PDFs and Alaska - how to get more e-filings
When the USPTO's e-filing initiative began some four years ago, the assumption was that e-filed patent applications would be XML documents. XML tagging would permit USPTO systems to extract any desired information (e.g. claim language, abstract) from the XML file for various uses. Publication would be easier given the computer-readable format provided by the filer, and preparing the issued patent would likewise be easier. The Patent Office would be able to develop process-flow systems which would take full advantage of the XML tags.
One of the institutional goals of the USPTO was (and is) to get lots of people to file their patent applications electronically rather than on paper.
The USPTO developed two XML authoring tools, TSA and PASAT. The XML-tagged output from either of these tools is provided to ePave which generates an application data sheet and fee transmittal, and assembles everything into a submission package which may be e-filed with the USPTO.
TSA has not come to be widely used and USPTO is phasing it out. User experience with PASAT has been somewhat varied.
Most importantly, even after several years of PASAT, TSA, and ePave, and after a lot of outreach by the USPTO to would-be e-filers, the fraction of applications being e-filed is still not as high as the USPTO would like. Possible contributing factors include:
The USPTO has developed a new XML authoring tool, EFS-ABX, which is now in beta testing by pilot users. EFS-ABX may help to overcome the second factor but will not necessarily overcome the first factor.
What, then, can be done to bring about a substantial increase in the fraction of US patent applications being e-filed? How about a discount for e-filed applications, or a surcharge for paper-filed applications? Anecdotal reactions from e-filers suggests that even a $500 difference in filing fees might not make much of a change in e-filing habits. Our firm, for example, averages an extra $1000 or so in professional time to e-file as compared with paper filing. This paper suggests two approaches for increasing the fraction of e-filings: permitting the use of PDF specifications, and changing the official time for e-filings from Eastern Time to Alaska Time.
PDF specifications. Over the past few years, participants in the EFS list server have brought up several times the question whether it might worth considering to allow filers to e-file using PDF formatted patent applications. (See e.g. Peter Rashid, Rader, Fishman & Grauer, PLLC, January 30, 2001, Scott Krieger, Sharp Laboratories of America, November 29, 2001, Jonathan Grad, Vidas, Arrett & Steinkraus, P.A., February 18, 2002, Jim Henry, Blakely, Sokoloff, Taylor & Zafman LLP, March 19, 2002.) Ed Hotchkiss of Perkins Coie LLP noted a June 27, 2002 speech by Director Rogan suggesting that the USPTO might adopt the EPO's e-filing system which permits the use of PDF patent specifications.
Now WIPO is beta-testing its PCT-SAFE e-filing software. (You can see a comparison of the PCT-SAFE software and the patent e-filing software from the USPTO.) This software permits the filer to file with either an XML-tagged specification or a PDF specification, depending on the particular Receiving Office being used. Several PCT Receiving Offices, among them the International Bureau of WIPO, the European Patent Office, and the French Patent Office, have said they will accept PDF-formatted PCT e-filings.
Many US patent practitioners have told me that if there were a way to e-file US patent applications by means of PDF specifications they would immediately file that way.
One possible predictor of customer reaction if PDF specifications were to be permitted may be seen from the adoption rates for e-filed patent assignments. The percentage of patent assignments being e-filed is much higher than the percentage of patent applications being e-filed. The two e-filing processes have much in common -- both use ePave, both require a user to hand-type bibliographic information, both require a user to provide payment information and fee calculation, and both involve use of a cryptographic certificate. The chief difference between the two e-filing processes is that the patent e-filing process requires XML tagging by the user, while the assignment e-filing process does not. For assignment e-filing it is only required that the user scan a document and attach the scanned image file to the ePave project. This suggests that if image-based (PDF) application filing were permitted, the penetration rate for e-filed patent applications might well rise to the level of penetration for e-filed patent assignments.
Indeed the one way in which this comparison (e-filed applications and e-filed assignments) fails is that there are only two ways to file applications (electronic or paper) while there are three ways to file assignments (electronic, fax, and paper). If fax were not available for filing assignments, I suggest many or most of the would-be fax filers would e-file their assignments instead. So I suggest the likely consequence of permitting PDF patent e-filings would be a penetration rate that is related to the sum of the e-filed assignment percentage and the fax-filed assignment percentage.
