimulus, Anything is better than nothing. It just doesn't last. Most of the studies are very short term and indicate positive results. They don't report what happens after two or three months.
        >
        > > "Kato said the group plans to put the robot in facilities for testing next year,......
        > So he doesn't know what the long term results are!
        >
        > > Why would these companies invest tremendous amounts of money
        > Because it is their careers and business, and we all believe that somewhere down the line robots will be as useful and have the personalities we dream of.
        > We just are not near there yet.
        >
        > I am realistic and I probably have more first hand and second hand experience of robots interacting with children (and adults) than most of the others put together.
        >
        > My two large walking robots (TecFoots) have had interactions with about 1500 people over the last three years and you can add on a couple of hundred for the smaller robots, this is not demonstrations to people, this is where the people are on their own with the robots, especially the TecFoot walkers where they may have wandered a couple of hundred feet away by themselves.
        > In addition my Cycler robots have performed in front of about 1,500,000 children since 2003.
        > So we know how positively children interact with robots, and if you have never been in a hall with 300+ children and watched the difference between the way they relate to adults and the robots, try it sometime, it is electric, it is like the kids have a switch in their brain somewhere. When a Cycler is presenting a show the kids behave and listen and are quite and participate and what a Cycler 'says' gets across like you wouldn't believe. I keep suggesting to WasteWatch - the Charity which operates the Cyclers - that they write up their experience using Cyclers but it falls on deaf ears. Basically the adults are wary of robots but the children just accept them as things. They know they are not alive just as they know their teddy Bears are not alive, but they accept them almost as equivalent to being alive. At TechnoGames in 2002 where we had the first Cycler, I was talking to a lad of maybe 15 about how Cycler worked and was amazed when about half an hour later he came back and said "Goodbye Cycler".
        > The normal age range of children 'talked to' by Cyclers is younger but in all those children the only bad result is that about once a month a child has to be taken out of the presentation because they 'don't like it', however the teachers say those children have to be taken out of everything because they 'don't like it'.
        >
        > However I am quite sure if a Cycler were to be with the kids every day the positive results would very quickly drop off, just as the average playtime for toys is only about 20 minutes. I think that figure is maybe in a report by Pawson to the British Government under their Robotics initiative of the late 80s. I probably have the papers somewhere but it is pre web time so Google won't find it. Nevertheless if it was Pawson he got his information from the toy industry.
        >
        > So we will get there, we just aren't there yet and in my opinion 'book.readers' are not going to make it, we need to create complex personalities which are continually stimulating. That is what makes pet animals fun and pet rocks a novelty.
        >
        > DAvid
        >
        >
        >
        >
        >
        > ----- Original Message -----
        > From: Weston Turner
        > To: SeattleRobotics@yahoogroups.com
        > Sent: Wednesday, October 12, 2011 4:12 AM
        > Subject: Re: [SeattleRobotics] Re: Robots - What are they good for?
        >
        >
        >
        > David,
        >
        > Here is some research that the MIT Artificial Intelligence Laboratory has done in this area: http://web.mit.edu/sturkle/www/encounterswithkismet.pdf, as you can read, the robots in their study were quite a bit less functional than what I am proposing, and clearly they commanded the children's attention, stimulated their imagination, etc.. Here is an excerpt from the paper to back up that assertion.
        >
        > "In sum, children were tenacious in their efforts to obtain a response from the humanoid robots we studied. They personified Kismet and Cog with extravagant detail, and developed a range of novel strategies for seeing the objects not only as "sort of alive," but also as capable of being friends and companions. Children went to great lengths to maintain the sense that they were in a mutual relationship with these objects, that the objects recognized them and cared about them. Children preferred to see these robots as disabled creatures in need of nurturance rather than of broken machines in need of repair."
        >
        > As far as caretaker and companionship robotics goes, here are some references to back up my assertion that there exists a substantial market for these things, and it is worth mentioning that they actually help people:
        >
        > "Interacting with Paro found to improve brain function through measuring and analyzing the brain waves of elderly patients with cognition disorders. Robot therapy through use of Paro can be used to prevent cognition disorders and it can contribute to improvements in long term care."
        > http://www.aist.go.jp/aist_e/latest_research/2006/20060213/20060213.html
        >
        > A raft-load of research papers linked to by the PARO website: http://www.parorobots.com/whitepapers.asp
        >
        > "Kato said the group plans to put the robot in facilities for testing next year, and begin sales from 2015. The current target price is around ¥6 million ($78,000)."
        > http://www.cio.com.au/article/395754/japan_caretaker_robot_kneels_lift_people_off_floor/
        >
        > "Mitsubishi plans to release Wakamaru sometime next year. The robot is expected to sell for 1 million yen, or $8,300."
