Christopher Herot's Weblog

Supreme Court Overturns Obviousness Rule

Last week the US Supreme Court rendered a decision in a patent case that is just starting to get the attention it deserves.  In a unanimous opinion authored by Justice Kennedy, KSR International Co. v. Teleflex Inc. et. al, the court reversed a decision by the appeals court that had clearly, if narrowly defined the bar for obviousness in a patent.

Section 103 of the Patent Act of 1952 states a patent may not be obtained

if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.

KSR v. Teleflex addressed the issue of when the combination of two prior inventions would be obvious.  The patent at issue concerned the addition of an electronic sensor to an adjustable cars gas pedal.  There had been previous patents for adding electronic sensors to gas pedals, and previous patents for making gas pedals adjustable (to allow for varying heights of drivers),  but Teleflex held a patent for adding an electronic sensor to an adjustable gas pedal.  While this may seem obvious in hindsight, the Appeals Court held that to be considered obvious at the time the patent was granted there would have to be some teaching, suggestion, or motivation (TSM) in the prior art, the problems nature, or in the knowledge of a person having ordinary skill in the art.  While the TSM had been applied by the District Court, the appeals court held that the application was not strict enough in that the District Court did not make "findings as to the specific understanding or principle within the knowledge of a skilled artisan that would have motivated one with no knowledge of [the] invention" to combine the prior art in the way the patent described.  In overturning that opinion, the Supreme Court recognized that it is common engineering practice to try existing things in new combinations and therefor held that "The combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results."

The Supreme Court cited some examples of this doctrine:

In United States v. Adams the court considered the obviousness of a battery that contained water instead of acid and used electrodes of magnesium and cuprous chloride instead of zinc and silver chloride.  While ordinarily substituting one material for another would have been considered obvious, in this case the prior art warned of the risks involved in the new combination.  The fact that it worked was unexpected and thus held not to be obvious.

In Anderson's-Black Rock, Inc. v. Pavement Salvage Co. the Court invalidated a patent involving the addition of a radiant-heat burner to a paving machine, ruling that while the combination was useful it did nothing unexpected.

In Sakraida v. AG Pro, Inc. the Court ruled that when a patent "arranges old elements with each performing the same function it had been known to perform" and yields no more than one would expect, the combination is obvious.

The Supreme Court did go on to say that the analysis of obviousness can be complex and that a court that performs it should made its reasoning explicit, but that it "need not seek out precise teachings directed the the specific subject matter of the challenged claim, for a court can take account of the inferences and creative steps that a person of ordinary skill in the art would employ."  The opnion adds that "A person of ordinary skill is also a person of ordinary creativity, not an automaton."

It remains to be seen how this will effect software patents, although some major cases such as Verizon v. Vonage will provide an indication.  In general, if this should make it harder to get the kind of patents where the application takes a verb from the dictionary and appends "on the Internet," which would be a good thing.

Verizon vs. Vonage - Claim Construction

Thanks to Brough Turner, I have posted the claim construction for the recent patent case between Verizon and Vonage.  As I learned from my experience as an expert witness, claim construction is the painstaking process in which individual words of the claims of the patent are examined in the context of the specification of the patent in order to assign meanings to them, for example in this case deciding if "name" refers to the called party or to the called party's device.  Claim construction is a matter of law and is therefor decided by the a, whose decisions are rendered in the above-referenced document.  All the jury gets to decide is whether the claims (as so understood) are infringed by the defendant.  As Fred Goldstein has pointed out, the jury was not asked to decide if these patents are valid.  Good thing for Verizon, since the claims at issue are for looking up a name and returning an address, something that's been going on in the Internet at least since Paul Mockapetris invented of DNS in 1983.  I expect this will be in the courts for some time to come.

Verizon Services Corp. et al v. Vonage Holdings Corp. et al

Many people have written about the recently decided patent infringement case (1:06-cv-00682-CMH-BRP) Verizon brought against Vonage in the Federal District Court for the Eastern District of Virginia
but no one has posted any details about which patents were infringed.  So here goes...

The jury found that Vonage had infringed on the following Verizon patents.

6,104,711 - claim 20

15. A method comprising:

receiving a name translation request at a server coupled to a public packet data network;

executing a conditional analysis in response to the name translation request;

if the conditional analysis produces a first result, translating a name included in the name translation request into a first destination address;

if the conditional analysis produces a second result, translating the name included in the name translation request into a second destination address; and

transmitting a response message containing the first or the second destination address to a calling device for use in establishing communication at least partially through the public packet data network.

20. A method as in claim 15; wherein:

the first and second destination address includes a numeric Internet Protocol address; and

the second destination address further includes information relating to call routing via a public switched telephone network.

6,282,574 - claim 27

26. A method comprising:

receiving a name translation request at a server coupled to a public packet data network;

translating a name included in the request into a destination telephone number associated with a name included in the request; and

transmitting a reply containing both the destination telephone number and a packet data network address of a telephone gateway coupled between the public packet data network and a telephone network through the public packet data network to a calling device.

27. A method as in claim 26, wherein the address is an Internet Protocol address.

6,359,880 - claims 1, 6, 7, and 8

1. A method comprising: registering a wireless telephone terminal in a localized wireless gateway system;

transmitting registration data identifying the gateway system from the localized  wireless gateway system to a home location register database through a public packet  data communication network;

receiving a request from a calling computer coupled to the public packet data communication network for a call to the wireless telephone terminal;

in response to the request, accessing the home location register database and obtaining a packet data address for the localized wireless gateway system;

using the address to set up a voice communication through the public packet data communication network and the localized wireless gateway system between the calling computer and the wireless telephone terminal.