Note also that e-filed IDSs are now extremely popular and represent a large proportion of all IDSs. An e-filed IDS does not require any XML tagging by the user, yet it does require the user to master ePave and fee payments and a cryptographic certificate, and none of these requirements seems to get in the way of e-filing IDSs.
From all of this I can suggest that if we try to identify factors helping to explain why very few US patent applications are being e-filed, such factors do not include ePave or crypto certificates. A chief factor seems to be the time and expense of XML tagging.
But isn't it better in the long run (for the Patent Office) to have filings in XML rather than PDF? From the beginning I felt XML filing to be better for the Patent Office. Examiners could easily plug claim language into search tools. Patent publication would be cheaper and faster.
More recent events prompt me to feel differently. First, the USPTO now uses a patent application image management system that is purely image-based and which (so far as I can discern) makes no use of the XML tagging in an e-filed patent specification. Second, present-day amendment practice wholly discards the entirety of the claims-as-filed the moment that even one word is changed in a single claim. The amended claims are on paper (image-based). If the filer spent time and money going to the trouble to XML-tag the claims-as-filed, this time and money is discarded the moment any claim is amended.
Third, however important XML tagging is or may be at some future time, the poor penetration rate of e-filing seems a more pressing and urgent problem, and perhaps postponing XML tagging is a sacrifice worth making if the penetration rate for e-filing can be greatly improved.
I suggest that if the USPTO were to make PDF e-filing available, this should be accompanied by explicit and frequent reminders to users that it is still intended to move gradually to XML-tagged filings. There is no more potent reminder than an explicit fee structure -- one price for XML-tagged specifications, a higher price for PDF e-filings, and a third and still higher price for paper filings. Even if the price differentials were small at first, they could be increased later to incent users to continue toward a greater proportion of XML-tagged filings.
I note that WIPO has adopted this approach. At present, a PCT-EASY filing (in which the user saves the patent office from having to key in the bibliographic data) is cheaper than a paper filing. WIPO has said that when PCT-SAFE is released to the public, there will likewise be price differentials (XML vs. PDF, and e-filed versus paper). Such price differentials will provide an easy path toward XML incentives in future.
Alaska. Talk to any patent lawyer and he or she will tell you that it is the rarest of inventors and the rarest of clients who make it possible to file a patent application sooner than the last possible day. For most patent lawyers, the majority of patent applications are filed on the last possible day, and in many cases there are hurried communications between lawyer and client resulting in revision of the to-be-filed patent application up to the last minute prior to filing.
Present-day paper-filed US patent applications, under 37 CFR § 1.10, enjoy a filing date based upon the time of day at the filer's local post office. In contrast, present-day e-filed US patent applications receive only a filing date based upon the time of day in Virginia. For filers in US time zones that are west of Virginia, this means that it is always possible to get a better filing date by paper-filing than by e-filing. It virtually guarantees that a large fraction of US patent filings will always be done on paper since any decision to e-file results in an earlier cutoff of the work on the patent application. On some fact patterns it might well constitute legal malpractice to e-file when paper filing is also an option.
Deeming ePave filings to have been done based upon the time of day in Alaska would reverse matters and would give an extraordinary advantage to e-filings over paper filings. I suggested this to Director Q. Todd Dickinson in December 2000. I believe such a change would have two salutory effects -- it would motivate those who have not learned how to e-file US applications to learn how to do so, and it would motivate those who know how to e-file US applications to file many more of their applications electronically.
A very compelling example of how "getting a better filing date" can be a potent motivator to e-file rather than on paper, see the present practice of the USPTO regarding paper-filed and e-filed US trademark applications. By means of a rule change, the USPTO discontinued the availability of 37 CFR 1.10 for Express Mail filing of US trademark applications. Plainly put, an e-filer gets a better filing date than a paper filer so far as trademark applications are concerned. This is surely one of several factors helping to explain why the penetration rate for US trademark e-filing (by now well over 50%) is as high as it is.
What to do next. I have regular dialogue with many dozens of USPTO e-filers through the EFS list server and through AIPLA. I strongly believe the following to be true: