I know there’s a ton of patent questions on this board, and I know they’re generally frustrating – usually because it seems the posters talk about their ideas in vagueish generalities or analogies. And this post will be no exception
Suppose there’s such a thing as a search algorithm (hey, there is!) that can find the string “end” in words (so it’ll give you a hit for the word “friend.”) Suppose this search method is well-established, and furthermore suppose it is already patented*. Now make one more supposition, namely that I have the technical know-how to tweak the extant algorithm to enable it to find small graphical patterns in pictures rather than syllables in words – so now, if you search the American flag for a white star, you get 50 hits.
In my hypothetical, let this new method be novel and original (i.e., no one else could use a computer to find stars on the American flag before I came along). It’s an extension of a known method, but it’s not one that anyone else has thought of or been able to implement.
Is this sort of re-purposing patentable? Does it make a difference if I wrote the algorithm from scratch, rather than modified a pre-existing algorithm?
[sub]*I know, from reading other patent threads on this board, that software algorithms are historically difficult to patent. But, for the sake of argument, assume this one was successfully patented without contest.[/sub]
It might be patentable, but you couldn’t actually use it without violating the first patent. For instance, if someone invents a car, and you invent an improvement on the car, the steering wheel, you can’t build a car with a steering wheel without violating the patent on the car. The guy who holds the patent on the car can build a car, but he can’t build a car with a steering wheel.
Most patents are new applications of existing technology, and in fact one of the most common ways to “extend” a patent (which, for the United States, is currently only generally in effect from 20 years from the date of application) is to submit an additional patent that extends the original mechanism or process in a way that is integral to the original process but not part of the original patent application. In the case of the scenario listed by the previous poster, a steering wheel would no longer be patentable (and given its general use was probably never patented) but improvements upon the steering wheel could potentially be patented if they were sufficiently unique and not already in the public domain.
To elaborate on what Stranger On A Train said, this is exactly the mechanism that drug companies use. This is why you’ll find controlled (or sustained, etc) release drugs coming out towards the end of the original patent. This is the way that drug companies use to extend their patent and retain market share from the generics.
A month or so ago Spectrum (the IEEE monthly journal) had an article about some new changes in patent law.
As I recall it is now very difficult (perhaps impossible?) to patent an algorithm that is not associated with a device/machine of some sort.
I had similar problem.
I filled patent for New Use of Exiting Technology. But new use is in entirely different field. As like same technology which is used in field A. no one ever tried to use it in a entirely different field Z. So technology used in field A has been used first time ever in world by me in field Z. Is it patent able under process patent category. Please reply.
Assuming your invention is useful and novel, it comes down to whether or not it is obvious to someone skilled in the art. If it isn’t obvious, you can patent it. As pointed out above, you can’t practice your patent unless you have a license from the owner of the prior patent.
The classic case is the sewing machine. Lots of different inventors were working on one, and so lots of different patents on different parts were granted. But none of these inventors could make a sewing machine that incorporated all the other improvements that the other inventors had, because that would violate the patents on the other improvements.
Which meant that no one could make a sewing machine that was worth using. The inventors eventually formed a consortium to manufacture sewing machines with all these improvements.
So yes, you can patent a new use or improvement of an existing patented product. But that doesn’t mean you can actually manufacture or sell or use your improved product without violating the patent of the existing product. On the other hand, the existing manufacturer can’t produce the improved version either without violating your patent.
Let’s say the Pilkington process for making float glass is patented (no idea actually).
What if a candymaker were to use the same basic idea to make flat sheets of candy by floating the candy on a sort of oil? (same basic idea, different materials) Would this be patent infringement?
I think it’s safe to say the “non-obvious” test is not actually being used in practice as demonstrated by the number of painfully obvious patents awarded regularly.
If you can make a case for it. There’s the famous case of the Patent to Use a Laser to Play with your Cat (US Patent 5443036, granted Aug 22, 1995)
The patent was not on any device (such as a mechanical toy involving a laser) – it was on the idea of using a laser itself to amuse your cat. It’s not really a new idea – I had friends in the optics community that were doing this ages before this, using wall-plug HeNe lasers. In fact, the idea of using a focused flashlight to do the same thing had already been published.
So how did this patent get approved? I researched this issue and wrote an article about it. I got hold of the patent “wrapper”, which is an immense bound volume with al the patent paperwork. It turns out that the patent was initially rejected. The authors of the patent went through the objections, altering the patent claims or arguing against the objections , then re-submitting the patent until the examiner didn’t object any more.
Lawyers I’ve talked to told me that this patent would never have withstood a serious challenge. In any event, the patent holders failed to keep up the schedule of patent fees, and the patent lapsed before anyone ever tried to challenge it.
Of course, if you think that’s looney, have a look at #6368227, granted April 2 2002 – “Method of Swinging on a Swing”
Generally speaking, yes, taking a method or machine used in one art, and applying it to a different art, is potentially patentable. The question comes down to “How different are the arts?” If a person skilled in the second art could not reasonably be expected to look at the first art for potential solutions, then it’s probably patentable.
Of course the question of what is “different enough” can be (and often is) argued about during the examination process.
