Robots R Us

Back to Contents of Issue: February 2002


Entertainment robots are just the beginning. Japan's robot pioneers are planning to bring the bots out of the playroom.

by Sam Joseph

THE ROBOTS ARE HERE There was a slight acrid smell. It appeared to be coming from solder fumes, rising from groups of young high-school boys who sat huddled around large collections of electronics, toolkits, and sundry pieces of machinery. I was in a large hall filled with disc-shaped arenas, roughly five feet across. Standing in gradually wider circles around the arenas were competitors, mechanics, judges and, further away, the public and esteemed members of the press. The tension and focus of the participants was visible on the fresh faces of the young men as they carried their creations to the arenas. Here the machines would battle it out for first place in the 13th All Japan Robot Sumo Tournament, the largest robotic sporting event in the world.

JAPAN'S RELATIONSHIP WITH ROBOTS, as evidenced by things like Robot Sumo, is unique and can be traced back to the 12th century writings Konjaku Monogatari Shu, which include a description of a mechanical irrigation doll that would automatically pour water into the paddy fields during dry periods of the year. During the 17th and 18th centuries, robot-like puppets called karakuri-ningyo were developed and used for entertainment purposes in folk festivals, some of which are still in use today.

Gakutensoku, the first modern Japanese robot, developed in 1927 using Western technology, could smile, flutter his eyes and write. Yoshinori Haga, deputy general manager of the Bandai Robots division, described Gakutensoku as, "a sort of early Japanese animatronics." He suggested that early work like Gakutensoku fed the boom in robot manga, or comic books, after the Second World War.

Two archetypal manga series began in the 1950s: Osamu Tezuka's Atom Boy (known overseas as Astro Boy), and Mitsuteru Yokoyama's Ironman #28. Atom Boy, constructed by a scientist after the death of his son, could think and showed human-like emotions as he battled to save the world from calamity, whereas Ironman #28 was a huge remote-controlled robot, with no will of his own. This division is still reflected in the two categories of the Robot Sumo Tournament " autonomous and remote-controlled.

Pet Robots

Approaching the iconic status of Atom Boy is Sony's robot dog, Aibo. Sony's release of a walking robot pooch in late 1999 stunned the world and ushered in a pet robot boom, with multiple companies following with their own robotic pals. The Aibo has always been a pedigree mutt " the basic unit costs $1,500. The cut-down Aibos (see "Christmas Shopping," page 35, December 2001 issue) are slightly cheaper, but robot purists like myself prefer the Terminator look of the newest ERS-220 model. Still not having had the chance to play with one, I persuaded Aibo's international marketing manager Kazuhiko Akiyama to show me the Aibo II, released in October 2000. Aibo II had extra features over the original model such as being able to take photos, hook up to a wireless LAN and the ability to read email aloud.

Akiyama says, "In Japan robots are considered to be friendly, an image that has been created through popular characters like Atom Boy. In Japan everyone will buy an autonomous robot for their children, something that would not happen in the West. Remember, it's difficult to have a real pet in Tokyo."

Akiyama told me that user feedback from the original model suggested the main things people wanted were for Aibo to be able to self-charge (current battery life is under an hour), and to use voice/face recognition abilities. Aibo II can recognize up to 50 words and managed to chase the pink ball I threw at it well enough to kick it around. The self-charging ability is still not present, but it is an interesting challenge, since an ability to charge its own batteries would potentially allow Aibo to run indefinitely and allow it to meet several of the standard criteria for a "living" system. "Our goal is to have an Aibo in every household in 10 years," said Akiyama. Which leads to the big question " in 2011 will you have to fight off a pack of stray Aibos to use your electrical sockets?



One in Every Home

It's not just Sony that's setting its sights on ubiquitous household robots. Yuji Hatano of Honda's publicity department told me that, "Honda would like to see one Asimo in every home." Asimo is Honda's latest bipedal robot, with more advanced walking ability than the previous P3 model, which had to pause while turning. Asimo can walk round corners without breaking stride, as well as backwards and sideways. Not really a pet robot, Asimo's smaller and cuter than previous bipedal robots, about the size of an 11-year-old child. Honda is renting out Asimo rather than selling it and seems unwilling to publish an hourly rate " perhaps for fear of people going into shock.