        > http://www.wired.com/techbiz/media/news/2003/04/58593
        >
        > I ask a rhetorical question: Why would these companies invest tremendous amounts of money in these ideas over the duration of decades if it was all hype and there was no hope of making any money? Ultimately, companies tend to invest in areas where they will realize a return on their investment (either directly or indirectly).
        >
        > Without citations, what am I to make of your claims? Please provide references when you make assertions that are not common knowledge so that I can better understand your perspective.
        >
        > Best Regards,
        > Weston
        >
        >
        >
        > On Tue, Oct 11, 2011 at 4:10 PM, David Buckley <david@...> wrote:
        >
        >
        >
        > > little more than chat-bot behavior would be sufficient to entertain and intrigue small children.
        > Really?! For how long. Children are not stupid and can quickly determine an absence of intelligence, it's adults who get deluded.
        > Have you any idea what the average total playtime for a toy is? - 20 minutes.
        >
        > > Low maintenance companionship is exactly what elderly or disabled people need (in addition to human companionship of course).
        > > Also, there is a potentially very large market for this type of robot. And I imagine the demand and price for them would be
        > > commensurate to their capabilities, thus a motivation for building ever more functional robots.
        >
        > If you read other than the advertising hype and claims of the makers you will find Paro doesn't work for very long. That is the initial novelty wears off in a short time even with old folk. They can't be bothered with going the deluded route.
        >
        > DAvid
        >
        > ----- Original Message -----
        > From: Weston Turner
        > To: SeattleRobotics@yahoogroups.com
        > Sent: Tuesday, October 11, 2011 9:53 PM
        > Subject: Re: [SeattleRobotics] Re: Robots - What are they good for?
        >
        >
        >
        > Hi Jim, thanks for the thoughtful response to my robot idea.
        >
        > I definitely see the advantage of having the robot control an e-book, and just "read" the text electronically. It turns the problem into a solved speech synthesis problem, and developing some sort of e-book API. I understand that the technical challenges of turning pages in arbitrarily sized books, and doing OCR on arbitrary fonts and book orientations (we cant even expect that children would hold the book still let alone present them to the robot in a consistent manner), are quite a bit greater than your proposed solution. I think the best approach would be an incremental one, if getting the robot to interface with e-books is more readily achievable, then we should build that functionality with an aim toward modularity and code reuse. Then we can work on the harder problems later while leveraging the code-base and lessons learned form having solved the easier ones. I must admit that I have a fondness for paper books. Call it nostalgia, or maybe the dislike of eyestrain caused by reading from electronic screens, but regardless, I feel that paper books are relevant and will remain so for quite some time. There is just something to an author's/publisher's choice of paper format, rich color illustrations, being able to bookmark the heck out of an excellent reference book, etc.. So having the robot read from an actual book would be ideal for me.
        >
        > As far as the requirements for human interaction, little more than chat-bot behavior would be sufficient to entertain and intrigue small children. Good understanding of the semantics of speech is a really hard problem (as in pass the Turing test hard). I think regardless of how much patience children have they would still benefit from repeated exposure to a playful AI agent that tries its best to interact with them in meaningful ways. The kids may get distracted and do other things after a while, but would come back tomorrow to squeeze a little more algorithmic performance out of the robot's "personality" and repeatable behaviors. It would be a study in cause and effect, which is hugely important for kids.
        >
        > Your idea for robots as therapy agents for the disabled and/or elderly is right on the mark. I read about a Japanese group that developed what is essentially an electronic therapy animal for the elderly. They have very simple behaviors, but respond to touch and sound with their own movements and sound output. The benefit of this approach is that as opposed to a live animal, you do not have to feed it, clean up after it, take it for exercise, and they don't die from mistreatment. Low maintenance companionship is exactly what elderly or disabled people need (in addition to human companionship of course). Also, there is a potentially very large market for this type of robot. And I imagine the demand and price for them would be commensurate to their capabilities, thus a motivation for building ever more functional robots.
        >
        > A Soft Spot for Circuitry:
        > http://www.nytimes.com/2010/07/05/science/05robot.html?pagewanted=all
        >
        > The serious truth behind the adorable PARO baby seal-bot:
        > http://www.gizmag.com/paro-robot-baby-seal-companion/13753/
        >
        > Wakamaru Bot at Your Service:
        > http://www.wired.com/techbiz/media/news/2003/04/58593
        >
        > Best Regards,
        > Weston
        >
        >
        >
        >
        > On Tue, Oct 11, 2011 at 11:15 AM, actioncontrols <actioncontrols@...> wrote:
        >
        >
        > Hello Weston,
        >
        > I agree with you comment about the hostile tone of a couple of responders to this thread. We need both dreamers and pragmatists, but they could treat each other with consideration.
