I'm a big Blade Runner fan since it was released in 1982. Read the book by Philip K. Dick in the early 80s when I was starting high school. Director's Cut on DVD. Rabbit hole on filming locations, set decoration, and obscure cast members. So Rands' stirring homage to typography in Blade Runner's opening title cards caught my attention:
"Unlike our functional fixed-width typefaces, the role of typography in this title sequence is partly functional — to set the story — but the primary purpose is establishing mood. Director Ridley Scott expertly drops us into the middle of a dystopian future where we’ve enslaved the robots and, duh, they are rebelling."
Well, yes. But dropping us into the story with descriptive text and careful typography is not the only thing that's going on.
The audio consists of two parts: the minimalist score by Vangelis, and sound effects that include distant booms, twinkling computer notes, and a crescendo that builds to loud explosions and the whoosh of futuristic vehicles. It effectively immerses viewers into this incredible world and story.
This is very different than Star Wars' title sequence 5 years prior which also had an amazing score (by John Williams), but was firmly anchored in the tradition of Holst and contemporary Hollywood orchestral scores.
Keep in mind that a typical viewer entering a theater in 1982 to watch Blade Runner was stepping out of Chevy Nova with an 8-track cassette players loaded with Juice Newton, The Charlie Daniels Band, and Kool & The Gang. The typography, the score, the sound effects, and the images on screen would have been mind-blowing.
Every once in a while an invention comes along that changes the way people live their lives. Computers, mobile phones, television, washing machines and cars are but a few examples. I'd like to add to the list robot vacuum cleaners ... specifically, the iRobot Roomba Model 530 that we got last month. It has changed the way we do housework, and has really freed up a lot of time for us to do other things.
Why buy a robot vacuum cleaner?
A lack of time was the reason we got the robot. I have been cranking away on starting a new company producing a classifieds app and simply don't have the bandwidth to vacuum the house every week, and neither do other family members. I had heard about the Roomba for years but never considered buying one, owing to the expense and the fact that one of us usually had enough time to vacuum every week using a 15-year-old Sharp upright model. But when we both began new careers, the problem suddenly presented itself. Around the same time, I saw a brief side-by-side review of three robotic vacuum cleaners by America's Test Kitchen, and noted that the prices were coming down. I did some additional online research, and decided on the Roomba.
We put it to work right away. The results were impressive -- it cleaned up our bedroom, even getting underneath dressers and night tables, in less than 30 minutes. It got into corners and redid each spot multiple times, resulting in lots of debris in the dirt and dust traps. Next was the bathroom and hallway, and then the kids' rooms, and finally the downstairs area, including the kitchen. We have mostly hardwood floors throughout the house, which really suits the Roomba, although we've noticed it takes care of medium-sized rugs as well.
The greatest thing about the Roomba is it's pretty much fire-and-forget. You spend a minute or two prepping the rooms to be cleaned (moving toys out of the way, putting up loose electrical cords, shutting doors where you don't want the robot to go) press the button on the top, and let the robot do its thing. It seems to be moving randomly, but there is a method to its cleaning path, driven by trigonometry and some sophisticated algorithms. It's actually a great feeling being able to work in one part of the house while hearing the Roomba whirring away. Yesterday we even went out shopping and left the Roomba to clean the whole downstairs. When we came back the floor was clean and the robot had automatically returned to its charging station.
Roomba vs. vacuum cleaners: What are the drawbacks?
Are there drawbacks? Sure. One thing that some people will hate is having to clean the robot. Unlike a standard vacuum, which only requires changing bags or shaking out a dirt trap every now and then, the Roomba requires regular maintenance that contains about a half-dozen distinct tasks. You have to manually empty the dirt and dust traps (typically after a large room or the entire floor) and periodically unscrew the brush and rollers to remove tangled hair and dust. It doesn't take long, but it's dirty (with pets it would be even worse). It is also absolutely necessary to keep all of the parts moving.
There are some other minor irritants as well. It can get stuck on high barriers or under furniture that has hanging springs or rods, like our sofa bed. Cords and strings can mess it up. One time it brought down a heavy plug that I had placed on a table. It landed on top of the robot and shut it down, but there was no permanent damage (the robot is actually pretty sturdy). Another time I had to shut it down just before the end of a long string of yarn got swiped into its brushes. You can buy electronic barriers to keep it out of certain areas, but I've found the range on those are only about 10 feet, which means it can get around them if you're trying to block off part of a big room. As an alternative, I sometimes lay chairs on their sides to stop the robot from passing certain areas.
