By Glenn Foster
Disclaimer: The following represents the work product and opinion of the author, except where stated. I am interested in receiving comments, questions, or other viewpoints at gbfoster@comcsat.net and I will attempt to respond. The reader is advised that this is not to be used to represent my legal opinion (in which I refuse any responsibility therefore). I am providing these concepts in an effort to allow patent practitioners and clients to know there are alternate techniques for protecting innovative software from using pure software patents, as an effort to improve the overall quality of computer software patents being filed within the USPTO. Such established alternate techniques are more aligned with the machine or transformation test of Bilski, as well as guidelines the USPTO appears to be following.
The State Street Bank 149 F.3d 1368 (Fed. Cir. 1998) and ATT cases heralded the beginning of computer software patents as well as business method patents. Up to that time, patentable arts were limited to the electrical, mechanical, and chemical areas. Bilski vs. Kappos, decided by the Supreme Court in 2010, marks a major milepost in the way computer software patents should be prosecuted, and will likely severely limit the number of “pure” software patents that will be granted by the USPTO. There are certain alternative software patents that will continue to be allowed at a relatively high rate of well over 50 percent, but many of these alternative software patents rely on “machine or transformation” of aspects other than pure software for patentability. There are certain patent practitioners and corporate patent portfolios or individual patent portfolios that are much more successful than others. Following Bilski, the USPTO appears to be establishing guidelines. Patent practitioners should always be attempting to increase our rates of allowance, to benefit our clients. It is easy to assume that if another company, such as IBM or ATT Bell Labs, is obtaining a very high allowance rate, that each company or individual can also obtain such allowance rates. While there is some truth to this, there are certain reasons why certain organizations are not performing nearly up to these expectations, and some are within the single digits for patent allowance rates. This paper first describes different types of software patents, and the effects of Bilski on different types of software patents. Then, the paper considers why certain organizations are effective while others are not.
As an underlying theme to this paper, patents should be a positive, and even fun area in which to practice. After all, the inventors are often very optimistic about the hopes and dreams of their innovations whether the invention is individual or assigned to an organization. It has been said that patent attorneys are the only attorneys our clients wish to see. I sometimes refer to patent law as the ideal cocktail party profession, because almost everybody has, or at least knows a relative who has, an innovation, be the person semiconductor developers, medical professionals, housewife, aircraft engineers, outdoors enthusiasts, or whatever. It is human to be innovative and ask, why don’t they do it this way or make something that solves a problem. Engineers, technicians, and programmers may be considered as problem solvers. Patent practitioners may be considered as communicators of how the inventors solve their problems. Typically, no one knows the problem to be solved better than the innovators. Quite often the innovators also know the state of the prior art better then almost anyone. Patents conjure up such innovators as Edison, Ford, Bell, Boeing, and more recently Bill Gates, Steve Jobs, and a number of major corporations and small innovators. From an engineering perspective, what could be more innovative than Thomas Edison wandering around Menlo Park with his numerous patent attorneys? Edison still has more issued patents than anyone. While it is relatively easy to understand mechanical and electrical innovations, software innovations are less visible and more abstract. Comparing Edison’s light bulb innovation to some software process going on within a black box or computer network, for example, it is easy to understand why patent examiners are often challenged to find innovation in pure software patents.
For the purpose of this article, software patents and patent applications are characterized as either pure software or multi-technology software. Pure software patent applications often occur within a computer, network, or black box and follow a flow chart describing software-only processes that occur within a computer and are not tied into a machine or transformation, and these pure software patent applications often fail to satisfy the machine or transformation test. Pure software patents face a challenge of being considered as number crunchers, overly abstract, or simply being within a black box only. Recently, the USPTO allowance rates for pure software patent applications, that are not tied to machine or transformation test, have remained constantly low. Bilski vs. Kappos likely represents the end of the age of pure software patents.
There are other types of software patents from pure software patents that remain viable and even strong. Multi-technology patent applications often satisfy the machine or transformation test, and will likely pass muster under 35 USC §101. Those applications that satisfy 35 USC §101 within the USPTO will be further examined. This analysis of Bilski v. Kappos emphasizes the "machine or transformation" test will almost certainly be used as a major determiner (though evidently not the only one) for computer software patent in satisfying 35 USC 101. As such, the machine or transformation test applies to examination within the USPTO as well as litigation within the Patent Courts.
