“Expensive Fools,” then and now

Many billions of dollars have been spent by Japan’s top companies in the last decade to develop the next generation of robots — famous for such technical marvels as expressing human emotions on a silicon face, bipedal walking, roller-skating, playing musical instruments, and cooking meals. Japanese companies such as Hitachi, Toyota, and Toshiba are broadcasting ads and posting billboards across the globe that show-off their latest robotic achievements. These products aren’t for sale. You cannot buy they for industrial or commercial use. But their existence suggests a fantastic future where robots will be serving us in all sorts of tasks, both menial and complex.

Too bad this is so utterly removed from reality. For all of Japan’s collective effort in developing its robotics industry, TEPCO has had to turn to iRobot, — a Massachusetts-based company — for the Packbot, a simple yet practical robot equipped with simple claws and cameras, to enter the Fukushima plants. The robot first entered the plants on Sunday (a month after the disaster, after it became clear that no Japanese robots were up for the task) taking measurements and photographs of areas where it was unsafe for humans to venture. (Interestingly, but not surprisingly, Japan has refused robots provided by the Chinese military.)

On that note, iRobot is also the company that brought us the Roomba, a robot that vacuums the floor and which is probably the most popular robot among Japanese consumers — showing that Japan Inc.’s investment in robot technology cannot save it, even in its home market. For all the fanfare in recent years, in truth, Japan’s robots are so advanced that they have skipped a generation of usefulness. The robotics on display by Japan Inc. may be useful in ten years, but they are nothing more than gimmicks today.

Coincidentally, just hours after I read the above article on TEPCO having to go to iRobot for robots to use in Fukushima, I was browsing through Michael E. Porter’s tome, The Competitive Advantage of Nations (which is just as heavy a book as it sounds). Published in 1990, one case study in the book addresses Japan’s robot industry, and I found this section prophetic:

The first robots used in Japan were imported from the United States in 1967. The Japanese robotics industry had its beginning in 1968, when Kawasaki Heavy Industries signed a licensing agreement with Unimation. Kawasaki was both a major potential user of robots as well as a producer of related products and services. It produced a broad range of machinery and parts, including engines, motorcycles, aircraft, machinery, complete plants, and ships. In 1969, Kawasaki began to sell Unimate robots, the first robots produced in Japan. Kobe Steel was also an early licensee of American robot designs.

The early Japanese robots produced results that were somewhat less than expected. They were often referred to as “expensive fools” and many were relegated to the corners of factories to collect dust. However, Japanese firms began to improve upon the robots they imported. Kawasaki redesigned some of the parts of the Unimation machines and upgraded its quality. In the late 1960s, the mean time between failures (MTBF) of an imported robot was less than 300 hours. By 1974, Kawasaki had achieved an MTBF of 800 hours. By 1975, this figure was 1,000 hours.

It’s a total cliche about Japanese business from the late 20th century: “Japan doesn’t invent things well, but they are great at improving things.” It’s a cliche so tired that it’s rarely rolled out much anymore. Yet as I read this, I realized that the observation from 1990 is still true, and is being echoed in real life events more than twenty years later.

Japan must get back to doing what it does best — improving, not inventing. If Japan’s researchers were tasked with improving the Roomba or the Packbot, I bet they’d do an amazing job. But left to their own devices, with the task of making something new, Japanese researchers, for whatever reason — dedicate themselves to creating a robot that can cook okonomiyaki and play the violin — and for what? Another generation of “expensive fools” destined to collect dust, due to the minor value the products provide in industrial and consumer markets.

21 thoughts on ““Expensive Fools,” then and now”

  1. Check your facts. The Japanese did create robots resistant to radiation. I couldn’t find the article I read a month ago on the subject that said that the robots were available, but the nuclear plants did not purchase them. This article has the same scathing tone as your post, but at least it cites that the Japanese did in fact make robots for this type of situation (this author too could have looked into the reason why the RaBot disappeared from the market instead of asking a rhetorical question): http://news.cnet.com/8301-17938_105-20044970-1.html

  2. From 2005:

    “…Joe Engelberger, the father of modern robotics…emerged exasperated from a visit to the international robot show in Tokyo…’These are toys that are being made…Nothing serious. Just stunts. There are dogs, dolls, faces that contort and are supposed to express emotion on a robot’…It was pointless, expensive and unnecessary for Japan…to tinker with trivial inventions like robotic house sitters that rang to say there was a burglary going on, he said…What the $US8 billion robotics industry needs is for engineers to design practical robots for personal care. So why isn’t more work being done? Mainly, Mr Engelberger thinks, it’s because everyone is immersed in needless research and companies are distracted by the uneconomic quest for the humanoid, which he derides as toy making.”

