« Something To Be Said About Paying Your Bills | Main | BlogThink 2005 »
October 09, 2005
How to Write a Patent Application – Deconstruction and Rebuilding Approach
There are several different styles of patent applications, each being a certain response to a specific business situation and prosecution strategy.
I use a deconstruction and rebuilding approach, detailed below, for applications where the prior art is reasonably well known, a product is reasonably well known or anticipated, and thus the claims can be well tailored.
For example, if the inventor knows the art very well and has worked at or near the top of the field for a period of time, I have reasonable confidence that they can distinguish their invention and the prior art. Also, if I have a reliable search done using a draft set of claims, we may be confident enough in the prior art to know what would reasonably be allowed by the Patent Office.
The method for drafting the patent application is quite simple: deconstruct the invention down to the claims, then rebuild the description to match the claims. This method results in an application that is very tight and fully supports each element of the claims very well, but it might not have much fat or excess that could be used as backup claims if the original claim set must be significantly altered. These applications are also not specifically intended to spawn many continuation applications.
After doing a thorough interview, I spend a considerable amount of time drafting claims and getting the claims exactly right. Some inventors are willing and able to help with this process, where we winnow through each limitation and make sure it is proper.
Once the claims are finalized, I draw several figures that include limitations from every independent claim.
I detail each element with as much detail as I can in the specification. The actual implementation of the invention is usually shown as one embodiment, but I include every possible variation I can think of for each element or combination of elements. At this stage, I am following the script of the claims, rather than the script of the inventor’s disclosure, interview, or design methodology.
The subtle process of drafting the claims first, then writing the figures and specification adds a layer of abstraction to the invention. Rather than explaining how a specific implementation operates, like I would if I were using the inventor’s description of the invention, I am crafting a document that supports, expounds, and explains the elements of the claims, whatever those claims may be. Depending on the prior art, the claims may be very broad and have wide application, or may be narrow and directed toward a specific application. In either situation, the patent application is crafted directly around those claims and not around the inventor’s embodiment of the invention.
The resulting document should withstand most attacks by the PTO or by litigators, at least from a patent drafting standpoint.
Most inventors, especially the talented ones, see this process as adding a tremendous amount of value to their idea. Further, it gives the client the maximum amount of coverage for the minimum amount of cost.
Posted by krajec at October 9, 2005 03:10 PM
Trackback Pings
TrackBack URL for this entry:
http://www.krajec.com/mt/mt-tb.pl/190
Comments
Hi Russ,
I'm not sure if you can answer this question, but I am interested to know if you have any thoughts on a patent drafting approach for patents that you are sure will be attacked in an infringement action? My guess is that there is no real way to be sure that a patent will withstand a full attack during trial. The closest thing I can think of is to encourage the inventors to be as close to 100% convinced as possible that their invention is novel and non-obvious (i.e., they are so informed of all of the possible prior art as they can possibly be.)
Todd
ipcounsel.blogspot.com
Posted by: Todd at October 21, 2005 10:33 AM
Tim:
I really like this method for exactly the circumstances you suggest. The key is to know the art well enough that the claims are the proper scope, then spend all the effort to support those claims, with minimal extra information.
There is no way to know exactly what the prior art could be that could cause problems, because no search can be exhaustive, plus you cannot get to the applications that are within their 18 month window before publication. But very talented and well versed inventors that know their industry may be the next best thing.
Russ
Posted by: Russ Krajec at October 23, 2005 08:51 AM