Practically speaking, the multi-technology software patents likely satisfy the machine or transformation requirements, and do obtain considerably improved allowance rates at the USPTO. I first became involved in the multi-technology software patents as a result of my distinct technical degrees and professional experience in each of the electrical engineering, mechanical engineering, and computer software areas. Since electrical and mechanical patents have considerably higher rates of allowance than pure software patents, it made complete sense to continue practicing at least partially in the mechanical or electrical areas for the benefits of my clients. These distinct degrees and my subsequent patent experience in electrical, mechanical, and software areas allow me to target the patent application for a variety of particular technical and legal innovations. This diverse technical background also allows me to objectively consider how many of the USPTO examiners are responding to varied patent applications.
A number of attorneys including myself have found considerable success when filing and prosecuting multi-technology software patent applications that are spread between software and other technical areas such as illustrated in the attached writing samples. These multi-technology software applications include such types as microprocessor controlled, computer controlled, imagers, medical devices, transportation devices, etc. such as the writing samples appended to the end of this article. As a rule of thumb, the electrical and/or mechanical portion likely satisfies the machine or transformation test that may make the patent application more innovative, and thereby patentable.
Quite often, the claim language (as well as the flow charts) of the multi-tech software patents contain very mechanical, electrical, or other machine-like language. In such cases, the computer software may be viewed as supportive of an alternate or additional software embodiment. As such, the patentability focus for such multi-tech software patents includes the mechanical, electronic, electric, optical, or other non-software portion. Skilled patent practitioners can describe innovation within the mechanical, electrical, optical, medical, transportation, or other machine-like portions of the multi-tech software patents whose operation can be understood far easier than those same practitioners can describe innovation in data often performing some abstract and/ or proprietary process likely within a black box or computer network.
This machine or transformation emphasis is not an academic exercise, and appears to be consistent with how the USPTO has been operating for a considerable time. Numerous practitioners obtain very high rates of allowance that integrate software into patents to satisfy the machine or transformation test. From a perusal of the USPTO Search Web Pages (considering both the Issued Patents and the Patent Applications Published), a large percentage of pure software patent applications remain pending and face protracted prosecution while the USPTO has issued a large percentage of other types of patents such as mechanical, electrical, and multi-technology software patent applications that consistently satisfies the machine or transformation test.
There is reason to believe that the holding of Bilski will remain in effect through the Patent Courts and the USPTO for an extended duration. Though at the end of the Bush Administration (corresponding to Jon Dudas as Director of the USPTO), very few patents were granted while the USPTO under David Kappos is now granting many more patents and hiring a large number of patent examiners.
For certain clients, the multi-tech software patent applications have an allowance rate similar to the electrical and mechanical applications at well over 50 percent, and for certain clients approximately 90%. With the USPTO following the machine or transformation test of Bilski, it is likely the allowance rate of pure software patent applications will drop from 10 percent even further. Bilski contains no language which would tend to force USPTO patent examiners to grant more pure software patent applications that they do not believe satisfy 35 USC §101. Practically, it will be difficult for patent practitioners to justify further filing of pure software patent applications to our clients. By comparison, multi-technology software patents will be much easier to justify.
A lot can be gleaned by considering how well each company or firm is doing in patent prosecution, based on the USPTO databases (most notably patent applications filed as compared to patents granted).
Through my career, I have viewed the case law as providing clues to how to proceed with software as well as other technologies in patent law. The CAFC's first Chief Judge, the late Howard Markey, spoke at the USPTO in the late 1980s to patent examiners. He emphasized that examiners almost always got it right; and elaborated that some high percentage of patent examiners (e.g., 95%) agreed on patentability for some high percentage of patent applications (e.g., 95%). The only technologies that were patent eligible at that time were electrical, mechanical, and chemical, and with each of these technologies, patentability considerations were less confusing and abstract than for pure software patent applications. As such, he was commenting how the USPTO’s patent examination process was largely effective and efficient. He noted that while the USPTO provided little time for patent examination, it is not clear the overall examination process would improve significantly if examiners were provided more time.
Chief Judge Markey then commented that the CAFC should make patentability determinations on difficult or uncertain technologies. In this brief presentation, I am convinced the Late Chief Judge Howard Markey effectively summarized the structure and operation of modern patent law from the viewpoint of the CAFC. Since that time, the patent world has expanded to include computer software and business methods as patent eligible subject matter as per the State Street Bank and ATT cases. Time has shown that patent examiners have more difficulty finding patentability in most software patent applications than traditional mechanical or electrical patent applications.
