[OT] patent search process for a new product

Hi, All

Sorry for the off topic. I don't really know where I should ask.

We are building a new audio product. It uses a number of audio
processing algorithms. Some of these algorithms were derived from
recent publications. We wanna make sure we don't infringe any patent.
What should I do? Do a keyword search on USPTO? What if I miss some
"relevant" patents?

We are not interested in filing our own patent, just wanna make sure
nobody will go after us for patent issue.

Thanks

-syl

On Sep 6, 3:10 pm, SYL <sya...@gmail.com> wrote:
> Hi, All
>
> Sorry for the off topic. I don't really know where I should ask.
>
> We are building a new audio product. It uses a number of audio
> processing algorithms. Some of these algorithms were derived from
> recent publications. We wanna make sure we don't infringe any patent.
> What should I do? Do a keyword search on USPTO? What if I miss some
> "relevant" patents?
>
> We are not interested in filing our own patent, just wanna make sure
> nobody will go after us for patent issue.
>
> Thanks
>
> -syl

Hello Syl,

Unless you are a very experienced patent searcher, your lack of
finding patents that cover your implementation won't tell you much. If
on the other hand you do find such patents, then you are either
infringing on them or you may be okay - check the patent's dates. For
example MIT's patent on the RSA encryption algo is now expired. So all
can use it freely. Patents used to be in force for 17 years - now they
enjoy 20.

Usually for this kind of thing you need to get a patent attorny who
has seasoned researchers to look into this for you - yeap it will cost
some bucks.

Since you said that some of the algos come from recent publications,
these algos could have been submitted already for patenting, but they
won't likely be visible in the patent system until issued. Patent
applications are rarely published.


Unfortunately you have the problem of trying to prove the negative.

Of course if you are making a software based product, don't disclose
how you are doing it, but you run a risk. Is it worth it? If you find
out you are infringing someone's patent, you need to work out a deal
with them. If you knowingly infringe a patent you are subject to
triple damages (In the US).

Clay

SYL wrote:

> Hi, All
> 
> Sorry for the off topic. I don't really know where I should ask.
> 
> We are building a new audio product. It uses a number of audio
> processing algorithms. Some of these algorithms were derived from
> recent publications. We wanna make sure we don't infringe any patent.

Pretty much everything is patented, especially if it is has any relation 
to consumer market. But a patent is only worth money and effort that the 
holder of the patent is willing to spend protecting it. So the real 
question is what patents are owned by your competition.

> What should I do? Do a keyword search on USPTO?

Yes, to prepare yourself for a competent conversation with a patent lawyer.

> What if I miss some
> "relevant" patents?

It depends. At worst, your company can be kicked out of business.

> We are not interested in filing our own patent, just wanna make sure
> nobody will go after us for patent issue.

It is only the matter of how much money is involved. The reality is 
every successful company has to fight dozens of lawsuits on patents and 
other issues.



Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com

"google patents" is your friend: http://www.google.com/patents, unless you
have access to a professional search engine (and even then).

Don't let it bother you if you feel that you don't know where to start...
just keep going. 
Your advantage is that you understand exactly what you are looking for.
Getting some hired help later on is maybe needed, but it may be much more
efficient if you can put some leads on the table.

some hints (I'm not really an expert)
- Jump straight to claim 1
- If you find a patent that comes close, search for other patents that
reference it
- And the other way around, if you find something "close enough", trace
the references.
- Find the relevant patent families / classification

-Markus 

On Sep 6, 3:10 pm, SYL <sya...@gmail.com> wrote:

> We are building a new audio product. It uses a number of audio
> processing algorithms. Some of these algorithms were derived from
> recent publications. We wanna make sure we don't infringe any patent.
> What should I do?

For the recent publications, at least, I believe that my first step
would be to contact the authors to find out whether their intellectual
property is protected by patents or patent applications.

For any others, start with a keyword search at USPTO.gov.  That will
lead you to other keywords, and so on.  Standard search practice.

-- Greg

On Thu, 06 Sep 2007 12:10:10 -0700, SYL <syanli@gmail.com> wrote:

>Hi, All
>
>Sorry for the off topic. I don't really know where I should ask.
>
>We are building a new audio product. It uses a number of audio
>processing algorithms. Some of these algorithms were derived from
>recent publications. We wanna make sure we don't infringe any patent.
>What should I do? Do a keyword search on USPTO? What if I miss some
>"relevant" patents?
>
>We are not interested in filing our own patent, just wanna make sure
>nobody will go after us for patent issue.
>
>Thanks
>
>-syl

First, hire an attorney familiar with Intellectual Property issues
before you do *any* searches.   I'm not an attorney, so the following
should be taken for what it's worth (i.e., nothing):

There is a downside to doing searches, which is that if someone can
demonstrate that you willfully violated a patent the damages are three
times what they'd be if you violated it out of ignorance.  The fact
that you do searches means that someone who suspects that you've
infringed their patent might be able to convince a judge that you
willfully infringed.   If, however, you never, ever, do patent
searches out of a matter of policy, you can be immune from any
prospect of treble damages.

