The U.S. Patent Backlog

coondoggie writes "Even with its increased hiring estimates of 1,200 patent examiners each year for the next 5 years, the US Patent and Trademark Office patent application backlog is expected to increase to over 1.3 million at the end of fiscal year 2011 the Government Accounting Office reported today. The USPTO has also estimated that if it were able to hire 2,000 patent examiners per year in fiscal year 2007 and each of the next 5 years, the backlog would continue to increase by about 260,000 applications, to 953,643 at the end of fiscal year 2011, the GAO said. Despite its recent increases in hiring, the agency has acknowledged that it cannot hire its way out of the backlog and is now focused on slowing the growth of the backlog instead of reducing it. This too is but one of the goals of the Patent Reform Act currently making the rounds in the US Senate."

Therefore

[Dunbal (464142)]

I hereby submit a patent to use computing and human resources technology to increase the speed of the patent process.

Pay me, bitches.

Re:Therefore

[The MAZZTer (911996)]

Sorry, you have to wait until your patent is approved... catch 22.

Patent #1094398532 - One click approval

[neapolitan (1100101)]

I patent the "One-click" patent approval process.

I'll license it back to them for a fee.

What they told me

[john_is_war (310751)]

As a graduating computer engineer, I've been interviewing around, and USPTO was one of the places. Here's what they shared with me-
They are currently backlogged 5 years.
With their hiring surge of engineers, they want to bring the backlog to 2 years within 4 years IIRC
And apparently they crap money, with a starting salary of 63k with a 10k starting bonus for the first 4 years, plus a 10% bonus if a 130% efficiency rating is maintained for the 4 quarters.
The ones they are particularly hiring are EEs, CSs, and Comp Engs.

Now you know, and remember- Knowledge is power!

Re:What they told me

[ServerIrv (840609)]

plus a 10% bonus if a 130% efficiency rating is maintained for the 4 quarters.

There is a built in incentive for bad patents to get through. Patents get rubber stamped simply because of the need of efficiency to get out of the whole backlog mess. Instead of actually diligently checking and rechecking for prior art conflicting patents, the employee stamps it as good, as fast as possible, and walks away with their 10% bonus. This seems to be the same problem that tech support has with call tracking. The faster a person gets you off the phone, the more money they make, and the faster they get promoted.

Re:What they told me

[by Anonymous Coward]

> There is a built in incentive for bad patents to get through.

Looking at this job I am almost tempted to apply. The money is good even without the bonus. It would be interesting and varied work. And one would be in fine company, as Mr Einstein himself was once a patent clerk.

But I would last all of 5 minutes before getting fired. The problem is as computer scientist and inventor and someone who knows the difference between an abstract idea and well thought out and unique implementation that solves original problems I would have to apply ethical standards to my work.

Got a business patent? In the bin it goes.
Got an existing idea you want to add the words "web browser" or "computer" to? In the bin.
Only got mathematical algorithm? No language specific implementation? In the bin.
You want to add a tiny specific modification to an existing idea? Sorry, in the the bin.
Something that I can find published on Google with trivial searching? In the bin.

Well, you might think my manager would be happy as a pig in shit, praising a wonderful worker who is clearing the backlog by rigorously applying the rules. Wouldn't you?

Oh, but wait... Where does the patent office make its money? Approving patents.

Which is why the rules were changed to allow all this crap through in the first place. There aren't suddenly more ideas in the world, the bar for patentability has been drastically lowered and the system is broken because of it.

So, how about this for an idea. If he approves a patent that is subsequently challenged and voided the clerk loses twice their bonus and the patent office has to refund the application fee in full. And to make sure the applicant doesn't benefit from specious claims, they must pay a fine of 10 times the application fee to the government.

Then let's see who is so quick with the rubber stamp.

What is needed is incentive to find _good_ patents. To add this incentive, how about a royalty type bonus system. A clerk who approves a patent that runs its full term gets a small bonus at the end of its life, related to how much money it has actually made through the sale of real products (litigation payments would be excluded).

Re:What they told me

[greenreaper (205818)]

The system gets paid for dealing with applications, not for approving them. If they aren't approved, they still keep the money. Same with trademarks.

Re:

[mavenguy (126559)]

...10% bonus if a 130% efficiency rating is maintained for the 4 quarters...

