Homework 6: Patentable Subject Matter Paper

From Bill Goodwine's Wiki
Revision as of 09:44, 21 March 2011 by Kyle Tennant (talk | contribs) (Created page with "In order to determine the patentability of a process, let’s first turn to the formal definition of a process offered in Section 100(b) of the U.S. Code under Title 35: “a ‘...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

In order to determine the patentability of a process, let’s first turn to the formal definition of a process offered in Section 100(b) of the U.S. Code under Title 35: “a ‘process’ means process, art or method, and includes a new use of a known process, machine, manufacture, composition of matter, or material.” In Cochrane, as cited by Diehr, a process was further described as “an act, or a series of acts, performed upon the subject-matter to be transformed and reduced to a different state or thing.” Note that these definitions are not descriptive of any process, but only those which are patentable.

Now, how does one determine whether a process is indeed a “process” under patent law? The first step is to consider what the claims of the patent application are. In several cases, the applicants attempted to claim inventions which were by definition unpatentable. The court system has understood the U.S. Code to exclude three categories from being patented: laws of nature, physical phenomena, and abstract ideas. These exceptions, though not included in the written text, are in keeping with the spirit of novelty and utility found in the code. Thus, if an application amounts to claiming one of these principles by itself, then the claim is unpatentable.

This is not to say that any patent containing a mathematical formula or other fundamental principle is unpatentable. If the mathematical formula is part of a larger process and the applicant(s) seek(s) the entire process as an invention, then the process may indeed be patentable provided it complies with the other standards of patentability - novelty, nonobviousness, etc. This is the second step: if a mathematical formula or other law of nature is present, are there other substantial steps claimed in conjunction with it? One must be careful here, however, because if the application claims any use of the formula for any purpose, or if the other steps in the process are trivial and meant solely to frame the formula, then the invention is not patentable.

Special consideration must also be given to business practices and computer programs. While neither article is expressly forbidden by legislature (indeed, business methods are specifically mentioned in Section 273), it is much harder to prove patentability for these two types of work. Business methods generally do not produce a transformation of material or a function as in a machine, and computer programs often are simply mathematical algorithms, which amount to fundamental laws of nature. Patentability for computer programs is shown in the manipulation of data to produce some new result. In general, as stated in State Street, business methods should not be approved or barred for being “business methods,” but rather should be judged like any other process. In general, it is not advisable to categorize an invention as one of a type of works, be it process, machine, or otherwise. Instead, care should be taken to ascertain its practical utility; this utility is proved in many cases by the production of a “useful, concrete, and tangible result.”

In summary, a process is patentable if (1) it claims a series of actions which produce a real, tangible, and useful product, action, or piece of information; (2) it is tied to a particular machine; (3) it transforms something into something else; and/or (4) if claiming a formula or computer program, it does so in conjunction with other steps that together meet the criteria for a process. Note that a patentable process does not have to satisfy all 4 criteria; part of the difficulty the courts have had in this area stems from their rigid misapplication of one standard which narrowly defines a “process.” The verbiage of the code is quite broad so as to encourage a wide range of patentable subject matter.

The policy dictating patentable subject matter has been shaped by a number of factors. First, consider the exclusion of laws of nature, physical phenomena, and abstract ideas. These principles amount to general knowledge that better serve as such than as the exclusive right of one person; they are fundamental “first things” upon which discovery and innovation are based. The point of patents is to promote science and the useful arts; if mathematical formulae, basic laws of nature, natural scientific discoveries, or abstract ideas (like, for instance, hedging risk) were patentable, then such patents would be diametrically opposed to this goal. The progress of science would be retarded by the fact that one person had the rights to something so fundamental. In Gottschalk, the inventors sought to patent a method for converting BCD numbers to pure binary; this amounted to attempting to patent a mathematical formula which should by nature belong to the public. This type of knowledge by necessity must belong to no one so that it can belong to everyone.

Secondly, it is inappropriate to consider aspects of a patent that utilizes these principles apart from the rest of the claims. In Diehr, the function of the computer was separated from the traditional methods of creating molded rubber pieces. This ruling was confirmed in Arrythmia Research, in which a process for detecting v-tach in heart attack victims was patented. In both cases, the use of a computer program was claimed only in conjunction with the rest of the steps. One cannot separate a claim into old and new parts and ignore the old stuff when trying to determine patentability. Indeed, this method would preclude large numbers of patentable processes where the individual parts are well known and not new. However, questions of novelty are a separate matter from questions of patentable subject matter.

