.comment-link {margin-left:.6em;}


Ivan Kirigin's views on Robotics & Culture: future. perfect. progress.

Tuesday, September 28, 2004

HomeBots, CareBots, JoyBots: GoBots!

USATODAY.com - Domestic bliss through mechanical marvels?:: Divides the fields of service robots into neat little categories. Most of the projections into the future seem plausible, though the specifics are light.

Articles like these are interesting because this is the mode of news for most potential consumers of robotic products. Building an image of helper bots, rather than killer bots from Hollywood, is very positive. Of course, it isn't good to be overly optimistic, as people might not be interested in a technology after initial failures.

Who in their right mind would invest in a company about to make breakthroughs in A.I., given the multiple busts in predictions?
A company doing helper robots, machine learning for drug discovery, intelligent video games, or automated phone systems, well, those are different stories.

Saturday, September 25, 2004

Solution for higher education

The story below about search and rescue cyborg rats, reminded me of this story about a "Robot Scientist" said to equal humans at some tasks.

Here's my idea: combine the robot scientist with the cyborg rat to make a "research and rescue bot". It will find struggling PhD students, do their research for them, and let them graduate on time. Combine this with story-telling Valerie, the roboceptionist, and you can get a thesis defense in the bundle! Heh.

Rescue rats will sniff out buried victims

Slashdot links this article on cyborg rats used in search and rescue. I've been told that one big problem is giving a guarantee that the rats don't begin eating the people (dead or alive) they find. I suppose that the stimulation of pleasure centers of the brain would dominate other such carnal urges.
Each rat has electrodes implanted in three areas of the brain which process odour signals, plan movements and experience rewards.

The scientists stimulated the reward centre to generate feelings of pleasure when the rodent’s nose picked up a whiff of human. In this way, the rats were trained to seek out human odours.
All of this is desirable for a few reasons. The computer-rat brain interface research is also very applicable to computer-human brain interface. I just went to this very interesting talk on the subject. Further, very dexterous robots with high level perception are few and far between. A rat is amazingly mobile and also has an excellent perception suite. Of course, along the way, projects like this could save lives, and that is always wonderful.

Friday, September 24, 2004

In One Building, Outposts for a Globe Full of Toys

The New York Times on The Toy Center, with descriptions of the most popular robotic toys and the business behind them. The same way television has capitaved generations, interaction with intelligent toys will make static pieces of plastic a non-viable business option, for the most part.

Put another way, I'm only interested in buying for my future kids toys that I would be excited to play with :).

Thursday, September 23, 2004

Hardtalk | Robot risk 'is worth it'

BBC interviews Rodney Brooks two years ago, and makes the interview available online to watch.

I was shocked by how aggressive the interviewer is in her questioning. I suppose that's her shtick, but I haven't seen other additions of 'Hardtalk' to know for sure.

Brooks tries again and again to bring up the interesting parts about our future: humans getting robotic replacement parts, the meaning of emotion and consciousness, the potential for exponential growth in the aggregate world intelligence. The interviewer continuously brings back doom-and-gloom scenarios like something out of '2001' or 'The Matrix', trivializing the ideas and excluding what could have been an interesting discussion.

This whole discussion brings to mind the difficult task of convincing people that the robots are not a threat, and that they should be embraced for the potential improvement they could bring to our lives. To their credit, most people haven't been given a realistic scenario for a smooth transition from now to an era with ubiquitous robotic laborers, or the next step to artificial thinkers.

Marshall Brain spends a great deal of time trying to tell people that the transition will not be smooth. I think he fails to recognize the decreasing cost of education with intelligent software that is associated with the advent of semi-intelligent robotic laborers. A more rapidly education populace will, by necessity, offset the elimination of unskilled human labor.

Given the lagging nature of hardware compared to software, I think this is certainly the case.