To illustrate practical effects of such low allowance rates, a 10 percent allowance rate means, in effect, that if a client is paying $5K per application, then for each patent granted they are paying $50K and getting a patent the attorney spent only $5K of time on. A $5k patent is not likely a high quality patent! I have been advising my clients to steer away from these pure software patents, and focus on higher yielding quality patents that more clearly define innovative concepts and have higher allowance rates to provide the potential to establish a structured patent portfolio. The USPTO appears to be complaining about the lack of quality and little innovation of a large number of their pending applications, and many of the pending applications appear to be of the pure software variety.
To paraphrase President Clinton for patent law, “it’s the allowance rate, stupid”. For instance, it is impossible to generate licensing revenues or limit others from making, using, or selling the patented item without an issued patent. IBM, for example, has long been recognized as having the most patent granted by the USPTO, and it is no coincidence they also continually yield among the highest royalty rates. There is considerable evidence to suggest that those attorneys, and those clients, who understand the patent system best and how to interact within the USPTO unquestionably get quantifiably better results and higher allowance rates. There are a number of patent practitioners (myself included) that draft multi-technology software patents as illustrated in the writing samples. It is therefore not an issue that no one is getting any software patents issued now, this is simply not true. Numerous patent savvy prosecutors have effectively prosecuted, and will continue to prosecute, multi-technology software patents that are more likely to pass the machine or transformation test.
Considering how multi-technology software as compared to pure software that a particular clients’ patent portfolio may be may be quite easy. Look at the granted patents as compared to the pending patent applications. I have searched the USPTO database for a number of clients. For a number of software clients, the more multi-technology software patents are issuing at a far higher rate than the pure software patents. Patents that don’t issue can’t be licensed and/ or enforced, so quality and innovation in patents affect the patents even after issuance. It could be argued that the only post Bilski software patents that the USPTO will grant (and the patent courts will enforce) will be multi-technology software patent applications. Consider many of the innovative technologies that required multi-technology to describe innovations best for their respective period.
The option for those attorneys and their clients who disagree with the patent examiners’ rejections under 35 USC §101 is, through the Board of Appeals of Patent Interferences, the Court of Appeals of the Federal Circuit, or eventually the U.S. Supreme Court. These appeals options are expensive, time consuming, and uncertain since each client will likely be able to appeal only a limited number of applications. The USPTO can therefore exert incredibly strong pressures to force practitioners to comply. The USPTO might use Bilski as an opportunity to reduce its long-line of pending pure software patent applications. As prosecutors, we end should be prepared for an increase in rejections based on an “overly-abstract” specification, and should make every effort to make their applications more concrete, tangible, and/or machine-like.
What do the U.S. Supreme Court, the CAFC, and the USPTO have against pure computer software patent applications? Patents appear to be a poor mechanism to describe pure software innovations, and it is often difficult for examiners to precisely ascertain the state of the prior art with many pure software patent applications. Based on the complexity and variations of computer software source code and object code, how can one having ordinary skill in the art be certain how computer software really works? If a software subroutine is described in a patent application, how can a patent examiner, a judge, jury, or even a potential licensee be certain a very similar subroutine does not exist in another software? Because so much computer software is proprietary, complex, hidden, or unknown; how can a novelty or non-obviousness standard be applied to examine or compare different software? This uncertainty of operation likely is associated with the “abstract” label that the USPTO has tied to much of pure software that is not tied to machine or transformation. Open Source Software, for example, is one of the few software that outside users are free to examine or modify.
Consider many of the multi-technology software patents and applications incorporated at the end of this document. Almost all the independent claims to all these patent applications contain claim language that are more similar to mechanical, electrical, optical, medical, or other non-software arts. The machine or transformation limitation is being applied to the more mechanical or electrical directed claim language of multi-technology software patents, as compared to software-directed claim language of pure software patents. As such, these additional non-software arts in multi-technology software applications likely satisfy the machine or transformation test of Bilski.
Multi-technology software patent law has been much less mined than pure software patent law likely since more practitioners are capable of working on the latter, so the former tends to be more inventive-prone. In the advent of microprocessor controls, robotics, automation, fly-by-wire systems, computers, electronic-ink, displays, heavy industrial equipment, etc., it might be more straight-forward to find greater innovation. Such multi-technology software patents are immensely valuable since they contain such a broad range of equivalents perhaps across mechanical, electrical, hardware, software, and combined embodiments. Multi-technology software patents should be challenging to design around since they have such well set out roles of each of the mechanical, electrical, and computer software.