But overall, the Roomba is a remarkable invention. It's a huge timesaver. I estimate it saves us about 5-10 hours of labor per month. The cost is comparable with many upright models -- less than $300 when we bought ours. And while the maintenance is a minor pain, it's worth it.
Demonstrated today in our Media Ventures class: Autom. The speaker is Cory Kidd, formerly of the Media Lab, now an entrepreneur based in Hong Kong, where software talent is cheaper and local government incentives are available (including free rent) to set up his business.
Other posts about my MIT Sloan Fellows experience:
This was fun. A bunch of the students in my program went to MIT's building E-62 on Thursday night, lured by the promise of free food and beer, and what else did we find? Cool MIT club and classroom demos as well! The following interview that I conducted at the event is from MIT's "D-Lab", which is described as "Development through Dialogue, Design, and Dissemination":
The D-Lab demo is specifically from the "Cycle Ventures" project, which, according to the project website:
... Explores bicycle technology to provide human power for an increasing array of other purposes including water pumping, grain grinding and transport of loads in underserved communities with the aim to offer economic opportunity. The course provides a historical prespective on innovation in bicycle technology, reviews its engineering mechanisms and draws on the ubiquity of bikes to present them as innovative tools to foster socio-economic development.
(Footage shot an edited on an iPod touch using ReelDirector)
Learning about a robot opera was the highlight of a second trip to the MIT Media Lab (see the other robot videos and interviews I shot on the first trip in the link at the bottom of the page). During my visit, I talked to some of the students and staff who are working on the production of "Death and the Powers," including my old friend and bandmate Bob Hsiung (pictured). In my interview with Bob, you'll get a glimpse of the storyline, as well as a close-up look technologies that powered the show, such as OLPCs running Linux and a joystick:
Here is a brief, professionally produced snippet from the opera itself:
One thing I've wanted to do since arriving at MIT at the beginning of the summer is to see the MIT Media Lab. I mean, I'm a media guy interested in geeky technologies, so it should be one of the first stops, right?
Well, I never got around to going into the Media Lab building, until today. I heard from two non-Sloan friends of mine -- a current PhD candidate studying new ways to do gaming AIs, and a staff member who helped make the world's first robot opera -- that there was a Media Lab open house, in which most of the current projects (and a few past ones, too) would be on display with researchers on hand to explain them.
Firehose time. I spent a few hours visiting the offices and labs, drinking in projects that ranged from anthropomorphic robots (such as Lexi, inset) to a kids-oriented programming language to "tangible media.
Unfortunately, there was no way to see it all. But I was able to shoot a few highlights and conduct some interviews using my iPod touch. There are three videos:
MIT Media Lab Open House Video #1:Scratch + Legos = Cool (interview with researcher Sayamindu Dasgupta from 00:00 to 08:00, for the final two minutes I spend a few minutes looking at other Media Lab projects)
MIT Media Lab Open House Video #2: Social robots! (interview with Personal Robots Group researchers, including Kenton Williams, discussing Nexi and human-social interaction) (note: audio is slightly off-sync)
MIT Media Lab Open House Video #3: An army of robots? (I was unable to find the identity of the PhD student on the Media Lab/Personal Robotics Group website, but I am wondering if that's because of the military sponsor?)
The robots of Newton are here -- an army of green Waste Management trucks which pick up the blue trash barrels and green recycling barrels every Monday using a giant robotic arm (or mechanical arm?). They started in our neighborhood at the beginning of October. Here's what it looks like:
I showed this to a Japanese friend (a resident of the next town over, Waltham) who said, "this is in the United States?" Indeed it is! While in Japan robots are used for auto manufacturing and bizarre educational trials, here in Newton our robot overlords are starting with simple trash collection. The company that operates the service is Waste Management, which won the contract in our city based on a trial in one neighborhood and the claim that it will save lots of money, thanks to reduced manpower needs. Indeed, the collection is very quick, as you can see from the video.