At the opposite end of the scale, a company that is well on the way to achieving the one-in-every-home target is Sega Toys, which cashed in on the robot pet boom with the affordable Kokorobo series. Sega's Poochi robot dog, retailing at \3,000, was clearly much more likely to meet most parents' financial constraints. Poochi, released last year, was tamagotchi (remember them?) in robot form " a plastic toy that would sing, dance or sulk, depending on whether you stroked it or knocked it around. The market for entertainment robots is huge: Sega Toys has shifted 15 million of its robotic friends.

The different robots have a variety of sensors, including light, sound and touch that allow them a limited sense of the environment. Mikako Harada of Sega's Media Department introduced me to the new C-Bot series, which intends to build on the success of the robot pets. C-Bot has an animated display rather than a fixed face, and so can display a range of expressions, including text. Harada switched on the C-Bot and it started to flex its arms, Japanese characters scrolling across its features. "Pleased to meet you," it texted, "From today we'll be living together; please give me your favorable consideration." It loses something in translation, since it was definitely much cuter in Japanese. Either way, I was beginning to wonder if I was ready for this sort of commitment, settling down with a young robot and so on. C-Bot then sneezed a couple of times (at least that's what Harada told me he was doing), and I began to imagine the sort of relationship a young child would form with a toy that could talk.

Due to a slight slip entering my name, C-Bot referred to me as Saato-chan for the rest of the interview, but the settings can be easily adjusted. Entering the time provides a framework for biorhythms, so C-Bot will sleep at night and when people are out during the day. It will say goodbye when you leave, greet you as you come home, and ask all-important questions such as, "Do you really like me?" to which the user can input answers by using the two head touch sensors as yes/no buttons. As with all of the emotional relationships in our lives, answering such questions in the negative will lead to problems, while affirmations of affection will be rewarded with smiles, arm waving, and the traditional Japanese comic-book response of eyes turning heart shaped.

"The robo-toys are not really robots," says Harada, "they are toys " real robots do dangerous things such as working in environments unfit for humans. In 10 to 15 years maybe these more complex robots will have become commonplace and then we will start to see toys that are really robots."

Communication Robots

Tomy also released a canine robopal, called DOG.com, not long after the first Aibo was released. Kazuo Watanabe of the Tomy Public Relations Department told me, "We never really intended to produce robots, but Furby (marketed by Tomy in Japan) was such a huge hit here we saw a big market for communicative toys." Watanabe indicated that the most difficult thing was to create something that would have interesting behavior, but would still be affordable. Design was also a crucial factor " whether people liked DOG.com or not seemed to depend a lot on its color. The key component in the success of Furby and DOG.com was taken to be the communication features; both of the systems behave like robot tamagotchi, and respond to how they are treated.

Focusing on communication, Tomy's latest product has no mobility at all. The Memoni, a hand-held chatting robot, went on sale in October last year. It can pick out a couple of unknown words from a given sentence, such as, "Today I will meet Sam," and can ask, "What is Sam?" The user can then say "Sam is a friend" and it can remember that fact. Gradually, it builds up links between the different words, which can then be used to lead the conversation. So subsequently, when I say "I'm meeting Sam today," it can reply, "Ah, Sam is a friend," or "You met Sam last week." The resulting experience is somewhat like talking to a child that might unpredictably change the topic.

This is perhaps the first time that this kind of learning technology has been put into a non-PC-based product. "Getting the system to operate at a fast enough speed was a big problem," explained Watanabe. "There is a trade-off between conversation speed and conversation breadth. If the number of unknown words is too high, conversation becomes impractical." The Memoni comes with a 20,000-word vocabulary and apparently can be taught a further 3,000. At \19,000 it isn't the cheapest robot around. Nonetheless, Tomy is aiming to sell 50,000 units, to "ladies who enjoy talking." Tomy is still being secretive about sales figures, but Watanabe believes that the market for "pet robots," or rather "communication toys," can only get bigger.



Furry Robots 

The potential size of this market is not being lost on other companies, including Omron, which released a robotic cat, going by the charming name of NeCoRo in October last year. Starting with a limited run of 5,000 units, NeCoRo is an experimental product and the first robotic pet to come on the market with synthetic fur. It can't walk like other robotic pets, but makes up for this in the brains department " it's able to distinguish between its name being called by multiple different owners, perhaps different children in the same family. This allows it to create a personalized 'playful' behavior for each of its owners, so that it can greet each one with that special action or play routine. It can track moving objects and will turn to look at things, reaching out to paw at feathers and other long, thin implements that one might wave at it.

