Recently I saw a commercial promoting wind turbine technology from General Electric (NYSE: GE). GE is heavily invested in wind energy, continually seeking patents on various improvements. What caught my attention, however, was the fact that the commercial said that these new wind turbines can “follow the wind,” which is a lay way of saying that there is sophisticated software at play that is capable of improving the efficiency of windmills that employ the technology.
Indeed, for the past several years GE has been promoting what they call PowerUp, which can increase efficiency of existing wind farms by up to 5%.
“PowerUp is essentially software and control setting changes that allow us to increase the level of production from turbines that are already on the sites,” says Patrick Woodson, CEO of Eon Climate & Renewables North America Wind. “In some ways it like taking the cars we have that stay in fourth gear and finding fifth gear for them.”
“We will go in and change various parameters within the turbine, for instance in pitch, and they’ll start to narrow that band down to find the optimum blade angle at any given condition,” says David Riley, Site Manager, Pyron GE Wind Services. “So no real mechanical type modifications.”
But is any of this patentable?
I know that GE gets patents on software related wind turbine inventions. For example, within the last two weeks GE received U.S. Patent No. 9,109,577, which covers a method of operating a wind turbine. The method disclosed specifically seeks to detect malfunctions by comparing current operating values with historical reference values. The patent explains that the method and system “enables a better cause detection,” and allows trouble shooting “by even uneducated personnel… because most or all data may be gathered automatically out of difference sources and used for finding the root cause for a fault or malfunction.”
But doesn’t this just express an unpatentable idea in light of the Supreme Court’s current software jurisprudence?
Some time ago I shared an e-mail exchange with a patent practitioner who wished to remain anonymous. “I believe that there is no invention, whether software-implemented or otherwise, that cannot be distilled to some ‘abstract idea’ plus an amalgamation of known elements,” explained this experienced patent prosecutor who handles software patent applications regularly. “I’ve seen one office action after another in which the applicant’s preamble was simply declared to be an ‘abstract idea’ without any rationale or evidentiary support. In short, I see that Alice-based 101 rejections have simply become a crutch to Examiners, a way to avoid the hard work of actually examining the claims.”
Therein lies the problem with the Supreme Court’s software related patent jurisprudence. On some basic level everything can be characterized as an idea. It is also all too easy for those who are not technically trained to believe, no matter how wrongly, that implementation is a trivial or ministerial act. Just monitor the windmills, if they are operating at a less than optimal level adjust them, tilt the blades a little. No big deal. Anyone could have thought of that, and a college student could have written the code over a weekend. Moreover, windmills are extremely old technology, so merely applying a computer process to something so old can’t be patent eligible.
Sadly, that is what passes for Supreme Court understanding. The reality, however, is that implementing software on a large scale to make an existing mega-machine (like a windmill) work better is anything but trivial.
So what is the distinguishing aspect between software that runs a windmill more efficiently, providing up to 5% greater efficiency, and the financial methods that the Supreme Court has heretofore considered? That is the $64,000 question, but one for which we really don’t have an answer. Clearly, there has to be a difference. Software makes machines run better in many instances, and in some cases without software the machine couldn’t run. But the machine itself, at least in terms of its structural components, does not change. All that changes is the operating instructions, which are dictated by software. Requiring some uniqueness in the tangible components in order for the software to be protectable is preposterous, particularly in an innovation economy where over 50% of all innovations in one way or another have a software component.
The real problem with Supreme Court jurisprudence on software is that it relies on the yet to be defined “abstract idea doctrine.” The Supreme Court has never defined the term “abstract idea,” but they continue to apply the doctrine nevertheless. Unfortunately, there seem to be only nine individuals in the world who know what an abstract idea really is. They seem to be in unanimous agreement about what the term covers, but unwilling to share their knowledge in a coherent way that would distinguish between innovations that are patent eligible and ideas that are not patent eligible.
This brings me back to wind power. If all you do is tilt the blades a little one way or another based on data that suggests efficiencies could be had by doing so how is that anything other than an idea? The fact that it would be hard to accomplish and might take a very large team of engineers and programmers years to figure out doesn’t really mean anything under the current test, which is why the current test for patent eligibility makes absolutely no sense whatsoever.
How difficult it is to accomplish a task is a relevant consideration when we are talking about obviousness under 35 U.S.C. 103, but the level of difficulty is irrelevant under a 101 patent eligibility inquiry. There are no procedural or legal safeguards when 101 is applied with brute force. The abstract idea doctrine swings like Thor’s hammer, rather than precisely cutting like a scalpel. Historically, 101 jurisprudence was viewed as a threshold inquiry that would allow virtually everything through, acting as a bar only to weed out the most fundamentally unpatentable innovations. It is precisely because 101 has never been historically used that there are few, if any, procedural or legal safeguards to ensure that true innovations are not unnecessarily and inappropriately declared patent ineligible.
For some time now Martin Goetz, who is widely recognized as the recipient of the first software patent ever granted, has been arguing that we really need to stop referring to “software patents” because the term misses the entire point and focuses decision making attention on something that is nothing more than a red herring. It matters not how the innovation is brought into being, what matters is whether there is an innovation. Earlier this year Micky Minhas, who is Chief Patent Counsel at Microsoft, told me something very similar. Whether you use hardware or software is largely a design choice and it should not have any affect on patent eligibility if you choose one design path over the other.
Is GE’s system and method for providing up to 5% greater efficiency for wind farms an innovation? Clearly it is an innovation and one that should be patent eligible regardless of how it is brought into being. Hopefully someday soon the Courts will focus on the innovation and lose their irrational fear of software patents. After all, the patent system is meant to protect innovations, providing a critical incentive for the private sector to innovate in the first place. Picking winners and losers based on how efficiencies are obtained is unsound, unwise, and just plain stupid.