If you enter as a GS-5 your production quota is 60% of the nominal GS-12(100%) production quota for the expectancy assigned to the docket that you will be working in (varies by art). By the time you make Primary Examiner (GS-14) you have to crank out 135% of the GS-12 expectancy. That's a factor of 2.25 more. In a recent GAO Report [gao.gov] recent hires who are leaving the PTO in droves cited "outdated production goals" as one of the leadi

Re:What they told me

[john_is_war (310751)]

Average CS starting is 51k, Comp Eng is 56k, actually.

Re:What they told me

[by Anonymous Coward]

I'm graduating with a BA in CS this year, and barely looking at any CS jobs below 85k, and there are plenty above.

Reality is about to kick you very hard in the groin.

Enjoy.

Re:

[laffer1 (701823)]

He did say average. I think most of you forget about cost of living. I see this mistake at my university all the time. Graduating students take like $45,000 in LA or something because they don't realize that it's not the same as $45,000 in Ann Arbor, MI. Just because location x pays $85k does not mean that location y pays $85k or even that it's the "same" amount to live on.

Some products cost the same in different parts of the country. Gas is fairly close right now across the board. It's cheaper in New

Re:What they told me

[john_is_war (310751)]

From my understanding, the first 6 months, you get 10 patents every week or biweek, unclear on that. Then after 6 months you start getting appeals back from rejected patents (6 months being the max time). And that point, the number of new patents expected is diminished. And of course all the paperwork needs extensive background research for prior art etc. etc.

And as for that, it's not my main choice, but it's better than no job.

Re:

[mavenguy (126559)]

Production is measured in units called "balanced disposals" which is the average, over any period of measurment (bi-week, quarter, fiscal year) of the number of first actions on the merits (N) and the number of disposals (D, which are allowances, abandonments, or examiners answers on appeal). For each of the various art areas a historical expetancy X is assigned as so many hours per balanced disposal. the office wide expectancy for this is a bit over 20 hours per balanced disposal for a hypothetical GS-12 e

old news

[heptapod (243146)]

http://yro.slashdot.org/article.pl?sid=07/10/18/2036215 [slashdot.org]
http://www.washingtonpost.com/wp-dyn/content/article/2007/10/07/AR2007100701199_pf.html [washingtonpost.com]

Examiners don't quite do anything anyway

[CrazyJim1 (809850)]

I think examiners encourage the use of big words in a patent, not so that the patent is unique, but so they can claim ignorance when someone re-patents something in existence. If an investigation comes up then the patent examiner can claim ignorance. A lot of that techno-bable doesn't makes any sense to someone in the field.

Yes, I know.

[Animats (122034)]

I have an application 12 months into a 30 month queue. In a fast-moving field, this is a huge headache. My previous patents only took about a year to get to first office action.

There's a new express program, though. If you file no more than three claims, file online, and do a more diligent search, the USPTO promises to process the patent in less than a year. That was just starting when I filed, and I didn't take that option.

A giant rubber stamp is needed

[DigiShaman (671371)]

"Just rubber stamp it. The judicial branch will sort it out for us."

Is that what it's going to come down too?

Re:

[vtscott (1089271)]

I hope not. I would imagine that it's much cheaper to just have a competent patent examiner with enough time to do his/her research reject a patent than it is to get the courts involved.

Re:A giant rubber stamp is needed

[ls -la (937805)]

"Just rubber stamp it. The judicial branch will sort it out for us."

Is that what it's going to come down too[sic]?

Unfortunately, yes. Since job performance is entirely based on number of applications processed, the examiners have very little incentive to do a good job, so unless they have a clear reason to reject an application in the first 5-10 pages, they'll likely just grant it. The problem then REALLY comes when the judicial branch says, "the patent office granted it, so if it's not patentable they can sort it out," which is what they have been doing for some time now. That's part of the reason it's so difficult to get a patent overturned: both branches say the other should do it.

Obvious Jobs Program

[Doc Ruby (173196)]

It seems to me that the demand for patent examiners and the explosion of patent applications and money derived from them should add up to a lot bigger hire than just a few thousand more examiners. The PTO should charge an annual fee on patents that's a tiny percent of the revenue from their applications or licensing, which if enough to pay for enough examiners should still be under 1% of income under the patents. Then the amount of examiners will keep pace with the growth in the patents they have to examine.

In fact, the growth in patents and their revenue should even stimulate the production of American engineers. Offer full scholarships to engineers, funded by those fees, in exchange for them becoming paid examiners for a couple-few years at least, and returning for at least 6 months every 5-10 years for a couple-few decades. If they break that deal, they owe 2x their scholarship immediately, which can pay for more scholarships and paid examiners.