Many of such innovations actually require the software to operate properly, such as hybrid vehicles. Many of the major technical breakthroughs now involve such multi-technology software products. Consider smart power grids, hybrid technology, green technology, energy efficient technology, computer-controlled features, automated technology, and robotics; all of which are computer-software intensive as well as typically contain another underlying technology. Such multi-technology software are important to protect, and can be protected most effectively using multi-technology software techniques. The Apple iPhone, iPad, and iMac; as well as the Toyota Prius hybrid and Ford Fusion hybrid; are examples of multi-technology software since they include computer software as possibly combined with at least one mechanical, electromechanical, electronic, optical, or other innovative features.
Considering the various patentability standards for pure software and multiple-technology software patents from State Street to Bilski, from a prosecutor’s and former examiner’s viewpoint there has been more consistency than may appear. The vast majority of multiple-technology software patent applications that satisfy the Bilski machine or transformation test also likely satisfy the State Street useful, concrete, and tangible test. Similarly, the vast majority of pure software patent applications that fail the Bilski machine or transformation test as being too abstract would also likely fail the State Street useful, concrete, and tangible test. From the viewpoint of the prosecutors, our clients will appreciate the benefits if we work with the patent examiners to achieve what we both want.
By coldly applying the machine or transformation test of Bilski, the USPTO can act to effectively dispose of many of the pure software patent applications it might consider to have little innovation. There are clear benefits in applying innovative multi-technology software aspects to certain non-computer software patent applications that might, for example, be controlled by a processor to better enable the mechanical or electrical embodiments. It is unlikely that the USPTO will ever maintain that all computer software will fail Bilski. This would, in effect, return the patent system to before the State Street case is not an option since it is critical that our patents suitably cover such innovations realistically. Bilski does set forth a useful guideline that can be mechanically followed by patent examiners, as well as the patent courts. Perhaps, in this manner, Bilski allows the multi-tech software patent areas to become similar to the electrical, mechanical, and chemical areas as being one of the relatively straight forward examining areas that can be reliably and predictably examined by patent examiners with little input from the Patent Courts, as alluded to by the late CAFC Chief Judge Howard Markey earlier in this article. The Patent Courts of beer to indicate for inventions to be useful, they should not be abstract and should be more machinelike or tangible.
Patent practitioners do not need to have EE, CS, and ME degrees, like myself, to act as a multi-technology patent attorney. However, it indeed does help because we can tailor particularly suited patent protection for each situation. Perhaps mechanical, electrical, or other non-software patent attorney might interface more closely with computer software patent attorneys to derive more enabled, descriptive, or innovative patent applications by injecting software aspects. Basic computer software patent description could be integrated into the basic framework of many non-software applications. In crafting multi-technology patent applications, it might be most effective to use computer patent application framework, and then draft the mechanical, electrical, optical, transportation, imaging, or other non-software portion around the core computer software application. Reviewing the following writing samples should provide some indications how applications can be drafted more machine-like to satisfy 35 USC §101.
The above techniques provide some indication of filing higher quality, more innovative multi-technology patent applications. Multi-technology patent applications are largely based on a combination of software technical and legal aspects with a variety of electrical, mechanical, optical, medical, or other technical and legal aspects. A number of practitioners and corporate patent programs are much more effective at yielding more innovative patents, and therefore have a higher rate of allowance, than others. This raises the question of whether patent teams can be structure so the abilities of particular patent attorneys, or inventor, can be leveraged to yield higher quality patent applications based on combined patent law or combined technical skills. My experience suggests the answer to this question is yes, considering that patent drafting techniques, as with innovations, can always be improved.
Such innovative techniques as multi-technology software patents can result in improved innovation and increased rate of allowance, both for the individual practitioner as well as the corporate or firm patent department. There are other benefits to multi-technology software patents. Higher allowance rates correspond to greater value per patent issued. Patents that don’t issue can’t be licensed. If a patent is clearly patentable, it is likely to hold up better in court as well as in licensing. Whichever legal standard has been applied to computer software claims in the past years since State Street first allowed software patents, multi-technology software techniques have almost always fared better statistically in the USPTO and Patent Courts than pure software techniques. Though the “machine or transformation” test has difficulties, it consistently provides an innovative solution to a vexing computer software challenge. It is evident the US Supreme Court, the CAFC, and the USPTO support multi-tech software patents more than pure software patents. It is best not to go against the stated holdings of these three organizations as a patent practitioner, assignee, or inventor.
So why is it that some inventors, assignees, companies, and organizations have so much more effective of a patent program from others. By searching the USPTO Web Page for issued patents and non-issued patent applications, it is difficult not to observe that some assignees have well above fifty percent allowance rates while others are in the single digits. Having done patent work at patent firms and companies for both some highly successful assignees, as well as some not so successful assignees, there are some glaring differences I note.