  3. Japanese companies make the violin playing, okonomiyaki-making robots as part of PR efforts. All of this promotion has precision movement as its bottom line R&D impact, and this is a huge advantage in the area of industrial robots. The point isn’t to produce the best silly robot, but to draft useful prevision movement tech into PR. This has given a modest boost to Japanese companies reputations globally, but the main event is sales of industrial robots.

    Japan’s market share in industrial robots and similar precision programed machine tools is over 50% of the global total – this is a key Japanese export area and one of the reasons why Japan has nearly balanced trade with China (while the United States has a close to 3 to 1 import / export imbalance). Other notable statistics include “new installed industrial robots” for 2010 – Japan is 30% of the global total, compared to 30% for the EU and 15% for the United States. The United States lags badly in industrial robotics technology compared to Japan and German and this is a both a major growth area and one of the major areas of export from developed countries to developing growth engines such as China (and to a lesser extent, India, Vietnam, and so on).

    Packbot has an advantage over similar Japanese robots for a very good reason – it was developed as a military robot for bomb disposal. Most Japanese companies don’t want to touch something like this with a ten foot bomb disposal pole. Is anyone particularly surprised that American companies are leading the way in what is essentially military equipment? In high volume high profit areas like industrial robots, the US makes some good stuff, but is essentially an also ran behind Japan and Germany and increasingly South Korea and Taiwan.

    Comparing a military-use robot, which exists because of America’s rather unique political economy, with Japanese song and dance robots, however, is an apples and oranges exercise. Better to look at Japanese leading technology in the area of industrial robots and to ask whether or not Japan will be able to maintain this advantage (estimates have the market for industrial robots growing by 300% by 2030, will China start making its own or will Japan be ahead on the curve and stay ahead?).

    On other fronts, service robots for domestic use are still a minuscule market, but Japan is thought to have a competitive advantage here because of ageing-society related technology.

    America’s advantage in military-use robots is evident, but given coming cuts to American defense spending and China’s likely move into this area as an export market and it seems both shaky for the US, and likely not a feasible area for Japan even if military related exports were to become a national priority.

  4. While I agree with M-Bone that Curzon is talking mostly about “PR stunt robots,” I have been a bit worried in recent years that even the stunt robots don’t seem to be progressing much. I don’t think that Asimo has seen any major changes in the last 5 years. Sony killed off the Aibo when budgets had to be slashed, and I’d expect that the other stunt robots — at least the ones from companies that make actual useful objects — may be sunsetted as the bottom line descends.

  5. “I don’t think that Asimo has seen any major changes in the last 5 years”

    I’d argue that the point of Asimo wasn’t to make Asimo at all, but as a dumping ground sideshow project for all kinds of different robotics research that could, in separate forms, have business applications.

    I’m not 100% sure if Asimo was really necessary, but Honda claims that the Asimo balance and locomotion research has gone into its “Walking Assist” technology. http://world.honda.com/Walking-Assist/

    This sort of stuff is both impressive and could build a market.

    I think that it is important to keep in mind that Japanese companies don’t (usually) sit down and say “how can we build a robot that juggles”, they dump money into various areas of robotics R&D and, as an adjunct project, try to build something to show off along the way.

  6. A topic near and dear to my heart. Thank you for writing this.

    I agree with M-Bone that the real issue here is the lack of military applications in Japan. Military robots (from bomb-disposal robots to aerial drones) need to deal with the real world. That’s something Asimo has never had to do; its performances are all carefully choreographed set pieces with pesky real-world obstacles like humans being kept safely out of the picture.

    I had always seen Asimo as the modern equivalent of Honda’s F1 efforts in the Sixties and Seventies. But unlike race car engine innovations that presumably made their way into consumer models, Honda has never really managed to articulate how and where Asimo would make a difference in everyday lives. That is really critical for endeavors like this. And as Marxy mentions they’ve totally failed to debut any new innovations (or even iterations!) on the design in recent years. Is it a funding issue? Some technological hurdle?

    It’s also telling that one of the more high-profile applications in consumer robotics, cars that can drive themselves, emerged not from an automaker but rather Google. Japan is a hardware titan struggling in the software era.

    “This sort of stuff is both impressive and could build a market.”