        >
        > I'm currently working on a robot using ROS and the Kinect that would do much of what you describe below. I have a question. Is it necessary that the "book" be in paper form. My wife, along with a bunch of others, reads her books on a nook. Could the child in your scenario do the same? A very hard task for a robot (and a young child) is holding a book and turning pages. Of course the robot could OCR a book and read, but - if the presentation method is electronic - the robot already knows what the book says, so the reading part is a non-issue. What do you think?
        >
        > I think the greatest difficulty will be the interaction between the robot and the child. The child will say "I want THAT!" The robot may or may not be able to understand and/or provide whatever that is. Children don't have a lot of patience. Still there are a lot of interactive electronic games out there for kids. I don't know how popular they are. My grandkids are too old for such stuff.
        >
        > With regards to your last sentence about adults interested in being read to, my original audience for the robot I'm working on is the elder mentally challenged group. My robot wants to read to, play movies with, play checkers with - in short, entertain - folks in assisted living situations. Kind of similar to your posit.
        >
        > Keep the ideas flowing! Jim.
        >
        >
        >
        > --- In SeattleRobotics@yahoogroups.com, Weston Turner <wstnturner@> wrote:
        > >
        > > Despite the somewhat hostile tone of the thread at this point, I will posit
        > > an idea for a hobby/home robot. A robot in the form of a stuffed animal that
        > > can be positioned to read out of books and interact with children during
        > > play activities. The robot would recognize basic objects visually and speak
        > > aloud about them, i.e., identify them by name and talk about them in
        > > relation to other objects. The idea is that the robot would function as a
        > > learning aid, helping to develop their reading and language skills. The
        > > justification being that children are insatiable learners and although
        > > parents and teachers often try valiantly to satisfy the curiosity of
        > > children, they do not always have the time or the patience to be completely
        > > attentive. Locomotion is not of primary importance, but the ability to reach
        > > for and grasp nearby objects would be nice. Also the robot would poses
        > > moderate communication skills, be able to recognize words and basic
        > > sentences and change its behavior/topic of conversation based on context.
        > > The ability for the robot to learn would be an important feature, e.g.,
        > > recognize people using facial recognition, learn individuals speech
        > > patterns, etc.
        > >
        > > One advantage of designing a robot for this application is that young
        > > children who do not yet speak or read at a high level would be not be jaded
        > > about the limited abilities of the robot.
        > >
        > > Who knows, adults may have it read to them at bedtime too.
        > >
        > > Weston
        > >
        >
        > > On Tue, Aug 30, 2011 at 1:49 PM, Curran A <a.dushyant@> wrote:
        > >
        > > > **
        >
        > > >
        > > >
        > > > Peter,
        > > >
        > > > It's unfortunate that you think that way and maybe that is the difference.
        > > > When I said students, a Ph.D candidate is also still a student. If you look
        > > > at DLR's research "students", they are not fantasising, they have actual
        > > > prototypes. I had posted some pictures on this site from Automatica 2010,
        > > > you can see the difference. I suggest you look at the DLR's prototypes
        > > > carefully...you will see.
        > > > http://www.dlr.de/rm/en/desktopdefault.aspx/tabid-3758/
        > > >
        > > > Most students there don't distinguish between hobby robots and actually
        > > > applying them in the industry. Maybe that is another difference.
        > > >
        > > > You have only looked at the LWR superficially. Did you know that robot has
        > > > programmable stiffness and damping? Maybe that is another difference. A less
        > > > observant and predisposed mind to baseball and football.
        > > >
        > > > You mention the automotive industry. The difference is quite obvious. I
        > > > would rather drive a BMW than a Ford or GM. I have visited both plants in
        > > > the US and the difference is clearly in the attitude. Maybe that is another
        > > > difference.
        > > >
        > > > Some walk, some people run, and some get sit around and get fat.
        > > >
        > > > Curran
        > > >
        > > >
        > > > --- In SeattleRobotics@yahoogroups.com, "Peter Balch" <peterbalch@>
        >
        > > > wrote:
        > > > >
        > > > > From: Curran A
        > > > > > If you look at where German, Japanese and now Chinese students
        > > > > > are looking at Robotics you can notice the difference.
        > > > >
        > > > > Students always have ambitions way beyond what is achievable in the next
        > > > 10
        > > > > years. Good for them. But don't mistake their fantasies for reality.
        > > > >
        > > > > I suspect that one of the differences between American (or UK) students
        > > > and
        > > > > Chinese (or particularly Indian) students is that that American
        > > > professors
        > > > > keep their students' feet more firmly on the ground because American
        > > > > professors know what they're talking about.