A marketing gimmick perhaps, but NeCoRo is a great showcase for some of the latest sensors developed by Omron. Although the cat won't fool anyone into believing that it is anything other than synthetic, some of its responses and actions are surprisingly lifelike " a testament to Omron's sensors and actuators. Will we be seeing more lifelike robot pets? We'll have to wait and see how many NeCoRos they manage to sell.

 
Applied AI Systems (AAIS)

Small-scale robot companies appear few and far between, but Applied AI Systems is one of them. The main AAIS R&D Center is in Ottawa, Canada whereas much of the other business functions are run from Japan. AAIS is strongly linked to Japan through its Japanese president, Takashi Gomi. I met Dr. Gomi at his house in Shinkoiwa beside Tokyo's Arakawa river. His rooms were filled to overflowing with videocassettes, which I discovered were a library of short films, each a demonstration of a different AAIS robot species.

AAIS's main operation involves selling research robots to universities. Gomi founded the company in 1983 and by the late 1980s there were around 20 AI/robot companies in Canada, but AAIS is the only one to have survived through to the 21st century. Gomi-san suggests that the secret lies in his corporate philosophy: "We never wrote a business plan, assuming that money would be an 'emergent property' " the rule is 'don't plan for money.'" Gomi was forthright in his explanation of AAIS business practices. "We don't hire people for experience or background, we hire people based on the type of person they are. We ask them two questions: 'Are you honest?' and, "are you hardworking?'" Presumably, a positive answer to both questions is a prerequisite for employment. Apparently most people don't last more than five weeks.

So what inspires Gomi to build robots? He comes straight to the point: "Why are we building robots? Because we are trying to understand consciousness." This comment underlies an ongoing conflict between classical roboticists, who tend to work on top-down control systems, and the rebellious behavior-based roboticists, an approach promoted by MIT professor Rodney Brooks who has heavily influenced Gomi and AAIS.

Behavior-based robots work using a bottom-up approach, small component behaviors being combined into gradually more and more complex abilities. It's like starting off by building a robot's reflexes rather than focusing on its algebra-solving abilities.

Gomi showed me a series of demonstration videos highlighting the advantages of this approach. "The main difference with AAIS robots is that they are not constantly checking their position, as other robots do." One video involved two robots trying to get through the same gap from opposite directions. To solve the problem mathematically is a nightmare, but behavior-based robots let one another past by moving around reflexively. Using this approach, AAIS is gradually building more robust robots, some of which are now capable of carrying up to 100kg over rugged terrain. AAIS robots have appeared on television in Japan and Canada, and have given eye-catching demonstrations, such as a robot wheelchair (with a human passenger) adjusting its trajectory to steer around a car that unexpectedly pulls out.

 

Behavior-Based Robots 

One Japanese company that seems to have no problem with the bottom-up approach is Bandai. Like Sony, Bandai is targeting the higher end of the entertainment robotics market. I found the Bandai research labs nestled behind the blinking lights of Tokyo's denden town (electrical district), Akihabara. To complete the Bladerunner mood, it was raining heavily and it took me a while to dry out as I began to talk to Yoshinori Haga, the deputy general manager of Bandai's Strategy & Development Department.

Bandai's robotic kitten, BN1 has sold 3,000 units at \50,000 each as of October last year. It isn't sold in shops, only over the Internet and it is aimed at robot enthusiasts. Haga told me that Bandai was not really aiming at creating a pet as such: "A real pet is still probably more fun than a BN1."

BN1 has a similar range of behaviors to some of the cheaper robo-pets but has 360-degree infrared sensors, allowing it to sense nearby objects. It expresses emotions through the eyes, which can display different symbols, something copied by Sega's C-Bot and Tomy's Memoni. 


It can detect that it is being stroked or hit through a combination of light and sound sensors, and can right itself from an upside-down position. BN1is also a completely programmable robot " new programs can be downloaded via PC. I was fascinated to hear that the control system is a behavior-based architecture (Gomi would approve). Different behaviors can be given different priorities and linked together with variable degrees of association, somewhat like a neural net. Bandai is currently working on making this a learnable, or evolving system that will behave more effectively over time.

