Patent applications need to be screened by experts in their industry before patents are issued.
Yesterday I published an article that looked at the state of the system in the U.S. and the changes that may happen when the Supreme Court hears the case of Alice Corp. and CLS Bank International. A ruling from the Supreme Court may serve to clarify the bounds of what is and is not patentable when it comes to areas such as business methods, financial systems and software patents.
However, this does nothing to fix the time it takes to get a patent issued or the number of patents that get through the system that should never have been allowed in the first place. This, in my opinion, is one of the biggest problems with the patent system. It means that frivolous suits can then be filed, which then have to be settled using the court system rather than preventing them from being issued in the first place.
There is an easy fix to this problem. The industry needs to help the U.S. Patent Office. We already have patent applications being published before patents are issued so people can see what patents are being sought in their areas of expertise. But right now there is no review period. There should be an open review period after the patent is filed but before the patent is issued, during which anyone can supply prior art or evidence that it has been in practice for a long time making the obviousness test fail. As with any chip design, it’s easier to fix a problem earlier than later.
Consider an example of a patent application in our space that was recently filed. It basically describes the way the industry has been performing verification for the past 20 years, comparing the output of two simulations at different levels of abstraction.
How many of you have written papers about this very thing, used it in your everyday work or had to sit through presentations from EDA vendors that talk about this very concept? I know I was giving such a talk two decades ago.
So what does the application say about the prior art? It lists no patents or publications, but does say previous approaches have been less efficient.
It would be very quick and easy for me to provide a list of prior art as long as my arm, saving the patent office a lot of time. Patent Office workers may be well intentioned, but there’s no way they’ll be as knowledgeable about these subjects as experts in the field. This very simple modification of the patent system—allowing prior comment—would help enable companies to prevent patents getting through the system in a very cheap and simple way. It also would go a long way toward making the patents that do get issued a lot stronger.
We need to allow the industry to help weed out the patents that would result in frivolous claims and be the food for non-practicing entities. Can you imagine this kind of patent in the hands of a non-practicing entity? They could sue every verification company out there, not to mention companies that have used these techniques to verify their designs.
 |
 |
 |
 |
 |
 |
 |
 |
 |  |
Albert Einstein sure doesn’t work for the Patent Office anymore! No “experts” that can be employed elsewhere do. So, finding a way for outside experts to provide input is good, but those experts need to be able to recognize a new idea when they see one, independent of ego or economic & political pressure. How do you organize and fund that?
What was the number of the verification patent you mentioned?
I’ve seen a number of patents where (after struggling to understand the opaque legal description) I grab a decades-old text from my book case and thumb through until I find a description of the same or similar technique, often described as the standard approach. I can understand that the patent applicant may not have been aware of the prior art, the patent attorney isn’t interested in searching for prior art, and the patent examiner is perhaps inadequately familiar with the area. Allowing a review and comment period would allow stopping patents which should never be issued before they progress through the system.
this is a good way to start on software patents: http://www.joelonsoftware.com/items/2013/07/22.html
I agree with you. We have hit the point where the USPTO cannot hire enough “practitioners in the arts” to understand what is obvious or prior art.
A very good idea, with however one caveat. The “experts” must be absolutely free of conflict of interest, not an easy thing as almost by definition an expert will be employed or in service to a player in the technology of the matter. I don’t see how that can be overcome. However the Patent office could and should have industry experts in its direct employ who can be effective in siting prior art.
Although this sounds like a really good idea, it is problematic. By the time a patent attorney takes your idea and translates it into patenteese, it required both technical expertise and some experience in writing or reading patents. Also there is a problem of keeping the data safe.
It would be nice if the USPTO and the EPO would hire technical experts as ‘post examination reviewers’ or find a way to outsource it to a pool of experts once they can find a way of of protecting the data. this will be much easier if they can be called for after 18 months when the application is made public.
Hello Brian, In the UK patents have similar problems where patents are somewhat verbatum, I agree that there should be something done. These days you would expect that a simple word search is available to seperate the area of the patent application along with a wild card search both searches should be done on extinguised, existing and newly applied for applications.
One other area of debat should be around Design. As in my case my patent was applied for two weeks prior to a design but because of the timescale of appication a design had be appied for in my case in China, sold to a UK company and used for 10 years bfore i managed to remove it. The time scale i mentioned is patent publications taked a minimum of 10 months a design application can be speeded up and can be available in 3 months. the diference in IP rights are you can produce a patent in the country that the patent covers and that is about it. You would not have the right to stop the design from being imported and sold along side the patent.
So, I would also consider that there would be some benifit in having either date of application and /or cross partion where other IP had priority not just in publication
other benifits would be lower insurance cover for IP rights
Brian, the patent law and the constitutional articles are specific that patents are granted to art that is NEW in the USA. As such, the procedures at USPTO would have examiners checking literature and products. However, reality is these examiners have pretty extreme quotas to complete and simply do not have the ability to look at much outside of USPTO references, and even in USPTO there are a lot of “hidden cubby holes” where patents may be put (hidden) if they have the right wording to fit some inane classification. Worse, the patent law and regulations specifically allow the applicant to use his own terms to describe his invention. I have looked at patents where an expert in the field would immediately say, “Hey, that automated ink dispenser you described is really a computer printer!!!” Obfuscation, intentional or unintentional, abounds in the applications.
As such, the patent office has gone through several pendulum swings of how to deal with an inability to make their customers and Congress happy, while not angering the courts. What you propose would be a great idea, but will cause a bureaucratic war because the mind set of the old timers at USPTO is about protecting an idea until the time the patent issues. The newer laws that dictate “first to file” make such protections obsolete. The change will also require coordination with WIPO and other bodies to prevent international backlash.
Another significant change that would help greatly would be to make the holder of a patent legally responsible for abandoning claims that are known to have been inappropriately granted or not valid. This would make it possible to prosecute patent trolls who are spamming an entire industry with claims that even one of the targets shows are not valid, so causing the trolls to leave him alone. Today, that merely causes the trolls to find a new victim.