Robosapien gets Pocket PC brain

This ad hoc throwing together of parts might look rough, but is extremely valuable. Not unlike the work I do, putting a smart perception system on top of a dumb (sense-response/BEAM) mechanical base is probably going to yield the best results in the short term.

Eventually, a more dynamic balancing system would be in order.

Wednesday, September 22, 2004

robot skin

Flexible sensors make robot skin. This could have a number of applications. The first two I imagine are a richer interface between machines and humans and advanced manipulation.

If cheap enough, the machine can understand the precise location and posture of a human. Mentioned in the article are car seats. Imagine a bed which adjusted itself to minimize pressure points.

I should mention a project out of CMU by Chris Atkeson and Daniel Wilson, where he put only a few cheap accelerometers in the floorboard of a house. The algorithm processing these sensors could localize humans in the rooms with remarkable accuracy. The challenge then becomes sensor fusion and system integration, in using this information to boost performance of the entire system. For instance, a human tracker using vision alone would be dwarfed by such a system which had a reasonable seed guess from pressure sensors.

The second application is for rich manipulation. A robot grasping a glass must do so with enough pressure to not drop it, but also enough sensitivity to not break it. I doubt humans use significant higher reasoning in this process, unlike the advantage humans have over computer vision programs. Rather, robots could sense the weight fairly easily, but also the type of surface, and learn how brittle such a surface is.

Future Shock, for America?

TCS: Future Shock, for America? James Pinkerton reviews "Sky Captain and the World of Tomorrow" and "Ghost in the Shell 2: Innocence", while commenting on whether American and Japanese cultures are shunning or embracing the current rapid pace of changes in technology. Good 2nd hand quote:
"Ghost" director Oshii is on a mission to instruct. In an interview with The Washington Post, he observed, "People are very different from animals. We don't accept our original bodies. Humans wear clothes, have earrings and tattoos, do cosmetic surgery, take vitamins. If they are sick, they get organ transplants. And now we have radios, telephones, microphones, watches, computers, microchips outside the body now, but soon we will utilize these machines inside our bodies and then we will be part cyborg. This is inevitable. The process has already begun."
Meanwhile people in America find outsourcing scary, only the latest demon of progress. For more on views of the future and progress, I highly suggest Virginia Postrel's book, "The Future and its Enemies."

self-service woes

The Economist thinks that there are risks with self-service. I would say that the final remark is exactly the forcing function to bring about more intelligence with automation, "Above all, self-service is no substitute for good service."

There is an omnipresent drive to cut costs on the seller's side, so eliminating an expensive, human-powered service is always desirable. Keeping the customer happy is also a grudging desire. These two come together to yield an effort for computer systems to be more usable and more intelligent.

Replacing the human action is not enough: you need to replace the utility of human-interaction. For example, self-scanning automation is OK in supermarkets, because I don't need to deal with a human when in a hurry, but I am not as skilled as the human at that particular task and don’t have the grocery store domain knowledge. Securely identifying the products I'm buying and my personal accounts would allow me to just walk out of the store. Now human check-out laborers are eliminated AND service is better.

For an example of replacing a human poorly, look towards CMU and my least favorite project here: the roboceptionist. What a waste of time.

getting up is easy...not falling is hard

Combat robots wow crowds:: Go and follow the link just to see the videos of the robots falling down. They get up, which is good. They seemed to put a lot more effort into scripts that allow the robot to get up from a fallen position than dynamic balancing algorithms which would keep the robot from falling down in the first place.

Either way, electric motor bipeds are dead in the water. The power requirements are huge and you can't really run without compliant legs or linear actuators. Have you ever seen ASIMO do more than take baby steps? It isn’t just because they are careful with such an expensive bot. For more information, go to this great site on robotic legs.

First Forked Post


I forked this blog to separate my views on technology from my views on politics. Who would have thought that some might be interested in one but not the other?!

I hope that I'll be able to get interesting information about new trends and my views on them in a timely manner. Check back here often!