ms3

[jharvey]

I see another frivolous patent attempt from MS land. http://www.google.com/patents/about?id= ... =BG+Soflex

I also see the patent reform act of 2009 allows anyone to object about a patent application, or at least if I believe what I read on wikipedia that is.

From http://en.wikipedia.org/wiki/Patent_Reform_Act_of_2009 "Procedures to allow third parties to submit prior art while applications are pending."

To me it sounds like they again applied for something that all major MFG's have been using for 30 years. If I knew where to submit a prior art note, I'd be sure to fill out some paper work.

If Bruce didn't threaten me with patent infringement a while back, I wouldn't care, but he spited me. Anyone know where to file some papers?

[Fred]

You should file some papers for that older patent, but this one is fairly unique AFAIK. They talked about it a while back and want to take what amounts to a sound recording of the waveform of the sensor and then filter it digitally and then use the data to determine positions. This is quite different to current MS and other edge sensing and timing calculations that are done commonly.

[jharvey]

Sounds like Autocorrelation to me http://en.wikipedia.org/wiki/Autocorrelation

I took a DSP class on that back in 01. This newer application doesn't sound unique to me. If I paint my muffler black, can I really expect to get a patent on the muffler?

[MotoFab]

I just read the patent claims, and to me it seems that if there is anything unique, it is the combination of several common methodologies into one 'component'. As fewer truly novel things are discovered in such a well covered field as controlling automobile engines, combination patents have become more common. Along those lines, greater processing power is also helping.

They are certainly doing some high-grade processing, and the combination of things may be unique, but taken individually I do not think their claims are unique. Derivative calculations to determine zero crossing for an AC signal, and mean square error correction for velocity deviation, don't seem to qualify as unique.

Acourse, unique isn't the measure of a patent's validity is it. To be granted, the claim must be unobvious. And on that point I'm not sure it is unobvious to do more things with more processing power.

Maybe if they found out that velocity errors could be predicted using never-before-used metrics, that'd be unobvious.

Here is something that might be unique...
The crank velocity isn't constant over the period of one revolution. A graph of the integral of the velocity deviation could be constructed and a 4D map made of rpm, throttle angle, velocity deviation, and crank angle. You wouldn't need to map velocity deviation at many points around the circle, maybe just the crank angle at the point of max and min deviation during each cylinder's power stroke. And interpolate from there.

The above might be unobvious, but I don't think there's much to be gained in using the above over applying simple mean square error correction to the periods between teeth. An algorithm like the above might have some useful application on a very low rpm single cylinder motor, with only 4 teeth on the crank wheel. But on a motor turning over 2,000 rpm, and a 30 tooth crank wheel, I can't see any practical benefit, unobvious though it may be.

- Jim

[Fred]

Good post, Jim :-)

[MotoFab]

Thanks Fred. I hope the MS team get their patent. My thoughts on what constitutes unobvious have to do with the US Patent Office, not the MS guys.

Patents don't affect the personal use, and sometimes the not-for-profit use, of the claims. In that regard patents are an excellent open source disclosure mechanism.


[hijack]
The USPTO has so far cranked out 7,650,647 patents, 3200 per week average. 167,350 utility patents in calendar year 2009. C'mon, seriously? That many things are unobvious every week?

The fees and maintenance for the simplest utility patent claim will net the patent office $10,170 over the life of the patent. The 167,350 patents issued last year is a potential $1,701,949,500 in revenue. Maybe that's one of the reasons for lowering the bar for the patent claims test.

A graph of the last 30 years of the patents-issued-per-year is starting to look like a log curve. That graph would parallel the total number of people in the patent/legal departments of large companies.

IBM is getting close to 5000 per year.
Check out the list on this press release.
http://www.ificlaims.com/IFI%202009%20p ... 0final.htm

More and more the Patent Office is issuing patents based primarily on the claims meeting the slight test of never-been-done, than anything truly novel.

The more and more that patents become a tool to extort money and leverage small business out-of-business, the more I cozy up to the notion of full disclosure of source materials, drawings, formula, etc. But widespread open source would force actual competition, and it appears that's the last thing anyone wants.
[/hijack]

[Fred]

Good post, Jim :-)

Keep 'em coming :-)

[MotoFab]

I also see the patent reform act of 2009 allows anyone to object about a patent application, or at least if I believe what I read on wikipedia that is.

Many folks believe that Congress has changed the US Patent Law recently. You are not alone in thinking that JH.

So far, with one exception, these bills have died in committee in the chamber that introduced them. For those not familiar with US lawmaking, at that stage of a bill becoming law, the child is still in nappies.

Links to the proposed legislation from the three Congresses:

Patent Reform Act of 2005 in the 109th Congress
http://www.govtrack.us/congress/bill.xpd?bill=h109-2795

Patent Reform Act of 2007 in the 110th Congress
http://www.govtrack.us/congress/bill.xpd?bill=s110-1145
http://www.govtrack.us/congress/bill.xpd?bill=h110-1908

Patent Reform Act of 2009 in the 111th Congress
http://www.govtrack.us/congress/bill.xpd?bill=s111-515
http://www.govtrack.us/congress/bill.xpd?bill=s111-610
http://www.govtrack.us/congress/bill.xpd?bill=h111-1260

[jharvey]

That's an interesting page. I also see this one passed.

http://www.govtrack.us/congress/bill.xpd?bill=hr110-636

It appears to be a consolidation of prior bills.

The ones that were worked on in 2009, haven't had time to clear the system yet, so we don't know if they will pass or not. Also just because it doesn't pass congress, doesn't mean that our legal system or PTO can't implement them in part or in full. Perhaps more info could be obtained from the below link under the courts section.

http://bulk.resource.org/

[MotoFab]

That's an interesting page. I also see this one passed.
http://www.govtrack.us/congress/bill.xpd?bill=hr110-636

Perhaps more info could be obtained from the below link
http://bulk.resource.org/

Perhaps more info could be obtained from the below link
US General Printing Office: Constitution of the United States: Main Page
http://www.gpoaccess.gov/constitution/index.html


Regarding the other link: H.Res. 636 110th is a House resolution. Maybe you're not familiar with what a congressional resolution is.

As to your last point, only Congress can modify US Code. Agencies do not enact laws and cannot modify codified law. The People cannot modify US Code either. The People can modify State Code by referendum vote, but not US Code. There is no provision in the Constitution for a federal electorate. I just realized, maybe you're not a US citizen.