TRON Project Leader's Opinion

Ken Sakamura

Interfaculty Initiative in Information Studies, The University of Tokyo

For the last several years, I have been promoting not just to Japan but also to the world that we should pay attention to ubiquitous computing and ubiquitous networking. This is because in addition to the fact that the TRON Project itself has set "computing everywhere" as its development goal and advocated from more than 20 years ago that "in the future computers will flood into places where living spaces extend," various component technologies have considerably matured, and the opportunity through which we can realize that conception has risen.

Well then, if one asks--technically, given that there are issues such as chip stability, if we wait a while will these be resolved and practicalization take place smoothly? That's just not so. Even in the TRON Project, although we have stressed the importance of verification tests from the beginning, we have come to believe that at present we should carry out verification more vigorously. This is because mere expectations for new technology are far ahead, and in spite of the fact that at present we are in a phase where we must straightforwardly solve problem points, there are too many examples of people ignoring that and making reckless demands.

A typical example is an application that uses RFIDs in supply chain management in the distribution field. Certainly, in that it can prevent merchandise defects, reduce inventory labor, and so on, it is the easiest to understand among ubiquitous technology applications. People understand an image in which a revolutionary distribution system could be created if there are cheap passive RFIDs that we can read in several hundred tags at a time from a distance of about 10 meters via electromagnetic waves. However, with passive tags, because we supply electric power via electromagnetic waves from the reader/writer, in order to extend the distance, progressively stronger electromagnetic waves become necessary. With the electromagnetic waves that we can normally use, on top of there being limits to operations where only weak electric power can be obtained, they are contending with external noise in places where they are weakest. Even if we talk about the 900 MHz band that has a long reach range, just the performance of its ranging is good; on the contrary, it interferes with other reader/writers, and diffused reflected electromagnetic waves hinder read-in in the manner of ghosts. There is a problem with the dielectric constant of the things that tags are attached to, and there are also problems such as interference with other tags. Also, lots of electromagnetic waves are already flying about in living spaces. As a result, when a chip that has no problems in an ideal electromagnetic dark room is actually tested on location, things like being totally unable to read it happen.

If we sincerely consider these types of location conditions and technical limits, then when we don't do a follow-up on the operational aspects, we should expect to notice things that cannot come into practical use at the present point in time. Such things as reading in at a single time the merchandise inside a cardboard box mentioned earlier are a dream within a dream. That, as one would expect, is something even Wal-Mart seems to have given up on it, and recently it appears they are only considering the reading-in of cardboard box units, but there is a report that even in those conditions, even at tests at Wal-Mart, which is a proponent of the 900 MHz tags, on location they were only able to read about 60 percent at best. This means that recounting of merchandise by means of a separate method will become necessary, and that first of all it won't come into practical use with objectives other than monitoring the pulling out of merchandise (for monitoring pulling out, all you need is to release a warning based on a tag that has been read disappearing).

At the present point in time, a high-low mix of distribution in which we manage by tying together via the network boxes premised on reutilization onto which we attach a 13.56 MHz large-scale tags, which have high reliability but are difficult to make low cost, and individual items to which we attach 2.45 GHz ultra small-scale tags, which have strict read-in conditions but can be made low cost, is the most realistic, but even that voice doesn't seem to reach the 900 MHz believers.

Also, even if not just the operational side, but the technical problems are solved, there is not even a business model that has been established. We need a different approach in putting out an answer as to who will bear the cost of the tags, and so on. However, in regard to this also, we can see that the 900 MHz believers are fixated on the dream story if we can unify all tag demand with 900 MHz tags, then the price will interminably fall. Although I do not think that the 900 MHz tag itself is bad, I cannot but object to the not very well based scolding that "there must be only one type of tag," or "everything must be 900 MHz." Returning to my original point, we should go on to solve our problems with a frontal attack method in which we ask, "what do you really want to do?," and "where is the balance between the most outstanding technology for that and operating it?" Won't things go badly by saying such things as "the determination of the promoting side," or "because their annual turnover is great, they can give orders to suppliers on a strong posture." Furthermore, what kind of person is it who convinces himself that another person's assumptions are the winning horse and stops thinking. The risk of blindly aiming at a target someone else has decided upon is that of the goal completely disappearing after your opposite has changed his thinking before you know it. It's fine if you can ride well by "riding with the winning horse," but it's shameful in that people are being taken by that very America in question as "the bothersome lot who are blindly advancing toward a goal we have already abandoned." I think the persons concerned had better wake up quickly. You could say it's someone else's business; one can only try to say it's pitiful. This is something one should think about with his own head. On this occasion, even over there was somewhat of a feeling of blind acceptance, but the point on which America is originally amazing is that there are many people who think with their own heads. By means of that, if they think something is bad, even if they cast away the obstinacy and honor of a portion of the people, as an organization, they rapidly switch the goal--that realism is something we should learn.

From the start, it is not even necessary to consider everything with just passive types. In short, the proper material in the proper place. Even to obtain a reach range of just several meters, a high output reader/writer of several watts even in the 900 MHz band is needed with passive types; this is a high level that is almost dangerous for people who will approach the vicinity of the antenna over long periods of time. However, with an active tag utilizing weak wireless that we recently announced, even though the electromagnetic wave output is very weak, on the order of one several ten-thousands of that, anti-collision of about 1,000 tags can easily be realized. With active types, changing batteries is a problem, but the performance of long life batteries has risen, and the cost will go also down in the future. It is also possible to combine with autonomous power sources, such as solar cells and oscillatory power generation. I have taken to heart the fact that developing the appropriate technology suitable to the application is important.

Just as we say there is not a royal road in academics, new technology, after it is established, goes on to advance itself one step at a time. As I have repeatedly said recently, if we make 2002 the first year, then if we don't wait until about ten years later, 2012, I think the technology will not be mature and stable. Ubiquitous computing, even in computer science, is technology with a wide range of possibilities and applications rarely seen in recent years, but it is still very much in the process of developing. This kind of thing is a chant like crying "wolf;" for it to end up being degraded into a type of temporary meal in the IT field will be very regrettable. For Japan, which is taking the lead in progressing down the path of aging with a small number of children that has never existed--and, for the countries that will follow Japan--going beyond simple convenience and pleasantness, this is important technology that may be the key to survival. For this very reason, I would like to nurture it with care. When I look at movements in the world, I believe that we--the T-Engine Forum and the TRON Project--can contribute to the world with ubiquitous computing, and I think that I would like to continue those type of activities hereafter also.

The above opinion piece by TRON Project Leader Ken Sakamura appeared on page 1 of Vol. 93 of TRONWARE. It was translated and loaded onto this page with the permission of Personal Media Corporation.

Copyright © 2005 Personal Media Corporation

Copyright © 2005 Sakamura Laboratory, University Museum, University of Tokyo