However, what the city of Newton failed to take into account is the fact that the barrels are extremely heavy -- the barrels hold 64 gallons and are made of heavy-duty plastic, and must weigh upwards of 40 pounds with a typical load. Many local senior citizens are already complaining, and I can only imagine what it will be like when we have to wrangle these monsters down steps and over three-foot-high snowbanks in February.
My letter to the local newspaper about robots. Specifically, the new robotic arm-equipped garbage trucks that will start collecting trash from our neighborhood next week:
In the article about the new trash and recycling system that relies on robot arms to pick up and dump the barrels’ contents into trucks, you were told by the city’s spokesman that contractors currently cannot collect trash in the new blue and green barrels “because they would be too heavy to lift by hand.”
If people who do this for a living have trouble lifting the barrels, what about ordinary people? Are residents going to get robots to help manhandle the barrels up and down stairs, through the snow and onto the curb?
This automated trash pickup system was not well thought out, as evidenced by this situation, and the fact that it wasn’t put out to bid (aldermen approved Mayor Cohen’s plan, 23-1, last April). I hope the new mayor, aldermen and city officials can deal with “upgrades” like this with more realistic planning, including assessing a range of use-case scenarios of how city residents will be affected, including senior citizens, people without garages, apartment and condo dwellers, etc.
For some reason, the Newton Tab didn't print it, but did publish the letter online, along with another letter on the same topic that did not make it into the print version.
When the trucks finally did come to Newton, here's what they looked like in action:
(See updates at bottom of post) It's campaign season in my corner of the 'burbs, and the mayoral candidates are pulling out all the stops to make that all-important personal impression upon voters like me.
But I have to hand it to one of the hopefuls, Setti Warren. Not only is he the only candidate to ring my doorbell to personally introduce himself and his platform, he is also the only one who is apparently fully ambidextrous, as shown in the two postcards I received from him today:
I jest, of course. It looks like Setti's effort to craft a personal response was slightly botched by a pair of young campaign helpers working from the same distribution list and almost the same script (one asked for support in September, and the other in November).
Wouldn't it have been much easier and authentic to simply print out a postcard and have him sign them in his actual hand? Or, if a personal touch is absolutely essential, why not use technology to solve the problem, like this:
This machine may seem silly, but I have seen similar devices used in the higher education sector to solicit alumni. It's not a stretch to imagine them being used in political campaigns in which candidates simply don't have the bandwidth to personally sign hundreds or even thousands of pieces of election-related communications.
Our Founding Fathers surely never foresaw robotic pens and robocalls making an impact on our political system, but then again, they surely didn't anticipate other technological innovations such as whistlestop tours, loudspeakers, and television.
8/4/09 update and note to readers: This post is not intended as a criticism of Setti's platform or other campaign practices -- as I pointed out yesterday, Setti is the only mayoral candidate to visit my house in person so far. He actually spent about five minutes on my porch answering a question I asked about city hall. I also have a great deal of respect for his service in U.S. Navy Reserves, which required him to take leave of his family and life in Newton a few years back when he was called to active duty.
So, if you want to leave a comment, please be polite. And remember that in the grand scheme of Newton politics, having campaign staffers write and sign personal thank-you notes ranks very low in terms of importance to what I consider to be real problems impacting the city -- such as terrible long-term fiscal planning, crowded public schools, and the unmonitored use of scores of city vehicles by civil servants who live outside of Newton.
Second 8/4/2009 update: The Newton Tab explained what happened, after talking with Setti. He did sign one of the cards, apparently.
Update #3, 8/4/2009: Matt Frank, commenting over on the Universal Hub thread about this incident, gives a name to the device in the video above and confirms that it is indeed used by politicians:
The machine Ian refers to is in use by the big leagues already. Many campaigns just do the elevated print but down in DC it is not all that uncommon for Congressmen to have an "autopen" that is there for signing items and response letters to people back home.
Matt also describes an amusing faux pas involving an autopen and former NYC Mayor Rudy Giuliani.
File under dead-end educational technologies. It's cute, but there's no way that robots like this one will be an effective way to teach kids anytime soon. Check the video below, which shows the robot "Saya" reading a prerecorded text (why not show them a video?) right before the kids come up and poke and fiddle with its face.
The SFGate article claims Saya will have some ability to scold misbehaving kids and take roll call. Regardless, I can only imagine the mischief that would result if this were tried in most normal classrooms.