    Exoskeletal assist was all the rage a few years back, but once again, America has taken the lead thanks to the military. (Juxtaposing imagery of the HAL Body Assist Suit lifting sacks of rice against the Raytheon XOS-2 hoisting missiles and shadowboxing is about as clear-cut an example of the difference between Japanese and US robotics technology funding as one could hope for.)

  7. To build on one of M-Bone’s comments above, Japanese robot makers have been driving their R&D over the past few years in response to the aging society here — trying to build more human-like robots for the purpose of helping the growing number of technology-averse elderly people while the working-age population shrinks. In the US we get by using Mexican labor, so more of our robot budget comes from less cute defense and law enforcement applications, where there are preferences for American contractors written into the law.

    As far as self-driving cars, I think it’s pretty obvious why an American company got there before a Japanese company did. Manual driving in Japan is difficult enough with all the narrow streets, nonexistent sidewalks, and elevated expressways in earthquake zones.

  8. M-Bone:

    “America’s advantage in military-use robots is evident, but given coming cuts to American defense spending and China’s likely move into this area as an export market and it seems both shaky for the US, and likely not a feasible area for Japan even if military related exports were to become a national priority.”

    Indeed — but as I noted, iRobot is ahead of Japan not just on the nuclear-plant robot, but also on the sweep-your-floor robot.

  9. “Indeed—but as I noted, iRobot is ahead of Japan not just on the nuclear-plant robot, but also on the sweep-your-floor robot.”

    Can you separate iRobot’s home-use and military-use products, though? If you look through iRobot’s history page, you can see how many of their earlier endeavors were military funded. Minesweepers, space exploration, “tactical mobile robots,” DARPA contracts, etc.

    http://www.irobot.com/sp.cfm?pageid=203

  10. France proposed to help and send radiation-resistant robots too, for the purpose of gathering samples or removing fallen stuff, carry out excavation work, etc.
    Got refused by the Japanese side for “not being adapted to the situation” for some reason.
    True the robot looks way bulkier than the iRobot one.
    (link in French only, couldn’t find news in English on this, sorry)
    http://centre.france3.fr/info/le-japon-ne-veut-pas-des-robots-francais-67969326.html

    France began to develop this kind of machines along with Germany after Tchernobyl.

  11. Interesting point M-Bone — I don’t know the answer to your (rhetorical?) question but the answer may be “no.”

  12. It seems that the Japanese Societ of Robotics and Automation has something to say about this topic.
    http://www.rsj.or.jp/shinsai/RoboticsTF_1.pdf

    Maybe a sad thing , but I am now rather inclined to believe that developing military industry is probably not only useful for developing military technology but also for developing proper skills for risk management and after crisis management.

  13. “Maybe a sad thing , but I am now rather inclined to believe that developing military industry is probably not only useful for developing military technology but also for developing proper skills for risk management and after crisis management.”

    Yes. That’s why Katrina went off without a hitch.

  14. By itself, Asimo and friends aren’t that useful. I can’t help but wonder, though, what the knowledge that went into them could produce when combined with prosthetic design? We’re probably nowhere near able to produce full-body cyborgs like the ones in Shirow’s manga, but knowing how to make machines that can walk or imitate human expressions is a good step closer.

  15. Examples of some of the robot research going on in the US military industrial complex.

    http://io9.com/#!5793714/7-bizarre-robots-in-the-pentagons-defense-budget

    And the research is already having significant dividends for civilian projects. For example, a tiny remote sensor robot (“throwbot”) designed originally for anti-insurgency intelligence gathering is new going on sale for civilian police and fire fighters in the US to scout out buildings for safe approaches before entering.

    http://arstechnica.com/tech-policy/news/2011/04/military-surveillance-robot-unleashed-for-public-safety-use.ars

  16. This Japanese blog article cites Shukan Shincho and blames Koizumi for dumping Mitsubishi Heavy Industries nuclear accident robot.

    小泉竹中路線の構造改革のせい= 100k hits in Google

  17. ““Maybe a sad thing , but I am now rather inclined to believe that developing military industry is probably not only useful for developing military technology but also for developing proper skills for risk management and after crisis management.”

    Yes. That’s why Katrina went off without a hitch.”

    Why don’t we work on developing technology to be used for crises instead of creating crises elsewhere? While the Predator Drone was useful in responding to this disaster, Tomahawk missiles weren’t. When we are talking about an aging Japanese society with a stagnant economy. Developing a military industry like the Americans have with bring with it more costs than the occasional benefit of a roomba.

Comments are closed.