[sci.electronics] Patents

sparks@corpane.UUCP (John Sparks) (01/12/90)

Could someone give me a general rundown on how to apply for a patent?

Let's say I had an idea for an invention, what is needed to get a patent?
Do I just have to have a detailed description, or is an actual working model
necessary? 

If just a detailed description, how detailed? Are details of parts and how
they interact enough or is a full schematic needed?

The thing I am thinking of consists of various sensors (already existing types)
and a computer (any kind will do) that serve a specific function. That is: The
way the parts are put together and used is the invention, not the individual
parts. Can I patent such a thing?

What is done about using any parts that are already patented (the sensors)?

Do I have to detail just the parts involved and how they are used in the
gadget, or do I have to actually have to get into the programming and 
electronics of the computer and sensors?


-- 
John Sparks  | D.I.S.K. 24hrs 1200bps. Accessable via Starlink (Louisville KY)
sparks@corpane.UUCP <><><><><><><><><><><> D.I.S.K. ph:502/968-5401 thru -5406  
Don't take life too seriously. You'll never get out of it alive.

hollaar%basset.utah.edu@cs.utah.edu (Lee Hollaar) (01/13/90)

In reply to sparks@corpane.UUCP (John Sparks):

Could someone give me a general rundown on how to apply for a patent?
   Obviously these answers will be very general.  While conceptually
   simple -- the United States Government grants you for a period of time
   (17 years for a design patent) the right to exclude others from making,
   using, or selling your invention provided you are the first inventor
   and comply with all the requirements of the patent law (35 USC) and
   rules (37 CFR) -- there are a number of gotcha's in the procedures that
   can result in the loss of all or substantial patent rights if not
   handled properly.  Anyone seriously considering applying for a patent
   should get help from a registered patent attorney or agent.  (By the
   way, I'm a registered patent agent, but I'm not soliciting work.)

   For a good general introductions to United States patents, you can get
   the publication "General Information Concerning Patents" from the
   Government Printing Office.

Let's say I had an idea for an invention, what is needed to get a patent?
Do I just have to have a detailed description, or is an actual working model
necessary? 
   You always need a detailed description (37 CFR 1.71), but only need a
   working model if the Patent Office feels it is necessary (37 CFR 1.92).
   About the only time they feel it is necessary is when they don't believe
   the invention actually works, such as for a perpetual motion machine.

If just a detailed description, how detailed? Are details of parts and how
they interact enough or is a full schematic needed?
   From 37 CFR 1.71: "The specification must include a written description
   of the invention or discovery and of the manner and process of making
   and using the same, and is required to be in such full, clear, cocise,
   and exact terms as to enable any person skilled in the art or science
   to which the invention or discovery appertains, or with which it is most
   nearly connected, to make and use the same."

   The idea is that you have to give complete instructions on how to make
   and use the invention, so that others can build it after the patent
   expires based only the information contained in your patent.  The Patent
   Office rejects applications that don't contain sufficient detail.

   The Patent Office also doesn't allow the introduction of additional detail
   into an applications (new matter -- 37 CFR 1.118).  You would have to file
   a new application containing the new matter, but there are a number of
   time requirements for filing an application (35 USC 102) that may have
   lapsed by the time the Patent Office tells you your application does not
   have enough detail.  (Many of the time requirements are one year, while
   the Patent Office often doesn't start detailed examination for over a year
   after it is filed.)  This is why it is important that the first application
   be complete and properly written -- you can lose most or all of your
   possible rights if it isn't.

   You can also claim your invention (the claims are the part where you
   indicate exactly what is yours) in terms of "means providing a function"
   ("means for attaching two widgets" rather than "a screw attaching two
   widgets", with the latter also including glue, chewing gum, etc.).  This
   is another example of how someone experienced with patent procedures can
   help you get all the rights you deserve.

The thing I am thinking of consists of various sensors (already existing types)
and a computer (any kind will do) that serve a specific function. That is: The
way the parts are put together and used is the invention, not the individual
parts. Can I patent such a thing?
   Virtually every patent is for something consisting of previously known
   parts, but organized in a non-obvious way.  The key word is non-obvious,
   which is covered by 35 USC 103.

What is done about using any parts that are already patented (the sensors)?
   A patent is a right to stop others from using an invention, not a
   permission to practice an invention.  Unless you get a license from
   the patentee of the sensors, you can't practice your invention until
   the sensor patent expires, but nobody (including the sensor patentee)
   can practice your invention without your permission until your patent
   expires.  If the sensors are regularly sold, and you buy them from their
   patentee or somebody licensed by their patentee, you have no problem.

   As far as applying for the patent, that parts you use are already
   patented makes no difference.  The test will be if you are using them
   in a non-obvious way.

Do I have to detail just the parts involved and how they are used in the
gadget, or do I have to actually have to get into the programming and 
electronics of the computer and sensors?
   You have to be detailed enough so that somebody skilled in the area
   can build your invention.  You also have to describe the best way of
   implementing the invention, not just any general way.  If you don't,
   your application may be rejected, or a court can invalidate your
   patent.


   Again, if things are not properly done, you can lose your right to a
   patent or lose substantial protection for your invention.  If you really
   think your invention has market potential, get qualified help.

mmm@cup.portal.com (Mark Robert Thorson) (01/13/90)

You don't need to build a working model, unless requested by the Patent
Office.  (Here I'm assuming you haven't designed a perpetual motion machine.)
You must show "reduction to practice", which could be a model but could also
be "conservative reduction to practice" (i.e. a paper design).  The latter
requires that the design is detailed enough to allow someone "skilled in the
art" to understand and use it.  (Quoted phrases are important legal terms.)