For the past two or three decades, there's been a regular stream of stories about prototype Japanese robots delivering the mail, bringing medicines to sick people, serving as pets, etc. Despite the breathless pronouncements from journalists about such robots becoming commonplace in a few years, Nothing ever seems to come from the experiments -- they're too expensive, the AI is limited, and certain technologies (especially visual and speech recognition) aren't ready for prime time.
On the other hand, if educators are looking for ways to engage young children, there are plenty of options, ranging from other types of technologies (video, software, online, and even 3D games) to old-school concepts such as storytelling and time capsules.
Disclosures: I am a board member of Immersive Education, which seeks to promote the use of virtual worlds in online education. I am also a personal friend of the owner of Heritage Time Capsules.
Last night I finished reading Code Name: Ginger, a history of the development of the Segway transporter. It was a fascinating story -- I've read many accounts of software startups and product development tales, but this is the first hardware-oriented book I've read since Arthur Hailey's Wheels.
The focus of the book is not just the product. This is about the remarkable inventor and entrepreneur, Dean Kamen. The author of the book, Steve Kemper, manages to illustrate his character very well. After reading it, it's almost as if I know Kamen -- and, honestly speaking, while I would love to talk with him I don't think I could ever work with him after reading about the way he managed the Segway project.
But Kamen has done some remarkable things. While part II of the Segway story would be its apparent commercial flop (compared to the sales projections in 2001, at least) Kamen has since launched an important project that has the potential to help a huge population of severely disabled people. A few weeks ago, I saw this interview from the Wall Street Journal's All Things D conference in May, which featured Kamen discussing a special project to design a thought-controlled robotic arm for amputees:
Code Name: Ginger is a great introduction to the world of modern inventions, entrepreneurship, and engineering management. It's also an interesting case study of the way West Coast investors interact with the projects they take interest in. The sections about Jeff Bezos, John Doerr, and Steve Jobs getting involved with the Segway project and Kamen's management team are priceless.
At EmTech '08 yesterday, I heard Microsoft's research and strategy chief talk about the company's cutting-edge technologies. Craig Mundie spent a fair amount of time discussing cloud computing and advances in programming, but I was particularly interested in his discussions of 3D-related technologies. For years, I have been writing about virtual worlds and 3D tools, and observing how companies both big and small have been developing products and trying different business models. Google, Intel, Sony, Linden Lab, and Sun Microsystems have been very visible in this sector, but Microsoft has been a bit of an enigma ... until yesterday.
Mundie basically said that the user-generated virtual worlds which are popular now -- Second Life and the like -- are not scalable in the sense that lots of people will be interested in creating avatars and virtual objects. Rather, he thinks killer apps will be centered around the "Spatial Web," a virtual representation of the real world that uses the Internet, cloud computing, and new interfaces. Key to the creation of the Spatial Web will be Photosynth, a really neat modeling tool that I have covered here on my blog. Photosynth figured heavily in Mundie's demonstration at MIT yesterday. Here's what he showed in his demo:
In one brief demo he showed how a small, camera-equipped robot could be used to model a large room. In another, he showed a 3D model of a commercial district in Seattle that had been created with Photosynth, and demonstrated how a virtual visitor could come to the district using the Internet, enter an art shop in the area, and examine and buy a virtual sculpture that had also been "photosynthed" by the shop clerks or the artist.
You can read the rest of my Industry Standard article here, and I have a brief video from Mundie's demo below, showing how his staff created a 3D model of a real place (a room at MIT) using a camera-equipped robot.
A few other quotes from Mundie's presentation, which I did not include in the TIS article:
How does Microsoft view the future?
"What is the killer application and what is the user engagement model that drives adoption?"
On cloud computing:
"More and more, computing will be model based, rather than discrete applications."
On user interfaces:
Microsoft is looking for the "NUI to replace the GUI"
Today I visited the MIT Museum after seeing a listing of current exhibits in the Boston Phoenix. It looked interesting, and I was in Cambridge anyway (having Japanese lunch in Porter Square) so I decided to make the long walk down Mass. Ave and check out the museum.