Thirdly, policy has been shaped by changes both in the realm of scientific progress and in the speed with which it occurs. The 1952 Patent Act did not foresee the rise of computers, nor the rapid pace with which they are improved today. Other fields like business have also seen a paradigm shift: many businesses today, such as Accenture and Price Waterhouse Cooper, almost exclusively market business processes or financial methods instead of a tangible product like a plow or coated glass. As innovation drives technology and business forward, it becomes necessary to redefine patentability and processes to reflect modern interpretations of such terms. Outdated language could well have precluded all computer programs from being patentable; however, updates necessarily move the law to keep up with the state of the art. This is why the court system has stressed that the words in the law be interpreted in their “ordinary, contemporary, common meaning.” [cf. Diehr, Bilski] The perennial challenge of patent law becomes more difficult: the Information Age has enabled more and more individuals to innovate and seek compensation for their efforts. Thus, the problem arises once again: where should the line be drawn between encouraging innovation and discouraging monopolies?

This line has been sketched by a number of cases since the 1970s. Benson et al. filed for a patent for converting binary coded decimal numbers into pure binary numbers in a method not specific to a certain application or apparatus. They attempted to claim any use of their system in a general-purpose digital computer. The Supreme Courted granted certiorari to determine whether or not the claim constituted a "process" under U.S. patent law. The claims in question are algorithms, or procedures by which a computer may solve a mathematical problem. In this case, the algorithm converts numbers from one form to the other.

The Court did not say that a process is never patentable if it either does not transform an article or is not tied to a particular machine. Nor did it say that no process can be patented for a computer. Nor was the Court meaning to freeze patents and ignore the advent of new electronic technology. Rather, the Court ruled that the formula used in this patent has no application outside of its connection to a computer. It is not a transformation of material, nor is the algorithm a new or useful application, nor is the method separable from its mathematical formula. Thus, if the Court granted this patent, it would for all intents and purposes be patenting a scientific law.

Ten years later, the Court heard a case in which Diamond and his crew filed a petition for a patent that would mold raw rubber into precise shapes. This process had been difficult in the past because of the uncertainty surrounding the proper curing time. This uncertainty stemmed from the fact that determining the temperature inside the molding press was problematic. The cure time is related to the temperature by a known mathematical formula; the respondents developed a system by which the temperature inside the press was continuously fed into a computer program that would automatically open the press at the right time. They were at first denied, but the court later held that they were in fact entitled to a patent.

The examiner initially denied the patent because he believed the use of a computer negated patentability. However, just because a computer is involved does not mean the subject isn't patentable. Furthermore, the respondents were not looking to patent a mathematical formula; rather, they sought to use a mathematical formula in conjunction with other steps to solve a problem which had existed in the process of making rubber; they were using the formula as a tool to improve an existing method. In Benson, the mathematical formula was solely for use in a digital computer, and the Court treated it as a fundamental law of nature. Likewise in Flook, the claim was not for a greater process of determining values or gathering data, but solely for a mathematical formula for an "alarm limit." Thus, that claim was not patentable. In the current case, the patent is not only for the mathematical formula, but for all the other steps taken in the process. Patentability, at least for computer programs, lies in the contribution of the program to a larger process.

A similar conclusion was reached in 1992. Arrhythmia Research filed a plea to invalidate a patent belonging to Michael Simson. The case was heard in the CAFC. Simson's invention charted the electrical signals of the human heart to certain functions by use of electrocardiographs. Heart attack victims are vulnerable to an arrhythmia know as ventricular tachycardia, which may be detected by the use of an EKG. Simson, a cardiologist, patented a way by which a person's vulnerability to such an arrhythmia could be determined. The issue of patentability arose from the fact that a computer model calculated several of the values used in the patent, utilizing a number of natural laws and algorithms. The District Court held that this patent was invalid due to a lack of statutory subject matter. The Appeals Court found that the Simson material was analogous to the Diehr material in that the Simson patent did not seek the exclusive rights to the formula itself, but rather the formula in conjunction with all the other steps which helped identify high-risk patients. The Appellate Court sent the case back to the District Court so that matter of infringement could be determined.

