Identifying “pegged” art units at the USPTO

USPTO used to “peg” its patent examination backlog at 108 months (nine years).  Then in about 2014 (see blog article) the USPTO changed the “peg” to 60 months (five years).  Now in 2018 the USPTO has changed the “peg” to 30 months (two and a half years).


I guess what is going on here is that the USPTO has been fairly successful at reducing some of the backlogs for some of the art units.  It surely must have been embarrassing for the USPTO in those old days when there were plenty of art units with backlogs exceeding six years (or even nine years).  But indeed in recent years USPTO has been making quite a lot of progress.

USPTO provides a First Office Action Estimator where you can plug in an art unit and it will give you an indication of the backlog.  You can plug in one art unit after another, and these days you will never get an estimate that is longer than 30 months.  The USPTO software clearly has a line of code that “pegs” the prediction at 30 months no matter how big the backlog might really be.

Back in 2014, USPTO actually offered a spreadsheet which a member of the public could download, listing all of the art units and, for each art unit, the backlog.  You could then sort the spreadsheet and identify the “pegged” art units (the ones for which the backlog was 60).  The spreadsheet never showed a number bigger than 60.  You could also sort the other way and identify the art units with the smallest backlogs.

But that spreadsheet seems to be gone from the USPTO web site.  To find out how things stand, a member of the public would have to go to the estimator (linked above) and plug in every art unit one by one to get the answers.  I don’t have the energy to plug in every art unit one by one.  But I did plug in a few art units at random and found some backlog figures:

  • 1631 (biotech) – 25 months
  • 1761 (organic chemistry) – 14 months
  • 2411 (multiplex electronics) – 8 months
  • 2464 (telecommunications) – 30 months
  • 2491 (cryptography) – 30 months
  • 2641 (telecommunications) – 8 months
  • 2811 (semiconductors) – 13 months
  • 2911 (designs) – 11 months
  • 3691 (business methods) – 30 months
  • 3731 (medical devices) – 21 months

I clicked around for a bunch of our firm’s pending cases and I kept running into thirty-month predictions over and over again.  I guess that’s just how it works out depending on the technology area of a firm’s clients.  It’s just that some art units (2464 and 2491 and 3691 are examples) are “pegged”.  The backlog for such an art unit is so big that the USPTO simply does not report the true backlog but instead lists the number “30” as the predicted backlog.

We also have cases in art units 3692 and 3693 and those art units are “pegged” at 30 months.

In our firm we almost never actually see a case run all the way to the “peg” because we use Patent Prosecution Highway a lot.  Most of our pegged cases end up on the Highway and we get an Office Action much, much sooner than one would expect from the USPTO’s First Office Action Prediction.

For someone whose case is in one of the “pegged” art units, and whose case is not on the Highway, I hate to think how long they might have to wait for an Office Action.

If you have a case in an art unit where the FOAP is 30 months (or more?) could you please post the art unit number in a comment below?

One Reply to “Identifying “pegged” art units at the USPTO”

  1. As I recall, back in the day, the Gazette used to indicate the filing date of the oldest case awaiting action by unit. Now, it gives the average filing date of cases having received a first action. Since many of those currently have average dates in 2014, it is clear that the 30 month figure is fiction.

    The oldest case awaiting action was important in the absence of PAIR since, absent making an inquiry to the PTO, it was the only way of knowing if a case you were waiting on was getting close to examination or whether an office action had been missed.

Leave a Reply

Your email address will not be published. Required fields are marked *