6. A method as in claim 1, wherein the public packet data communication network is a packet switched network.

7. A method as in claim 6, wherein the packet switched network comprises a system of interlinked data networks using TCP/IP protocol.

8. A method as in claim 7, wherein the system of interlinked data networks comprises  the Internet.

Clint Ricker has a good analysis on his blog.

Patenting a Law of Nature?

The US Supreme Court heard oral arguments yesterday in the case of Laboratory Corporation of America Holdings v. Metabolite Laboratories Inc., et al.  This case has been described one of whether basic truths of nature can be patented, most colorfully by Michael Crichton in a New York Times Op-Ed piece last Sunday.  As is usually true of high profile legal stories, there are more subtle aspects that have not appeared in the popular press.

The story begins with three medical school professors, Sally Stabler and Robert Allen of the University of Colorado and John Lindenbaum of Columbia University, who discovered that high levels of the amino acid homocysteine are indicative of deficiencies in cobalamin (vitamin B12) and folate (folic acid), both of which are essential.  As is often the case with scientific discoveries, the established authorities were initial skeptical but years later came to accept their work as the standard.  Meanwhile, the professors developed a clinical test for measuring serum homocysteine which doctors could use to test for B12 deficiency, received U.S. Patent No. 4,940,658.  The patent was assigned to Competitive Technologies, Inc. (CTI) who in turn graanted a license to the newly formed Metabolite Laboratories, Inc. which in turn provided a license and know-how to Laboratory Corporation of America (LabCorp) who performed tests for doctors.   This arrangement went on for six years until Abbott Laboratories came up with their own test for homocysteine, whereupon LabCorp stopped paying royalties to Metabolite and started using an automated test provided by Abbott.  As it turns out, while the bulk of '658 patent covers the somewhat complex process of doing the test, one claim covers the relationship between homocysteine and cobalamin or folate:

Claim 13: A method for detecting a deficiency of cobalamin or folate in warm-blooded animals comprising the steps of:
assaying a body fluid for an elevated level of total homocysteine; and
correlating an elevated level of total homocysteine in said body fluid with a deficiency of cobalamin or folate.

Metabolite has used this claim to argue that LabCorp still owes it royalties no matter which test it uses.  In fact, in Tuesday's arguments, Metabolite's attorney conceded, under questioning by Justice John Paul Stevens, that a doctor could infringe the patent by ordering a homocysteine test with the intent of determining B vitamin levels and then using the test results to infer whether there was indeed a deficiency.

In attempting to invalidate the patent, LabCorp has argues that this claim would "allow an effective monopoly over a scientific principle" and thus violates the Courts longstanding rule on patenting laws of nature.

Of course it gets even more complicated, because Metabolite argues that LabCorp didn't raise this issue earlier in the appeals process.  Nonetheless, the question the Supreme Court has agreed to consider is:

Whether a method patent setting forth an indefinite, undescribed, and non-enabling step directing a party simply to “correlate[e]” test results can validly claim a monopoly over a basic scientific relationship used in medical treatment such that any doctor necessarily infringes the patent merely by thinking about the relationship after looking at a test result.

Invalidating the '658 patent could, in turn invalidate a lot of other biotech patents, including those based on discoveries of the human genome.  Whether you find that scary depends on whether you own those patents.  Ultimately I think one needs to return to the original intent of the patent system as set forth in the U.S. Constitution [Article I, Section 8, clause 8]: "To promote the progress of Science and useful arts."  So, the question is whether discoveries such as those of Stabler, Allen, and Lindenbaum would have been made and, more importantly exploited, without the promise of a monopoly on any such exploitation.  Also, would subsequent improvements be more or less likely if they had to negotiate around such monopolies.  And on a more abstract level, would scientific progress be furthered if laws of nature could be patented?

EOLAS and the Update to Internet Explorer

Microsoft recently released update 912945 to Internet Explorer.  Curiously it is offered as a ‘non-security update’ in the “Optional software updates” category on Windows Update.  Although there is no mention of this in any Microsoft documents that I could find, this update is in response to Microsoft's setbacks in the long-running EOLAS patent case.   EOLAS, a spin-off of UC Berkeley, was issued US Patent 5,838,906 for DISTRIBUTED HYPERMEDIA METHOD FOR AUTOMATICALLY INVOKING EXTERNAL  APPLICATION PROVIDING INTERACTION AND DISPLAY OF EMBEDDED OBJECTS WITHIN A  HYPERMEDIA DOCUMENT, which they claim is infringed by ActiveX.  After years of litigation, and an impassioned letter from Tim Berners-Lee, the US Patent and Trademark Office agreed.  (See the docket report, especially the Notice of Intent to Issue a Reexam Certificate.)  The deciding issue was that although there was plenty of prior art that embedded graphics in a document, none of the materials submitted in the case showed dynamic, interactive embedded content, at least not without requiring the user to click first.

One interesting sidelight is that the PTO rejected the argument that Viola was prior art on the grounds that it's embedded applications were implemented in an interpreted language, a distinction that may have lost its importance in recent years.

Rather than pay $521 million to EOLAS, Microsoft issued an update which engineers around the '906 patent.  There appear to be two changes that are provided:

1. When a user encounters an ActiveX object, such as a Flash movie, he/she first sees an annoying text box admonishing the user to click first.
2. To avoid inflicting this text box on their users, developers need to change the way they embed the ActiveX control.  Instead of doing it inline with an EMBED tag, they need to use JScript to invoke a separate file the uses the DOM to embed the object,

The net result is that developers have to go to a lot of work, users will encounter a lot of annoyance, lawyers collected lots of fees, and otherwise nothing has changed.  You would think EOLAS and Microsoft would have settled for a much smaller number and walked away happy, but we will never know.