In 1998, the CAFC decided a case in which Signature Financial Group appealed the decision that their patent for a financial services program was invalid due to non-statutory subject matter. The patent was for a data processing system which provided mutual fund account support for Signature's clients. The system, termed Hub and Spoke, pooled mutual funds, or spokes, together into a single portfolio, or hub, in a partnership. State Street and Signature had been negotiating terms for the use of Hub and Spoke; State Street, upon the dissolution of negotiations, filed a declaratory judgment action stating that the patent was invalid. The Appellate Court had to determine whether the District Court was correct in issuing a summary judgment on the case; that is, whether there were any questions of the facts which would require a trial to determine the material fact of the case. They found that State Street was not entitled to summary judgment because the patent was, in fact, valid and directed to statutory subject matter.

As before, the patent was filed not for the mathematical formulae, but for the entire system which utilized the formulae, specifically the computer which was designed to perform these functions. Since the patent was filed for the machine, it is therefore valid since it pertains to statutory subject matter. In this case, it is not the formula but the machine or application which is protected by the patent. In addition, the District Court cited the broadness of the patent as a means of rendering it unpatentable. However, this decision falls under the regulation of sections 102, 103, and 112, not 101. Broadness does not preclude patentability.

Finally, in 2010, the Supreme Court found in Bilski that hedging risk, a business practice, and a method for doing so in the energy market were unpatentable. They agreed with the CAFC, but disapproved of the method in which it determined this ruling. In determining whether a process is patentable, the "machine or transformation" test is not the only method. Doing so errs in two ways: it allows the court to put limits and conditions on the laws not actually disclosed in the legislation, and it does not interpret the meaning of the legislation in an ordinary, contemporary way. Saying that a process must be related to a machine or transformation of an article does not assign a normal, common meaning to the word "process." In addition, the use of the word "process" in 101 does not exclude all business methods from this protection. In fact, some of the verbiage implies that business methods may be patented: for instance, the permissible defense of prior use to argue against infringement implies that business methods may be patented. However, another section (273) explicitly states that business methods are included in the general definition of "method." Be that as it may, the claims at issue remained unpatentable. One claim is for hedging risk, and the other is for doing so in the energy market. According to historical precedent, neither claim is patentable. Some of the claims describe abstract ideas and the mathematical formulae that quantify them; the rest indicate meager attempts to add to the abstract idea by explaining hedging and how certain pertinent information may be determined. This required no new rulings or interpretations on the Court's part: the prior rulings and the wording of 101 serve to effectively preclude the claims in this case. Though the machine-or-transformation test has been shown not to be a definitive test, the Court does not take issue with the Federal Court attempting to define a test so long as it does not contradict the laws or prior rulings.

While the courts have struggled to define what is patentable and what is not, it is clear that the matter at hand is blurred by the evolution of technology, business practice, and invention itself. As we enter a modern, digital age, wording like “tangible results” takes on a different meaning. Rather than being a physical object we touch, processes often produce information, recommendations, or guidelines for actions, decisions, and so on. We are in the midst of an Information Age in which new products and inventions are increasingly found in a digital, mental, or otherwise intangible realm. In my view, the court system, while acting within the letter of the law, has been largely unfair to business methods and technological innovations. While I agree that some cases, such as Benson, were ruled correctly, I fear that the precedent which has been set is too narrow to properly accommodate, or indeed even acknowledge, the progress which has been made in these fields. Business itself is now a business, and computers are revolutionizing the way we interact, learn, play, and conduct our daily lives.

Perhaps, however, it is necessary for business methods and technological advances to be held to a higher standard. Apple’s iPad, for instance, was on market for 14 months before the release today of the iPad 2.0, and it was barely three years ago that the business world realized its excessive, irresponsible method of conducting finances was akin to building castles in the sky. If Apple had sole rights to MultiTouch technology, then tablets like Motorola’s Xoom (arguably the result of reverse engineering the iPad 1.0) wouldn’t push the art forward. Yet, if Apple was able to patent MultiTouch technology, it would be compelled to fully disclose its method, which would add to the source of knowledge available to other companies. Decisions, decisions.

My proposal is this. Processes are patentable as long as they provide a transformation of matter, are connected in a significant way to a machine, or develop new information which is not the sole output of a law of nature. Business methods should not be subject to alternative tests because they are business-related; this is actually the current standard found in Bilski. However, computer programs should be patentable not just in conjunction with a process (technological in Arrhythmia, manufacturing in Diehr, and business in State Street), but also if the program provides a new function for the device in which it operates (like MultiTouch) or offers a demonstrable advantage over previous computer methods of achieving the same result (e.g., updated operating systems). Many computer programs can’t and shouldn’t be oversimplified as merely digital expressions of an algorithm. They should be treated more like manufacturing processes: if it does the same thing better, more quickly, or more efficiently, then it’s patentable.