Wrong Patents
You can’t trust anyone.
Lately I attended a nice course. It dealt with clock synchronization. Without getting too technical, the synchronization problem is that many times an (electrical) system sends a signal which is supposed to reach a few other systems at the same time. Practically, the signal reaches each destination at a slightly different time, which can cause some problems; so there are various techniques to make sure that all destinations receive the same data, and that it is correct in all of them. So after analyzing various reasons that may cause this problem (a bunch of technical terms - skew, jitter, drift, high-rate… this sort of stuff), the guy - who happens to work as a professor at the Technion - presented some possible techniques to overcome it.
At one point he told us a story about a doctorate student he once had. His student’s job was to go over many designs of synchronizers (which are special electrical circuits designated to solve this problem) and see if they’re working. As one source for these synchronizer designs he used the US Patent office, which is a good source to find such designs, he figured: when you issue a patent, you’re supposed to reveal your design, they say.
His student took such a design, and proved it not working. Very proud, he presented his work, and the satisfied profesor took it into his slides and started to present it in his lectures and teach other people with it. As it happened, after a while he presented these slides in a certain company, and the designer who issued this patent was working in that company and sat at the audience. After the lecture, he presented himself to the profesor and they had this conversation.
It turned out, that the design submitted to the patent office was not the true design they used in the company. They just thought for themselves that this neat synchronizer they’ve got is too good, and they wanted to keep it propietary, but they still wanted the patent protection (?), so they changed it a little bit, and handed to the patent office a “fixed” version. After this lecture they gave the real version to the professor, and his student proved that this one is actually working - so the student was happy (he received his Ph.D.), the professor was happy (he received this story and a few slides for his lectures), and no harm was done.
Was it so? It got me to wander - how many false patents are out there? What is the point of knowingly issuing a wrong patent? Is there no reviewer on the proposals? Is he not someone that understands what he’s doing? And why do so many companies rush to issue a patent, even on their commercial secrets?
January 11th, 2007 at 9:45
Way more than you think. Enough said.
January 15th, 2007 at 11:29
I’d like to just mention that with physical inventions - such as electrical designs, you have some sort of description of the thing being patented. It may be bogus completely, or it may be broken (on purpose or not) but can still be “fixed” by reading the functional description - in either case, if such a patent is ever litigated it may likely not hold in court.
Similar but a much more serious problem, are patents on algorithms or business processes (i.e. any software that is being patented not on the innovation of its algorithms), in which there is no disclosure whatsoever on the actual design and implementation - they simply patent the idea of such an innovation. The patenter gets his patent protection and the public gets nothing.
January 15th, 2007 at 12:01
The lack of disclosure isn’t the biggest problem there (altough huge by itself), the fact is that algorithms and business processes are even considered inventions (rather than math etc.), is bigger. Patenting a business process means preventing competition for any new business field, and where’s the sense in that?
Lack of disclosure is a bigger problem IMHO with copyrights on closed source software, I don’t see how the public benefits from granting copyrights on things that are never distributed to it (even after the copyright term would expire).
January 17th, 2007 at 21:17
Here are a few interesting things…
http://en.wikipedia.org/wiki/Steorn
This is a company that claims to have built an “Over-Unity Machine” but they found resistance with patent offices for “not giving a patent to perpetual motion designs without a working model”. Hence, they probably give a patent to any machine without a working model.
Second, it’s interesting that in the movie Armageddon, NASA “broke into the patent office” to steal the designs of the drill and they couldn’t get it to work.
January 18th, 2007 at 10:40
With the exception of cases involving perpetual motion, a model is not ordinarily required by the Office to demonstrate the operability of a device.
January 18th, 2007 at 10:41
That should have included a link to the source - http://www.uspto.gov/web/offices/pac/mpep/documents/0600_608_03.htm