First published in Robotics Business Review
Robotics and Section 101
The robotics industry is no stranger to the patent system, and has long employed the patent system to protect robotic inventions. Today's robotic devices, however, are a blend of complex mechanical and electronic technologies, with the invention often lying in the intersection between hardware and the software that controls it. Will such software-based inventions remain eligible for patent protection?
Unfortunately, today there is no easy answer. On March 31st, the U.S. Supreme Court will hear argument in review of the Federal Circuit's decision in CLS Bank International v. Alice Corp.1 and, when the Court renders its decision—expected by June 2014—we will hopefully have answers. In CLS Bank, an en banc court of the ten sitting judges at the time found the software method and computer-readable media claims in question were not patent eligible, but disagreeed as to why or what test should be used. The court split 5-5 as to the patent-eligibility of the corresponding system claims.
The debate over whether software-based inventions should be patent-eligible has raged on for many years. Those arguing "no" say patenting software thwarts innovation and is to blame for the perceived increase in patent litigation by non-practicing entities. Those on the other side agree with then-PTO Director David Kappos, who acknowledged that "many breathtaking software-implemented innovations power our modern world, at levels of efficiency and performance unthinkable even just a few years ago,"2 and are therefore deserving of patent protection. This statement is no less true of robotic inventions.
Patent eligibility should not be confused with determinations as to whether an invention is "new" or "non-obvious." Patent eligibility is the determination of whether the subject matter of a claim should be considered for patent protection, and is governed under Section 101 of the U.S. Patent Act of 1952, which provides that "any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof" is eligible for a patent.
In two early Supreme Court cases (Gottschalk v. Benson in 1972 and Parker v. Flook in 1978), the Court found computer-implemented claims not patent-eligible holding that the processes claimed therein were merely abstract algorithms with no practical applications. But after noting in Diamond v. Chakraborty in 1980 that Congress intended statutory subject matter to "include anything under the sun that is made by man," the Supreme Court opened the door for software patents with its 1981 decision in Diamond v. Diehr, in which it confirmed that processes implemented as computer programs may be specific applications of abstract ideas deserving of patent protection.
In 1982, Congress formed the U.S. Court of Appeals for the Federal Circuit and tasked it with hearing appeals in all patent cases with the hopes of bringing uniformity to patent law. Over the past 32 years, the Federal Circuit has issued a series of opinions attempting to apply the Supreme Court's guidance in Diehr to software and evolving technologies like robotics, and have proposed a series of tests for determining whether a particular claim is patent-eligible.
Using these tests, and guidance by the U.S. Patent and Trademark Office, patent practitioners had been able to draft claims that pass the patent-eligibility hurdle by, for example, ensuring the claims recited software that performed a transformation on data or was implemented as a tangible embodiment tied to particular hardware, operated upon by a computer processor, or stored on a computer readable media.
But in 2010, when considering business method claims in Bilski v. Kappos, the Supreme Court overturned the Federal Circuit's most recent "machine-or-transformation" test, but failed to provide an alternative.
Worse, the Federal Circuit seems unable on its own to provide new guidance. Opinions that have issued since 2010 have shown that the Federal Circuit is evenly split on nearly everything —from the weight to give to Congress' intent when drafting Section 101 and whether Section 101 should operate as a "threshold test" to how to interpret the wisdom of earlier Supreme Court cases and what "test" should be used to evaluate patent eligibility—as evidenced most recently by its fractured CLS Bank opinion.
The opinion has caused alarm in the high tech community, as it is by no means clear which way the Supreme Court will lean with its decision. Will the Court agree with the camp led by Judge Lourie which advocates requiring that method claims recite "significantly more" than the underlying abstract idea in order to be patentable?
Doing so would threaten to stop software-based inventions at the first threshold, given the Lourie camp's observation that "[a]t its most basic, a computer is just a calculator capable of performing mental steps faster than a human could. Unless the claims require a computer to perform operations that are not merely accelerated calculations, a computer does not itself confer patent eligibility."
Or will the Supreme Court agree with the camp led by Chief Judge Rader, which argues that it was Congress's intent for Section 101 to act only as a coarse filter, and as a "dynamic provision designed to encompass new and unforeseen inventions," such as softwareimplemented inventions? Moreover, a computer is a "machine" and a program is a "process", and both are expressly eligible subject matter under Section 101. Further, while virtually everything done by a computer can be done by a human, specific applications should be patent eligible.
Without Supreme Court guidance, the Federal Circuit seems destined to continue on its fractured path, issuing panel-dependent decisions like the recent SmartGene, Inc. v. Advanced Biological Laboratories, SA. In SmartGene, the three-judge panel (which included two members from the Lourie camp in CLS Bank) found the claim in question not patent eligible because it did not "identify any steps beyond those that doctors routinely and consciously perform" in their heads, and did not require any computer hardware beyond existing computers long in use.
So what does this mean for robotics and patents claiming inventions implemented in software? Until the Supreme Court rules this June, there is no clear test or method for determining patent-eligible subject matter. If the Supreme Court favors a test like that of the Lourie camp —one in which claims are systematically stripped down to an "inventive concept" a la Flook and Benson—it seems hard to imagine any claim that would be sufficiently "nonabstract" to survive the patent-eligibility hurdle.
A standard that acts as a coarse filter at the patent-eligibility stage, as advanced by the Rader camp, would appear to be more in line with the statute and legislative history. Contrary to popular belief, finding almost all claims patent eligible would not add to the existing problems of potentially undeserved patents, as such claims will continue to fail when considered for novelty or non-obviousness.
In any case, it is now up to the Supreme Court to resolve the very divisive issue of patent eligibility of software inventions. It seems highly unlikely, and contrary to legislative history and Supreme Court precedent, that all software-implemented inventions will be categorically barred. The fear of many in the high tech community, however, is that the scope will become unreasonably narrow and unfairly prevent some otherwise eligible inventions from being examined. Hopefully, the Court will reach a thoughtful balance and provide much-needed workable guidance to the robotics community.