It seems the behavior-based approach works well for a toy robot. "The BN1 is for fun, not for work," pointed out Haga. "If confused it tries out other sets of behaviors and it doesn't matter if it fails in a particular task as long as it is enjoyable to watch it fail." Bandai would like to include some reasoning ability. Aibo can recognize a pink ball and follow it while in view, but take it away and it cannot go and look for it; it has no concept of the ball, or that it might be hidden. Aibo's behavior is linked to the presence of the pink ball stimulus " remove the stimulus and the behavior goes away. Haga told me he would like to be able to include higher goals like, "I want to find the ball" in pet robots. The key is linking the high-level world model to the low-level reflex behaviors that support basic actions.

It seemed that I had found a really passionate roboticist amongst the press officers I had been talking to " someone actually influenced by Atom Boy and IronMan #28. I asked Mr. Haga why he got involved in robotics: "I was really entranced by the idea that you could make things move, by instructing them to do so." Haga remembers a primary school science fair at which a punch card robot was displayed. He remembers being inspired by the idea that you could increase the number of actions possible for the robot and then instruct it in more complex ways. The idea of making robots that can help humans still inspires him: "It's much easier to build robots now than it was. When I was growing up there were no microchips or graphical user interfaces, just transistors and wires. One of the big differences is the change in processor speed. The BN1 has a 16-bit processor that 20 years ago would have only been found in a top of the line business computer."



Snake Robots

I found some of the next generation of robot engineers at the Tokyo Institute of Technology (Titech). Professor Shigeo Hirose had invited me to an end-of-year display of robots by third-year students from his mechanical engineering course. I only caught the end of the event, but managed to see tea-drinking robots, calligraphy robots and pool-playing robots. Hirose is famous for creating the first operational snake robot that propels itself forward purely by undulation. The Hirose-Yoneda research group is working with Mitsubishi on special robots for cleaning up after nuclear accidents, but has also produced wall-climbing robots, roller-skating robots, three-legged robots " the list goes on. "Generally most companies have doubts about the practicality of robots," explained the professor. "They are very focused on what the mental age of the robot is."

Hirose is not a strong proponent of behavior-based robotics, preferring to see robots following explicit sets of rules: "Behavior-based systems are all very well for demonstration purposes, however if you do things explicitly you maintain a logical understanding of how things are working, and can thus fix a problem far more easily than if all you have is a black box of behaviors. It's also desirable to be able to specify concrete limits on the system, a black box with no limits can be very frightening." Hirose is a pragmatist when it comes to the development of robotics; asked about the future, he sees things we don't think of as robots becoming complex enough to be given the label 'robot.' He includes things like station ticket barriers and cars that gradually become more and more intelligent in order to cope with a wider variety of situations and circumstances.

All Japan Robot Sumo Tournament

The robotics students at Titech had told me about the All Japan Robot Sumo tournament, and I was determined to find out more. Robot Sumo is the largest robot competition in the world, outstripping the slightly better known RoboCup (RoboCup 2001 had 120 entrants, the 13th Robot Sumo had over 4,000). The Robot Sumo first prize is \1 million (around $10,000) and the competition is open to everyone. Still, most of the entrants are high school and university students and the robots they produce are often part of their studies.

I caught some of the opening rounds of the autonomous competition at the Kanagawa RoboFesta in Sagamihara. It was clear that one of the more successful tactics appeared to be to build a compact robot that was fronted with a bulldozer-like shovel that maintained contact with the floor, trying to push the opponent from the ring in a series of staggered attacks. This generally successful ploy would sometimes be foiled by the robots with long appendages that would distract competitors and circle around to attack from the rear. The bouts had a similar fascinating quality to real sumo. In sumo there is " detectable to precious few " a real tension that builds up around the bout, particularly when it goes on any longer than a few seconds. Robot Sumo manages to capture some of that in holding the crowd's attention.

RoboFesta

The Robot Sumo would have been reason enough to find myself in Sagamihara on a fresh autumn day, but there was more: RoboFesta, a series of events happening throughout last year celebrating Japanese robots in one form or another. RoboFesta Sagimihara was a kind of cross between a traveling circus, a museum and a DIY robot competition. Maybe this is just what the Japanese do in order to get the kids out of the house during the summer holidays, but the recent Japanese explosion in the entertainment robotics field had created the climate for a new kind of event. RoboFesta Sagamihara was being held at a sports arena that was conveniently situated next to a large high school (site of the Robot Sumo). The sports arena was used for lectures on how to build robots, and the exhibition center had sprouted a series of displays and booths, some presenting a particular piece of technology, such as a wheelchair that could climb stairs, a saxophone-playing robot, an exhibition of Japanese robots through the ages and a couple of sound stages that hosted robot shows. The robot shows generally seemed to consist of a huge audience of kids, listening excitedly to a futuristically attired female announcer introducing a mobile robot like Honda's Asimo, or ATR's Robovie.