The patent system has many problems. Primary is that the American people subsidize the creation of intellectual property by paying for the expensive examination and challenge system, which we can ill afford with our current budget problems (and which was never fair to the public, anyway). Also too few examiners of too little quality and commitment. Calibrating a fee to hire examiners and create them by scholarship to the volume of applications should make the system more self-regulating. And good for engineers: Albert Einstein had most of his good ideas working in the Swiss patent office, which no doubt benefited from his talent and imagination. Let's see America protect both itself and its inventors with a simple device that balances both.

You may consider this design to be placed in the public domain :).

Re:

[monxrtr (1105563)]

Pay the public for discovering errors. Mandate that all patent applications be posted for public review. Any discovery of prior art should be the *burden* of those submitting patent applications. Double or triple the patent applications fees. Mandate a $10,000 penalty (which must be deposited with the application fee) to be paid by all applicants who submit patent applications containing prior art or too obvious an idea that is forfeited if their application is rejected for any reason. This $10,000 is up fo

Re:

[Doc Ruby (173196)]

Well, the problem is that increasing the filing fee will lock out small-scale inventors, who often produce quite a lot of the innovation, and who need patent protection a lot more than do the big, rich inventors who could also afford to market the invention before a competitor competes with them even without patent protection.

In fact, patent filing should be free, but would need at least a nominal fee to deter people from applying at ridiculous rates. Remember that the patent is supposed to protect an inven

Re:

[Doc Ruby (173196)]

So what if patents are submitted by non-domestic entities? If they want their patents to stay registered in the US, they'll pay the fees, just like they pay the fees for registration.

Engineers are part of a service industry. They don't actually manufacture anything, they're info workers. Fresh engineers will come in greater numbers when their education budget is less risky from both scholarships and more jobs when they graduate.

Solutions?

[by Anonymous Coward]

How about regional processing centers around the county (if they really are paying engineers 65K/year that goes along way in the midwest but not far in DC area). I would be happy to process patents in my technical field from home, with the proper training and tools prior art and patent serching. Let the academics, engineers, comptuer sci/e people get trained and process patentent part time from home!

A possible solution

[Venik (915777)]

I think in this difficult time for our government all patriotic Americans should refrain from inventing stuff for the next five years.

DELETE WHERE ToLower(body) LIKE '%on the internet%

[RingDev (879105)]

Honestly, taking any existing patent and tossing it on the internet should be tossed immediately as obvious.

Knock out software patents, and patents on processes, and blamo! problem solved.

-Rick

Re:DELETE WHERE ToLower(body) LIKE '%on the intern

[penix1 (722987)]

Far easier...

Deny any patent with the words, "A method to..."

Problem solved. I bet the backlog drops by at least 3/4 what it is today.

why check them at all?

[the cheong (1053282)]

i must be a n00b in the patent process, because i don't understand why we need people to review patents in the first place. why don't we simply publish every patent on some online database, and review patents _only if_ disputes arise? if someone wants to patent something, then he must sift through the patent records _himself_ and make sure he's not infringing on anyone's rights. if a dispute (i.e. lawsuit) should ever be filed, _then_ we check the patent records to verify and take the appropriate course of action. with this system, people who want to patent things will be a lot more careful about their research on prior patents, no? maybe they'll even contact people with similar patents and clear everything up so that no disputes arise in the future? and the cost to the USPTO is simply publishing all the patents online and checking over disputed patents?

Re:

[aj50 (789101)]

Because if I'm some inventor, and I come up with e.g. a top for baby beakers that really doesn't spill when it gets thrown across the floor, get an agreement with a supermarket to fund the manufacture and start producing and selling the things then without patent protection it will be a couple of months before every plastic utensil maker with products marketed at babies is also making them and because they've got a better manufacturing setup and can afford to invest more money in the product than I can, the

Simple solution -- working prototypes

[Derling Whirvish (636322)]

There is a simple solution to the problem of too many patents to examine. Go back to requiring a working prototype. You would have to supply working source code for any software patent. Business "methods" patents would not be acceptable unless you could demonstrate the method actually in use.