You should immediately commit it to paper and have it properly witnessed,
because proof of early invention is important to obtaining a U.S. patent.
Unlike most countries, the U.S. is not a "first to file" nation.

Your collection of sensors and a computer running customized software seems
like a patentable apparatus to me.

The next step would be to do a patent search.  There are a couple dozen
places around the U.S. where patent collections are available to the
public.  They're listed on the inside cover of a magazine called the
OFFICIAL GAZETTE, which you can find at any large university library.
The magazine is published by the Patent Office, and has abstracts from
all newly-issued patents.  You can almost do a patent search in a large
collection of the Official Gazette, but it's much better to go to a
patent library.  They will be able to tell you how to do the search.

peg@psuecl.bitnet (PAUL E. GANTER) (01/13/90)

> The next step would be to do a patent search.  There are a couple dozen
> places around the U.S. where patent collections are available to the
> public.  They're listed on the inside cover of a magazine called the
> OFFICIAL GAZETTE, which you can find at any large university library.
> The magazine is published by the Patent Office, and has abstracts from
> all newly-issued patents.  You can almost do a patent search in a large
> collection of the Official Gazette, but it's much better to go to a
> patent library.  They will be able to tell you how to do the search.

Apparently most large university libraries have patents on microfilm and
partially on CD-ROM.  Here at Penn State, we have a nice computer system
that will search the CD-ROM's in various ways, giving you patent numbers
to refer to and occasionally brief descriptions.  I recall reading that
some idiot was stealing microfilms of patents from universities around the
country, so I assume most other universities have the films (note: an "idiot"
because you can buy the microfilms at fairly low cost).

Good luck!

Paul

Nagle@cup.portal.com (John - Nagle) (01/14/90)

     As an inventor who gets royalties on patents, I can reply to this.

     First, a good starting point is the classic "Patent It Yourself",
from Nolo Press.  It has its flaws, but will give you an idea of what
patents are all about and the mechanics of filing them.  It is NOT
current as to the international form now used, but is OK if you don't
intend to file internationally.

     For a simple patent, if you let a patent lawyer do the whole
job, it will cost a few thousand.  You can get this down considerably
if you work at it.  It's a good thing to have a patent lawyer look over
your application, but if you can write clear technical prose, good
enough for publication, there is no reason you shouldn't write most 
of the patent yourself.  The "claims" section of the patent must conform
to very strict rules and needs the hand of a lawyer, but even here,
you should provide the first draft.

     Patent drawing requirements are no longer as strange as they used
to be, and you can do a quite acceptable job with a CAD or drawing system
and a laser printer, instead of the classic India ink on Bristol board.
The style rules must be carefully followed, but they are not all that
complicated.  The whole job can be sent out to a patent draughtsman,
but it's not really necessary.  

     The patent search you should do yourself, and the Nolo Press book
can tell you how to go about this.  After you've done it yourself, a
second go-round with a patent search firm, using as a starting point
all the patents you found in your search, is worthwhile to catch anything
you missed.  It's easier and cheaper for a patent search firm to start
from such information than from your description.

     Always bear in mind that most patents don't bring in any money.

					John Nagle

davidc@vlsisj.VLSI.COM (David Chapman) (01/17/90)

In article <1311@corpane.UUCP> sparks@corpane.UUCP (John Sparks) writes:
>Could someone give me a general rundown on how to apply for a patent?
>
>Let's say I had an idea for an invention, what is needed to get a patent?
>Do I just have to have a detailed description, or is an actual working model
>necessary? 

You need only a description to get a patent.  This is why the Patent Office
gets applications for perpetual motion machines.  :-)

>If just a detailed description, how detailed? Are details of parts and how
>they interact enough or is a full schematic needed?

The patent application needs to be detailed enough that one "skilled in the
art" of whatever field it falls in (Electrical Engineering, here) could
build one from your description.  You tell the world how to build it in
return for a 17-year monopoly on it.

>The thing I am thinking of consists of various sensors (already existing types)
>and a computer (any kind will do) that serve a specific function. That is: The
>way the parts are put together and used is the invention, not the individual
>parts. Can I patent such a thing?
>
>What is done about using any parts that are already patented (the sensors)?

My patent (#4,849,313 :-) is a "Method for Making a Reticle Mask" (for IC
fabrication) such that alignment tolerances can be checked.  It works in
large part because of the way that Electron-Beam exposure systems are
designed.  Obviously I didn't invent them and I don't have to pay the
owners of the patents covering E-beam systems anything.

So the answer to your question is "yes, the assemblage can be patented."

>Do I have to detail just the parts involved and how they are used in the
>gadget, or do I have to actually have to get into the programming and 
>electronics of the computer and sensors?

See "skilled in the art" above.  Beyond that, I'm not qualified to say.

So far I've told you the good news - yes, your invention may be patentable.
The bad news is that it is not easy to get a patent.  The language on a
patent application is legalese at its worst.  You almost have to get a 
patent attorney to file it for you.  Mine was developed in the course of 
my work here at VLSI Technology, so they paid for the patent application 
process, and I'm glad.  A patent search alone (to ensure that no one else
has patented what you want to patent) can cost $300.  You can expect to 
pay $3K (I think) for the overall patent process, and wait up to 2 years 
after filing.

If you want, I can send you a bit more information via E-mail about the 
work I did to get my patent.  If there's enough interest I'll post.

P.S.  I note that distribution on this posting is "usa"; be forewarned that
      patent law is dramatically different in most other nations.  A patent
      here does not protect you against people using your invention in other
      countries!
-- 
		David Chapman

{known world}!decwrl!vlsisj!fndry!davidc
vlsisj!fndry!davidc@decwrl.dec.com