The special attraction at Sagamihara RoboFesta was a collection of space-faring robots from the Japanese Space Program. Tamiko Masutani, RoboFesta's head publicity officer told me that this was the first time that all of the Japanese-made Moon and Mars rovers had been brought together. Masutani explained that the money for the events had come from a mixture of companies and local government sources. All the famous pet robots were out in force at the events too, with playpens full of Poochi, DOG.com, Aibo, BN1 and C-bots for the kids to play with (or more cynically, to mold them into good little consumers at an early age.) One of the most fascinating exhibits was the "Nurses Assistant Power Suit" developed by the Kanagawa Institute of Technology. The suit is a kind of exoskeleton that provides the wearer with extra lifting strength. Keijiro Yamamoto explained that in tests a nurse weighing 64 kilograms was able to lift a patient weighing 70 kilograms. The prototype suit weighs 18 kilograms and cost $20,000 to build, but Yamamoto believes he will be able to substantially cut both figures in any commercial version.

Other exhibits included a human-shaped gymnastics robot that could flip itself round a horizontal bar and a robot that could write on a whiteboard. Perhaps the most revealing event of the day was the Insect Robot Tournament. During the week various children had been attending robot-building seminars at the sports center and now the boldest of them were about to pit the robots they had built against one another in speed trials. There was no shortage of fanfare as the 20 or so 10 to 12 year olds (mostly boys) were paraded in front of an appreciative audience. Those further away were able to see the expressions on the youngsters' faces on a video wall replaying the competition live. It would not have been difficult to believe that the event was being televised, and maybe it felt that way to the young makers of robots, who had assembled insect-shaped androids from the Bandai Wonderborg kit. The insectoid creations had to try and make their way around a maze filled with such obstacles as floors made of marbles, sand and other difficult-to-navigate terrains. I sat there snapping away with the camera, thinking about how all this fun was laying the groundwork for Japan's next generation of Robot enthusiasts.

By the end of the fourth and final RoboFesta at Yokohama over half a million people had attended the series of events. I asked Masutani if they had plans to turn this into a regular event. She replied with a tired look in her eyes: "Let's see how we feel when we've managed to recover from this one."



Robot Events for Grownups

I met Gomi of AAIS again at Evolutionary Robotics 2001, a symposium he has been organizing on an annual basis for eight years. Falling between the efforts of other academic or industrial groups, Gomi focuses on bringing together some of the most radical thinkers in the robotics field and working hard to try and encourage the audience members to think laterally, offer their own viewpoints and other similar things that the Japanese stereotypically find hard to do. ER2001 featured only one Japanese speaker, Dr. Hiroaki Kitano of Sony Computer Science Laboratories, but this symposium was about exposing Japanese researchers to the kinds of people they wouldn't generally meet in Japan.

I grabbed a moment with professor Rodney Brooks, the man who started the behavior-based robotics field, challenging the orthodoxy of the robotic control systems theorists who insisted on building robots that followed logical theory. He told me that while the Honda people had initially claimed that the control system was developed internally, that they had later admitted to him that it was based on the work of professor Atsuso Takanishi at Waseda University. Asimo performs lots of math in order to work out where to put its legs; the antithesis of Brooks's behavior-based robotics, which emphasize robotic control that is built up by combining many low-level reflexes. He suggested that the work being done at Sony was more along the lines of his own theories.

I asked him if he thought that there was some innate contradiction between a very top-down society like Japan trying to implement his very bottom-up ideas. Brooks caught my eye: "You expect me to take that and run with it?" he laughed, "In front of a journalist?" I backed off before I created a diplomatic incident, but Gomi was less shy when it came to criticizing the Japanese approach to robotics, suggesting groups like MITI did not really understand Brooks's "bottom-up" message.

It was at ER2001 that I managed to ask Gomi what, if anything, was preventing his robot wheelchair from coming to market. He replied that the technology was all there, but he had two words for me: "public liability." He gave me an example about one test user of the robot wheelchair in Europe, who had injured a hand going through a doorway whilst seated in the device. The robot wheelchair is designed to prevent the wheelchair itself from hitting obstacles, but won't change direction if the user sticks their hand out. Who is responsible for this kind of accident?