So Arthur C. Clarke would not have been able to patent the idea of geostationary satellites. He didn't so nothing was lost. Were current patent procedures in place in the late 40s, most certainly a patent troll would have patented it. But what harm is there in forcing comeone to actually get a satellite to geostationary orbit before allowing a patent? It would certainly encourage research and development rather than litigation and argument. Forcing Edison to actually get a filament that worked before granting him a patent on the light bulb worked out for the better, rather than allowing him to patent the "idea" of using an electrically heated filament to generate light. If he had gotten the patent without the working model he could have sat back and just sued anyone implementing electic lights for the next 17 years. It would have set back progress tremendously.

How to clear the backlog

[pokerdad (1124121)]

Step 1: Hire more patent officers.

Step 2: Raise the price of applying to meet costs of #1.

Step 3: Once backlog is clear create stricter application process.

Step 4: Based on number of man hours required for new process introduced in 3, raise prices again.

Step 5: Review demand now that it costs more and is less likely to success; adust staffing to meet new demand.

Broken system, but I have an idea.

[MichailS (923773)]

I have a suggestion: how about scrapping the current concept of patents, and instead award time-limited exclusive rights to entities that SHOW A PRODUCT USING THE COVETED TECHNOLOGY instead of just filing a paper?

I never managed to wrap my head around the fact that I can own the rights to almost anything - as long as nobody else did it first - by just having to file for an application to verify this and pay for the process.

In the next step somebody writes the application as vaguely as possible to give away as little information as possible while trying to grab as much as possible. Then someone will stare incredulously at my application with a stamp twitching in the hand, while tics cause their cheek to spasm. One second and an exaperated curse at the incomprehensible text later, WHAM, I am awarded a billion dollar paper that says I own something I may have never conceived, touched or even spent many minutes pondering about.

Show me an invention that isn't obvious to the expert! They exist, of course - in abundance - yet probably make up for a microscopic fraction of all the patents. But most of the time evolution and developemnt stand on the shoulders of giants and your expert peers will say "Yeah, I thought about that years ago, I just never made anything about it" about your inventions.

Thus, just procuring an idea on paper should not be enough to get a patent. You should also be able to demonstrate that you are actually UTILIZING the concept!

Simple Fix 1, 2, 3

[EQ (28372)]

1. Pass legislation reversing the court the ruling that allows for business process patents. These constitute a huge number of the pending patents I bet - and have been the basis for most so-called software patents.

2. Specify that neither business methods nor software can be patented.

3. Invalidate ALL standing patents that were issued under the previous rule.

That simple.

Re:Software patents

[by Anonymous Coward]

I'd imagine most are actually business method patents. Software patents are stupid. Business method patents are even stupider. (Yes, there is some overlap, like software-implemented business method patents)

Re:Software patents

[ls -la (937805)]

No, patents have their place or the founding fathers would have forbade them altogether. The current problems stem largely from
(1) business method patents
(2) software patents
(3) genome patents
(4) the patenting process (including the difficulty and cost of overturning a patent, compared to getting an obvious patent through)
(5) patent trolls abusing (4).

Patents on physical inventions which are clearly new, innovative, and unique are fine.

Re:Software patents

[pembo13 (770295)]

Why do people speak as if the founding fathers were infallible? Not that I don't agree with your points. I would add medical patents to that list though, at least for non Viagra type medicines.

Re:

[innerweb (721995)]

Not infallible, just very very wise compared to modern people. They had plenty of issues, not the least of which were moral perspectives that have been proven dead wrong (slavery for instance). However, even though they had their faults, they seem to have had a much better grasp on what makes as good government than those who are in/modifying our government(s) today. They understood basic things like the passive tyranny of religion and state being mixed together. They understood the evils of having too

Re:Software patents

[CastrTroy (595695)]

Just patents made sense 200 years ago, doesn't mean they made sense today. For instance, having a patent valid for 17 years used to make sense because it would probably take that long before your invention had adequate market penetration. However products don't even last 17 years anymore before the company, or inventor moves on to something else. 17 years seems like a really long time in our fast paced society. The world wide web [wikipedia.org] wasn't even something most of the public knew about 17 years ago.

Re:

[gnuman99 (746007)]

Imagine a world where a patent troll patented HTTP or SMTP 15 years ago.

Yes, no Internet.

Maybe the patent system is broken if it does the opposite of what it suppose to do then?

Re:

[torkus (1133985)]

Actually the founding fathers knew they WERE fallible and WOULD make mistakes. That's why the attempted to build a 'self-healing' system. Unfortunately they were far more wise and had much more integrity than most people and especially corporation do today.