First, most effective patent departments have a more direct communication path between the drafting patent attorney and the individual or team inventors. The more the patent attorney can act as the communicator of the inventors, the better. If there re multiple layers of review, or downright micromanagement, it can be very difficult for the drafting attorney to adequately represent the inventors innovation. By signing the oath or declaration papers, both the drafting attorney and inventors are affirming that there is to their best knowledge something new, novel, and non-obvious (patentable) within the application. If the reviewing attorneys are removing or severely modifying such matter, then there is a considerable risk that the innovation will become jumbled or outright lost. This risk becomes particularly high if the reviewing attorneys do not have the technical or legal expertise of the drafting attorneys.
This does not attempt to limit the review process particularly for inexperienced drafting prosecutors, but there has been a recent shift toward patent firms and companies whose experienced patent attorneys are arranged more “horizontally”, implying more direct client contact and less onerous reviews of the drafting attorneys by reviewing attorneys. Such horizontally structured patent firms and companies allow more personal interaction between the inventors and drafting attorneys, and often the cost per application is reduced due to more limited reviewing attorney involvement. Recently, a number of established old-line firms have failed, and evidently the more horizontal firms are giving them more than a run for their money.
Patent preparation and prosecution should be very positive, and even fun, largely because we are often dealing with inventors and companies dreams and futures. While it is unreasonable to believe that there will ever be no pressure at a patent department, those patent departments that do have high rates of allowance are generally much more positive and fun to work at than those that don’t. After all, what client wants to pay for an extensive portfolio when only a small percentage of the patent applications are being granted? The client can only receive royalties and can only limit others from making, using, or selling the patented object only if the patent issues. Additionally, inventors are keenly aware if their patent issues. As such, there are so many more reasons for those patent practitioners who can to use multi-technology software patent techniques in combination with traditional electrical, mechanical, chemical, and other well-established patent drafting techniques.
WRITING SAMPLES
The following are examples of the types of multi-technology software patents that I have drafted. Many of these incorporate legal, computer science, electrical engineering, and mechanical engineering aspects. These have always relied on innovative interaction between diverse engineers, programmers, etc.
U.S. Patent No. 5,054,995 –Apparatus For Controlling Fluid Compression System. My first multi-technology software patent for basic microprocessor-type controllers of “machines”, in this case compressors. These mechanical – software multi-technology software patents yield higher patentability than pure software. See also U.S. Patent No. 5,107,436 involving numerical control processing.
U.S. Patent No. 6,654,511 –Optical Modulator Apparatus. Relating to transmitting optical signals through the thin upper silicon layer of a silicon on insulator (SOI) wafer, modulating that optical signal, as well as coupling the optical signal to the waveguide to get the optical signal in and out of the waveguide. This patent could be used as a platform to build an optical computer or optical phone.
U.S. Patent No. 6,389,510 – entitled Method and Apparatus for Caching Web-Based Information.
U.S. Patent Nos. 7,623,625 and 7,627,085 –Compton Scattered X-Ray Visualizing, Imaging, or Information Providing. Also U.S. Patent No. 7,702,066 and others pertain to fluorescence imagers. These devices provide for localized depth imaging through tissue such as with a scope, or larger scale depth imaging using lower doses than other X-ray imagers. Very mechanical structure defined.
U.S. Patent No. 6,258,220 – Electro-Chemical Deposition System. One of Applied Materials first Electro Chemical Deposition, or electroplating, patent applications. Applied Materials now uses this technology extensively for metallization during semiconductor processing. Robotics-like description.
U.S. Patent Application S/N 2001/00083669 One energy-related patent application that I have drafted and am a co-inventor is a highly efficient (hybrid) jet engine. The increased fuel efficiency and aircraft range could approximate the 30 percent for hybrid automobiles, and also provide noise, weight, reliability, and safety improvements.
U.S. Patent Application S/N 20070080798, entitled Mote Device Signal Aspects and related to motes, which are minute sensor/control devices. A number of motes may be distributed throughout an area to define a mote network, some of which can communicate with each other. The energy level of such motes can be determined, and can be indicated by changing at least one color or surface feature. The user or automatic device can thereupon collect and/or recharge the energy-depleted devices.
U.S. Patent No. 7,708,493 – Modifiable Display Marker. Involves applying mote-type displays to roadways, etc. such that lanes can be modified and information or guidance can be provided to roadway vehicles. Toll booth and parking lot configurations can be modified.