Humanoids Workshop

Part of the larger Conference on Humanoid Robotics at Waseda University in November 2001, the Italy-Japan workshop promised some alternative viewpoints. They had designers and philosophers in much the same vein as ER2001, but claimed uniqueness, with perhaps the first ever presentation by a Catholic priest on the subject of humanoid robotics. The workshop featured other key figures in Japan's robot field. One of the key speakers was Takanishi, a student of the late professor Ichiro Kato, who had built WABOT-1 in 1973, the first ever full-scale anthropomorphic (human-like) robot. Takanishi is now a core member of the newly established Humanoid Robotics Institute at Waseda University.

Takanishi reiterated the point made by Gomi that "the real barriers to the introduction of robots into society are insurance, license and culpability." He went on to suggest that perhaps the solution was to have a robot "black box" similar to those found in aircraft in order to be able to determine responsibility after an incident. Like many others I spoke to, he made the point that Japan and other Asian societies have a theological background that makes it acceptable, even natural, for them to think of artificial objects having spirits or souls.

Speakers from other fields, such as the designer Shunji Yamanaka, the philosopher Masao Kurosaki and the Catholic priest Jose M. Galvan, all had interesting things to say on the topic, but I ended up feeling that they were lagging behind the robotics revolutionaries. Kurosaki focused on saying that the "minds" perceived in robots are always a projection by the human observer. Yamanaka, the designer, presented his beautifully sculpted robot that could turn to look at observers, and Father Galvan talked about how no machine could have a concept of self.

Yamanaka's design was exquisite, but conceptually the idea that gaze direction is important in communication is well known in the robotics field, and the philosopher and the priest seemed stuck in some industrial conception of machines as automata that could not think or learn " no consideration of the developments in genetic algorithms and neural systems that have allowed software entities to learn and evolve. It is easy to be critical of specialists from other fields who are struggling to keep up with the latest developments in AI and robotics, but perhaps this is the danger; that the rest of society will get left behind in the wake of rapidly advancing technology.

Perhaps the Japanese aren't particularly worried about this. Kitano had suggested at ER2001 that Westerners are slightly uncomfortable with humanoid robots, while the Japanese suffer no such inhibitions. He described how he and his colleagues had been surprised by how easily Aibo had ended up becoming a "family member" for many of the Japanese who purchased the robot dog. Kitano also talked about Pino, a humanoid robot developed outside of his work at Sony, with the Japan Science and Technology Corporation. I had originally seen Pino walking clumsily at the RoboFesta and not paid much attention, absorbed as I was with the smooth walk of Asimo and the coordinated dancing of the Sony humanoid. It is easy to dismiss Pino on the basis of its wobbly walk, but that is to miss what Kitano and his colleagues at the Symbiotic Systems project have achieved.

Pino is breaking new ground on a number of fronts. Not satisfied with being the first robot to be exhibited at the Museum of Modern Art in New York, Pino has appeared in a video with the Japanese pop superstar, Hikaru Utada. He also has a more nimble cousin called Morph, who can dance and turn somersaults. Two venture companies, ZMP and the Flower Robotics Corporation, have been set up to handle robot design and marketing associated with Pino and Morph, and Kitano is not shy about suggesting that he expects that these companies will eventually become the Ferraris of the humanoid robot industry; something that went down well with the mixed Japanese-Italian audience.

Amongst the hype and head-turning (literally) technology it's easy to miss perhaps the most important development. Pino, like the Linux operating system, is an 'open source' robot. The complete specifications for how to build the Pino humanoid are being distributed under the GNU public license (GPL), the same license used in Linux distributions. Pino has been designed to be assembled cheaply from readily available components; thus its walk is not as smooth as high-precision, expensive systems like Honda's Asimo. Because the design specifications are available publicly (www.openpino.org) and free of charge, anyone can go out and buy the off-the-shelf components required to construct a Pino. Since Pino requires no commercial license, academic institutions and private individuals will be able to develop the technology in all sorts of directions.

In this fashion Kitano and his colleagues are investing in the fundamentals of their field. In the same way that the lack of copyright on Hyper Text Markup language (HTML) and Hyper Text Transfer Protocol (HTTP) made the World Wide Web, and in fact the Internet as a whole, an open platform on which everyone could develop new business models, so perhaps Pino might form a similar platform for humanoid robotics. It takes this kind of visionary release of technology not to create new businesses, but to engender whole new ways of doing business. Who says the Japanese can't innovate? @



Note: The function "email this page" is currently not supported for this page.