As has been mentioned, there are zillions, nay, bazillions of patents
covering all sort of mundane things that may or may not be relevant to
what you're doing.   I don't think one could complete a genuine
comprehensive patent search in any given area without investing years
and a lot of money in attorney's fees deciphering claims.   Many
claims are deliberately written to be somewhat vague in order to
maximize the likelihood of being able to be interpreted to cover a
relevant topic or instance.  Who can interpret it?   The only
interpretation that matters is the one arrived at by a judge, so until
then you're just making estimates.

So, in my non-attorney worth-nothing opinion, the prospects of
actually being able to do a comprehensive patent search such that
you'll know every patent which could be used to claim infringement
against your system are pretty grim.  I don't personally think it's a
very good investment of time, since for every patent you actually
find, there's another obscure one that you probably won't, and they're
all potential litigants against you.   So carefully consider what you
actually want to achieve from a patent search process and consult an
attorney as to whether that would actually be worthwhile in your case,
because there's a potential serious downside to doing any searches at
all.

If you do searches, do them from a computer in a library or some other
public place.   I strongly suspect that many of the on-line patent
search engines (e.g., IBM's) are not offered for free for altruistic
reasons.   IBM has one of the largest patent portfolios in the world,
and if they can show that you accessed one of their patents and then
infringed, you're on the hook for treble damages.

Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.ericjacobsen.org

Eric Jacobsen wrote:
> On Thu, 06 Sep 2007 12:10:10 -0700, SYL <syanli@gmail.com> wrote:
> 
>> Hi, All
>>
>> Sorry for the off topic. I don't really know where I should ask.
>>
>> We are building a new audio product. It uses a number of audio
>> processing algorithms. Some of these algorithms were derived from
>> recent publications. We wanna make sure we don't infringe any patent.
>> What should I do? Do a keyword search on USPTO? What if I miss some
>> "relevant" patents?
>>
>> We are not interested in filing our own patent, just wanna make sure
>> nobody will go after us for patent issue.
>>
>> Thanks
>>
>> -syl
> 
> First, hire an attorney familiar with Intellectual Property issues
> before you do *any* searches.   I'm not an attorney, so the following
> should be taken for what it's worth (i.e., nothing):
> 
> There is a downside to doing searches, which is that if someone can
> demonstrate that you willfully violated a patent the damages are three
> times what they'd be if you violated it out of ignorance.  The fact
> that you do searches means that someone who suspects that you've
> infringed their patent might be able to convince a judge that you
> willfully infringed.   If, however, you never, ever, do patent
> searches out of a matter of policy, you can be immune from any
> prospect of treble damages.

Is that really a defence? "I couldn't be bothered looking" sounds pretty 
willful ignorance to me. "I looked and couldn't find a match" sounds a 
much more defendable position, as patents are never worded in ways that 
are easy to fathom. After the lawyers have finished, I don't even 
recognise my own patent applications. :-\

Regards,
Steve

Eric Jacobsen wrote:

(snip)

> There is a downside to doing searches, which is that if someone can
> demonstrate that you willfully violated a patent the damages are three
> times what they'd be if you violated it out of ignorance.  The fact
> that you do searches means that someone who suspects that you've
> infringed their patent might be able to convince a judge that you
> willfully infringed.   If, however, you never, ever, do patent
> searches out of a matter of policy, you can be immune from any
> prospect of treble damages.

That might be true.  I am also not a lawyer.  I think I was told
once, though, that if you do know about a patent that you might
as well read it.  It may be so well known that the judge would
assume you had read it, and you can at least try to avoid conflict.

(snip)

> If you do searches, do them from a computer in a library or some other
> public place.   I strongly suspect that many of the on-line patent
> search engines (e.g., IBM's) are not offered for free for altruistic
> reasons.   IBM has one of the largest patent portfolios in the world,
> and if they can show that you accessed one of their patents and then
> infringed, you're on the hook for treble damages.

Stories I used to hear about IBM was that they would patent things to
protect against others suing them, but they didn't tend to sue.
(I heard that not long after they bought the patent for virtual
memory.)

Another story I heard was that company A found out that company B was
infringing on a patent.  Instead of suing company B they instead sue
Xerox.  No point in suing someone who doesn't have any money!

-- glen

glen herrmannsfeldt wrote:

   ...

> Another story I heard was that company A found out that company B was
> infringing on a patent.  Instead of suing company B they instead sue
> Xerox.  No point in suing someone who doesn't have any money!

Did they win?

Jerry
-- 
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Jerry Avins wrote:

> glen herrmannsfeldt wrote:

>> Another story I heard was that company A found out that company B was
>> infringing on a patent.  Instead of suing company B they instead sue
>> Xerox.  No point in suing someone who doesn't have any money!

> Did they win?

I think so, but it was some time ago that I heard about it.
It wouldn't have been a good story if they lost.
(It was part of a conference talk, not a personal discussion.)

-- glen