But hey, all they had to lose was their lives, families, and infant country. We've got $trillions wrapped up in this stuff today!

Re:Software patents

[fastest fascist (1086001)]

What is WITH this founding fathers cult? Can't you trust logic and argumentation, must you invoke a bunch of ancients as some kind of semi-divine authority to back your opinions up?

Re:

[steelfood (895457)]

Because as much as it sucks, including an appeal to authority is the only way most people will listen to an argument, logical or otherwise. Remember that 98% of this world appeals to some form of authority when looking for guidance or otherwise.

Essentially, fitting the founding fathers' idea of America gets equated with legitimacy in the minds of most Americans. The part that makes it acceptable is that arguments that fall on the side of the founding fathers usually aren't wrong; it's just so much easier an

Re:Software patents

[hardburn (141468)]

In principle, not evil at all. The idea is that the government will grant you a limited time monopoly on your invention, provided you document everything so that once your time is up, anyone can create and improve on the idea. This is in contrast to trade secrets, where you get to keep your invention for as long as you can keep it a secret.

(As a side note, the NSA has cheated this system, where some of their algorithms are currently a trade secret, and will suddenly become a patent if they're ever revealed).

The system today has severe implementation flaws, but the idea behind it is brilliant.

Re:There is no real issue. Problem solved.

[Overzeetop (214511)]

Bzzzt...thank you for playing. at $400 each, it doesn't even cover the cost of the prime examiner. A post above gave about 8 hrs of allowed time for a patent examination (10 per bi-week). Even if they all went to fresh-hires (i.e. inexperienced) at $63k/yr+ 10k bonus, with a typical "efficient" overhead and G&A of 80%, and accounting for sick, vacation, and holiday leave (264 hrs/yr to start), I get a net cost of $541 per patent. And that ignores training, startup, any other incentives, higher cost of experienced examiners, re-examination, etc.

Even with all the cash they have, they can't hire enough to get them back to even.

Re:

[Idiomatick (976696)]

They'd each need to do around a patent per day to get this finished with your numbers. But they are obviously flawed you forgot that you are dealing with a government program.

Patent requests: 400$ * 1,300,000 = +345,000,000$
RIAA support: 55,000,000
Physical infrastructure costs: 1 * 10,000,000 = -50,000,000
Execs: 2,000,000 * 7 = -35,000,000
Coordinators: 500,000 * 10 = -15,000,000
People who we don't know what they do (Management??): 80,000 * 300 = -24,000,000
Political bargaining: -35,000,000
M

Re:

[ls -la (937805)]

To me, maths dictates that if the patent office hired 12,000 examiners this year and did so for the next 5 years, the problem described would begin to decrease at year 3 and disappear at year 5. That would be an achievement by US standards. Why don't these officials just do this grade 9 math?

That would cost them more money. People are still going to apply for patents no matter how long it takes, so they really aren't losing much (financially) by keeping the backlog, whereas it would cost them ~600-700k/yr in salary to do what you suggest to kill the backlog.

Re:

[mdfst13 (664665)]

That would require the patent office to be able to find 12,000 people in the US each year that want to be patent examiners.

I still think that the solution is to eliminate the idea of patents being "granted" after a review by the patent office. Instead they should change to a system where patent applications are recorded and only reviewed when someone tries to enforce the patent. At that time, the defendant can be responsible for doing the prior art search and the patent office only needs to be responsible

Re:God I want this Problem

[clickety6 (141178)]

If they start outsourcing to india, this kind of surplus could generate more revenue than oil.

If they start outsourcing to India, we'd probably see a lot of new inventions coming out of India ;-)

Re:

[Shados (741919)]

It must be great to live in such a simple world :)

Really though. If you have a bunch of people who discover a process in a lab to, let say, make a type of semi-conductor that would allow processors to be 10 times faster...well, you may not exactly have the facilities to make CPUs... so the only people allowed to research them (and make money from it) is Intel, AMD and co?

Man you must love your large corporations!

Re:How many Patent examiners does it take to screw

[ProfBooty (172603)]

i think you have examiners confused with the court system.

The examining corp, nor the office itself ever allowed software patents, rather it was a court decision

see State Street Bank & Trust Company v. Signature Financial Group, Inc., 149 F.3d 1368 (Fed. Cir. 1998), 47 USPQ2d 1596