It was small, but neat. When it comes to technology, I am a software guy at heart, so I most liked the AI exhibit. I didn't realize MIT has had a dedicated AI and robotics program since the 1950s, and it was fascinating seeing some of the displays from those early experiments -- in particular, a robotic arm designed by Marvin Minsky in the late 1960s and an early LISP hardware device which was designed to replace the use of the programming language on mainframes. This latter technology eventually resulted in two startups being formed, but both went bankrupt in the 1980s and early 1990s as personal computers became powerful enough to use for LISP programming. I was reminded of the fate of Silicon Graphics, which produced expensive, powerful hardware to handle graphics design and rendering in the 1980s and 1990s. It also went out of business as cheaper PC-based solutions became possible. Moore's Law meets the law of the marketplace ...
The museum has other sections devoted to mechanical engineering, nautical engineering, the history of MIT, and even a very neat mechanized art section. A bunch of displays are hands-on, which can be fun, but when I tried out one of them -- a 3D simulator used to train remotely operated submersibles -- I was frustrated by the interface. "Bad UI," as I like to say when confronted by poorly designed or explained interfaces. RTFM? I would, but there wasn't any.
If you're in that part of Cambridge, the MIT Museum is an interesting diversion for one or two hours. It costs $7 for adults, less for students and seniors, and it's free for people with MIT ID cards.
Photo: Kismet, a 1990s AI/robotics project now on display at the MIT Museum.
I don't post too often about robotics on this blog, but I thought I would share this video. It shows Boston Dynamics "BigDog" robot going through some very impressive terrain tests. Watch the whole video, because some of the best tests take place at the end. Also, be sure to turn down your volume before you click the "play" triangle, because while this thing looks cool, it sounds like a weed whacker going full tilt:
Some technical background was provided by a Slashdot commenter. The robot was funded by DARPA, a government agency that has funded scores of interesting science and technology projects over the past four decades, including early Internet research. The obvious applications of this robot include search, rescue, and resupply in difficult terrain, but some imaginative commenters on Slashdot predict military applications as well. Frankly, the shock factor and noise is enough to terrify practically anyone standing in its path.
What do you get when you mix robots, roaches, and the insects' socialization instincts? According to the New York Times, Lemming-like behavior that encourages the bugs to deny safety-related instincts:
When 16 cockroaches were placed in the arena, they naturally gravitated toward the darker disc, following what the researchers believe is an internal calculation of the amount of light and the number of other roaches, finding comfort in company.
Dr. Halloy then replaced four of the cockroaches with four robots equipped with sensors to measure light and the proximity of other robots. When the robots emulated the real roaches, the group continued to seek the dark and crowded place.
When the four robotic roaches were reprogrammed to prefer the lighter disc, however, the real roaches followed them about 60 percent of the time, in essence deferring their own judgment as the preference grew more popular. (The other 40 percent of the time, the robotic roaches succumbed to peer pressure and headed for the darkest place.)
I am trying to think up some practical applications here ... pest control services that more effectively entice roaches to come out from under stoves and refrigerators?
The New York Timeshas a report on a new graphics technology from Image Metrics that maps real, human-generated facial expressions onto an avatar. The technology is very accurate in capturing movements and even transferring the facial "soul" of the actor to the computer-generated face, says the Times' Sharon Waxman.
It’s not just that the virtual woman looks happy when the actress looks happy or relieved when the actress looks relieved. It’s that the virtual woman actually seems to have adopted the actress’s personality, resembling her in ways that go beyond pursed lips or knitted brow. The avatar seems to possess something more subtle, more ineffable, something that seems to go beneath the skin. And it’s more than a little bit creepy.
While using humans to model computer graphics is old hat (see an example from a 2004 car commercial), it traditionally has been used to capture strong body movements, rather than slight facial expressions. And older "hand-drawn" computer graphics can't capture eye movements, which Image Metrics' technology is able to do.
Potential uses of the technolgy, says the article, are re-animating dead actors, or letting living actors use a younger version of themselves in films. The technology can also save time and money -- it will be used in an animated full-length movie called Foodfight!, and will supposedly take just months to make instead of the years required for typical Pixar or Disney fare.
The technology was created by British experts who were working on image analysis, which was first used to better understand spinal X-Rays, says the New York Times article.
The Image Metrics website notes that the technology helped make the film The Polar Express, which, as noted in a previous Web3DNews article, some people felt to be "creepy."
, a history of the development of the Segway transporter. It was a fascinating story -- I've read many accounts of software startups and product development tales, but this is the first hardware-oriented book I've read since Arthur Hailey's 