The key to using a directed search to find your little corner of the prior art is to carefully sand down the rough edges of your inventive concept.  Once you have settled on a relatively broad search query (perhaps by using one or more of the techniques provided in the previous post), you can really hone in on the patentable space by thinking about the specifics of your idea.  Ask yourself: what are those features which are truly central to your invention?

Continuing on the theme of the previous post, let’s say that your idea is an electric unicycle with a series of colored lights on the wheel hubcap which sequentially light up as the vehicle increases in speed.  Perhaps “low speed” lights are in red, “middle speed” lights are in yellow, and “high speed” lights are in green.  As in the determination of any good search query, it is important to select a specific set of search terms which neither unnecessarily narrows your ideas, but doesn’t overly broaden them either.  Indeed, though you may feel that you already have a pretty narrow set of concepts, each individual term may prove surprising problematic.

Begin with a broad search query, such as (electric) (unicycle OR monocycle) -toy* after:publication:20000101.  This returns 5055 results.  From there, you may then tack on some of the critical concepts of your invention.  For example, adding (light*) to the query narrows the number of results to 2810.  Alternatively, adding (wheel*) to the query narrows the number of results to 3259.  As another option, adding (speed) to the query narrows the number of results to 3096.  Adding all three produces 1018 results.

Of course, concepts such as lights, wheels, and speed must certainly be integral components to many electric unicycle innovations.  It will be crucial to narrow your scope further – especially since 1000+ publications presents a lot of material to wade through.  However, as discussed last time, thoughtlessly narrowing your search may omit important pieces of prior art.  For example, adding (hubcap) to all combined terms from above leaves a mere 10 results.  Intuitively, you should be doubtful that this covers all, or even a portion, of the relevant prior art.  Indeed, there are several related words to hubcap which may be utilized instead (hub, wheel cover, wheel trim, etc.)  Further, what if the lights were disposed on the spokes?  Or on the rim of the wheel?  Or capping the lug openings?

What this exercise should impress upon you is that even concepts which may seem essential to an invention do not always constitute good search terms.  What you need to nail down are terms which are absolutely essential in any description of your invention.  For example, adding (color*) (red OR yellow OR green) will almost certainly be necessary for any description of your specified patterning.  Indeed, the search query (electric) (unicycle OR monocycle) (wheel*) (light*) (speed) (color*) (red OR yellow OR green) -toy* after:publication:20000101 returns 86 results.

 You can take one more step and place quotation marks around each of your broadest concepts: (“electric”) (“unicycle” OR “monocycle”) (wheel*) (light*) (speed) (color*) (red OR yellow OR green) -toy* after:publication:20000101) to obtain 75 results.  This ensures that close terms which Google characterizes as synonymous – “electrical,” “monocyclic,” etc. – are not returned.

Of course, it is still essential to have your attorney review, as there are further aspects of the field which may be more difficult to characterize.  This initial searching may still prove useful, however, as a selection of close prior art from your search may help your attorney grasp your specialized field.  Indeed, by analyzing the detailed description and prosecution history of each publication, your attorney will be able to make a more informed judgment call on the likelihood of your patent application being issued.  Further, the search strategies presented thus far don’t even begin to consider obviousness rejections.  However, armed with a few close publications, your attorney may be able to predict which combinations of the prior art are likely to be made by examiners to reject your application and where the patentable space actually exists.

You may have noticed that there has been no discussion of really generic language cases which may recite your invention (e.g., “electric vehicle having one wheel”).  Nor have we discussed how to effectively search non-English language prior art.  Next, we will see how you can rigorously address both of these situations by combining CPC subclasses with forwards/backwards searching.

Comment