        > > > >
        > > > > "Ah, but a man's reach should exceed his grasp, or what's a heaven for?"
        > > > >
        > > > > > Space Robots for cleaning up Space Debris,
        > > > >
        > > > > Pie in the sky ;)
        > > > >
        > > > > You don't use robots for that, you use fly-paper.
        > > > >
        > > > > And students may be dreaming of it but none are doing it.
        > > > >
        > > > > > Robots for high rise window washing,
        > > > >
        > > > > That's hardly cutting edge. It's a decade old. Basically, it's an attempt
        > > > at
        > > > > locomotion. Nice stuff but hardly hobby robotics. To me, the problems are
        > > >
        > > > > 99% mechanical - robotics problems should be 50/50 software and hardware.
        > > > >
        > > > > > Oil and water pipeline cleaning and repair,
        > > > >
        > > > > Not hobby robotics. Been done for decades.
        > > > >
        > > > > > Robotic Surgery,
        > > > >
        > > > > Show me. Teleoperation, yes. Autonomous robot, no.
        > > > >
        > > > > Teleoperation is not robotics - it just uses some of the same hardware.
        > > > >
        > > > > > Fully automated assembly lines for solar panels, cars, planes,
        > > > > > and boats. Advanced Robotics for production, manufacturing,
        > > > > > and assembly lines for home appliances.
        > > > >
        > > > > Not hobby robotics.
        > > > >
        > > > > I agree that it's where robotics should be going but how much of that
        > > > kind
        > > > > of thing is done in the US? Pretty well none is done in the UK and hasn't
        > > >
        > > > > been since Margaret Thatcher destroyed Britain's industrial base in the
        > > > > 1980s. The Chinese are very interested in industrial robotics because
        > > > that's
        > > > > where the industry is. How many companies running assembly lines in the
        > > > US
        > > > > see themselves doing the same in 20 years? So how many are going to
        > > > invest
        > > > > in a 10-year project? They're all planning to export the jobs to cheaper
        > > > > countries. And those countries will be cheaper than robotic US workers.
        > > > >
        > > > > > Applications of robotics in construction,
        > > > >
        > > > > Not hobby robotics.
        > > > >
        > > > > Presumably, the (building, civil engineering) construction industry can't
        > > > be
        > > > > exported so, yes, that's an opportunity for robotics. However, it's
        > > > > extremely difficult to build a house with a robot. Assembling a car is
        > > > > trivially easy in comparison.
        > > > >
        > > > > If you search for articles, you'll find that universities around the
        > > > world
        > > > > are interested in the problem but, really, none have made any headway.
        > > > >
        > > > > Imagine using a robot on a building site. It must be able to move around
        > > > > using legs or tracks; it must go up stairs or a ladder. It must navigate
        > > > > using vision or LIDAR or whatever. It must recognise objects, say a
        > > > brick,
        > > > > and do a pick-and-place. It must be able to handle difficult materials,
        > > > say,
        > > > > electrical cables or plaster. Etc., etc., etc. All of these elements are
        > > > > actively being researched in universities in the US, Europe, and
        > > > elsewhere.
        > > > > All are unsolved problems. It's no use trying to build a robot for the
        > > > > construction industry until those tasks can be preformed effectively and
        > > > > repeatable in a laboratory, let alone on an uneven building site in the
        > > > wind
        > > > > and rain and dust.
        > > > >
        > > > > Lets walk before we try to run.
        > > > >
        > > > > > Has anyone here seen the video's on KUKA's LWR 4 robot?
        > > > >
        > > > > Looks like a very nice robot but I don't see it doing anything that a
        > > > robot
        > > > > arm couldn't do 15 years ago. It's nicely engineered and so is faster
        > > > than
        > > > > old robots. It has a laser and a good camera connected to a much faster
        > > > > computer than they had back then. Lovely piece of kit - I want one - but
        > > > I
        > > > > don't see it as a huge advance.
        > > > >
        > > > > > People here wanted to go the Moon and Mars and we
        > > > > > are talking about Laundry,Tree Pruning,and Baking?
        > > > >
        > > > > At the time of the moon landings, there was a saying: "a computer can go
        > > > to
        > > > > the moon but it can't go to the corner shop for a newspaper". That's
        > > > still
        > > > > true. Recognising and folding laundry is an extremely difficult problem.
        > > > A
        > > > > million times harder than navigating in space. Pruning trees is probably
        > > > > even harder.
        > > > >
        > > > > Peter
        > > > >
        > > >
        > > >
        > > >
        > >
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.