Slashdot Mirror
USPTO Peer Review Process To Begin Soon
An anonymous reader writes "As we've discussed several times before on Slashdot, the US patent office is looking to employ a Wiki-like process for reviewing patents. It's nowhere near as open as Wikipedia, but there are still numerous comparisons drawn to the well-known project in this Washington Post story. Patent office officials site the huge workload their case officers must deal with in order to handle the modern cycle of product development. Last year some 332,000 applications were handled by only 4,000 employees. 'The tremendous workload has often left examiners with little time to conduct thorough reviews, according to sympathetic critics. Under the pilot project, some companies submitting patent applications will agree to have them reviewed via the Internet. The list of volunteers already contains some of the most prominent names in computing, including Microsoft, Intel, Hewlett-Packard and Oracle, as well as IBM, though other applicants are welcome.'"
5 days a week, times 47 [or so] weeks of work == 235 days. That gives them just under 3 (2.83) days to read the patent, understand it, look for prior art, and then say "yay/nay."
That's not actually a lot if you think about it. Sure some patents are probably trivially rejectable, but many probably require some deciphering before you get to the "omg that's obvious" stage.
Like this 7 page patent (yes, I'm picking on them) for table based multiplication. Not only is it an obvious idea, that even the average 10 year old could figure out, but it's already been used by many hardware manufacturers (ARM used it for the ARM7 multiplier for instance). From the first few claims the "invention" doesn't really seem invalid until you put it together, and realize that it's the mechanical equivalent of long hand multiplication.
While you or I would easily spot that patent and say "no shit," a patent examiner must be able to defend their decision, so they must actually cite prior art (or make a convincing argument the idea was obvious). That also takes time.
Tom
Someday, I'll have a real sig.
No they don't.
The purpose of patent protection is to grant the right of a person who comes up with something great to profit from it.
Without such protections, someone would invent something and then die of starvation before seeing any kind of profit. While he is trying to produce his invention, some megacorp would take his idea and put it in mass production and beat him to his own market. The inventor would be left with jack because noone would have to pay him to produce his innovation for themselves.
I'm not sure where the Washington Post got their numbers, but according to the USPTO's own Annual Report, they received more than 615,300 patent related applications, which were dealt with by 4,883 examiners. A reasonable calculation then suggests that they would have to process the applications in an average of 15 hours each to keep up with demand.
In actuality, only 332,535 patents were disposed in FY2006, which means the backlog (already in excess of on million patent applications) only grew. In a system where your application is not likely to even be looked at for the first 22 months, and it takes more than 2.5 years for the average application to be processed, they are desperately in need of help examining.
The most depressing part of the report is to look at their goals. The objective is not to reduce the backlog, or improve first action or total pendency time, it is simply to have the backlog increase by less than in previous years. With this kind of thinking, there is no end in sight. What is really needed is a radical change of leadership, such that the resources being allocated and the goals being set can actually improve the situation.
"Si vis pacem para bellum" -Publius Flavius Vegetius Renatus
... maybe they could start peer reviewingOnly to idiots, are orders laws.
-- Henning von Tresckow
His points defend the principle of patents. You are attacking certain abuses of the patent system. These are different arguments. I believe in democracy as a principle, but if 51% of the country voted that black people should be tortured, I would be against it. I'm in favour of the principle of the free market, but if some people lie starving in the street because they have no job or skills, I'm against that.
Patents are like anything else, there are abuses of the system, and extremes that can be cited, but in principle, we are better off with patents and copyright than we would be without them. The problematic cases and implementations need fixing, but don't throw out the whole system because parts of it need work.
It's easy to say "do away with it all". Its much harder to say what you would replace it with.
DRM-free indie games for the PC and Mac: Positech Games
Of course, no one denies that the patent system needs change--most likely a significant reform. The Patent Office knows this!! They are currently hiring thousands of examiners to help deal with the backlog. They have instituted hotelling programs to allow examiners to work from remote locations, thus freeing up valuable office space for new examiners. The Office is constantly developing new search tools to better help examiners locate prior art, especially for business method and software patents which can be very hard to invalidate.
Remember, the law as it currently stands states, "A person shall be entitled to a patent unless..." Thus, the burden is on the examiner to PROVE that a patent should not be granted. This can be VERY hard, even when the technology appears clearly to be unpatentable.
I have to say, I am very surprised at some of the comments coming from such an educated group of people. Destructive criticism will get us nowhere.
Notwithstanding the problems of our current system, patents ARE important for protecting innovation. Countries from around the world recognize this, and, believe it or not, try to emulate our system for the protection of intellectual property.
Give the Patent Office some credit here. This is a DRASTIC and REVOLUTIONARY change they are trying to institute here. It is VERY progressive, and it seems very in-tune with the open-source trend in information sharing and collaboration. They clearly recognize the need for change, and they really are working to find the right solution.
So before you start ranting about how the patent office sucks and how patents should be abolished. Take some time to think about why patents fundamentally encourage and protect innovation, and why the job facing the Patent Office is not so easy.
Again, everyone is looking for a better solution. That is why the Office is testing this program! Maybe it will work, maybe it won't. One thing, however, is sure: unhelpful and unreasonable criticisms from close-minded individuals do not help.
(Former patent examiner here)
The vast majority (I'd say 98%) of patent apps are initially reviewed and rejected, but the attorney usually argues the rejection and/or adjusts the claims and sends the aplicaiton back. So the examiner has to re-review the application, and this back-and-forth action often goes on for months or years before the patent is finally issued or abandoned.
So really, each examiner is looking at twice as many applications a week as you think they are. Even though half of them are ones they've already seen, it typically takes 3-6 months to get a reply, and by then it's hard to remember all the details. Also, the claims are often so changed that a whole new prior art search is required. As Tom noted, it sometimes takes a while to dig up the prior art to shoot down each claim, even if the invention is clearly obvious.
At least in my art unit, it was rare to find an application with a spec that was under 20 pages and had less than 30 claims. I'd say 40-90 claims was the average. That's a lot of material to sift through, especially for entry-level engineers with little experience in the industry (which comprise the PTO's vast majority of employees). When I worked there, it was pretty standard for patent examiners to work 10 hour days in addition to Saturdays or Sundays, and not get paid for the overtime.