Software Patent Claims Saved by Problem-Solution Approach
The U.S. Court of Appeals for the Federal Circuit reversed a lower court’s decision that software patent claims involving techniques for varying check data to verify data transmissions were ineligible for patent protection in KPN v. Gemalto, No. 2018-1863, 2019 U.S. App. LEXIS 34075 (Fed. Cir. Nov. 15, 2019). The lower court had found that the claims involved nothing more than data manipulation using conventional computer hardware. Following the trend for courts to find data manipulation to be nothing more than an abstract mental process, the lower court invalidated the claims.
The Federal Circuit pointed to the patent specification’s clear description of a problem with prior art verification techniques and specific solutions to that problem. Conventional data verification techniques generally relied on fixed functions for generating check data. Using those techniques, systematic errors repeating the same kind of defective transmission over and over could slip through. The patent stated that varying the way the check data was generated made verifying the transmission of the data using the varied check data more reliable. Specific techniques for varying generation of check data included periodically changing the generating function, adding random numbers to the input data, or changing the bit position in data blocks of the input data. Even if an error were able to slip through one specific technique temporarily used to generate the check data, varying the techniques over time reduces the likelihood that same error could still slip through. As a result, the danger that data transmission errors, particularly systematic errors, would slip through is greatly reduced.
Interestingly, the patent owner did not challenge the lower court’s determination as to it broadest independent claim. That claim required changing bit position in the data to be transmitted, but did not tie the changes in bit position to any variation over time. The dependent claims did require modifying the way bit position was changed in time. This specificity more explicitly tied the claims to the techniques and benefits discussed in the patent disclosure.
The Court explained that the remaining claims recited a specific technique for varying generation of check data, producing an improvement to an existing technological process. It compared the claims favorably to those confirmed patent-eligible in Finjan, which used a set of specified techniques to scan files not just for known viruses but for potentially dangerous or unwanted operations. The patentee’s strategy to focus only on the narrower, more specific claims also undermined that accused infringer’s argument that the improvement disclosed in the software patent was not sufficiently reflected in the claims themselves.
This decision highlights that patent claims to computer-implemented inventions in the U.S. are more likely to survive a challenge when the patent owner presents a clear narrative why the claims represent an improvement over the prior art and how the improvement is achieved using specific limitations recited in the claims. It also is emblematic of how close the U.S. approach to subject-matter eligibility has come to the problem-solution approach in Europe, particularly for software patents. While care should be taken in crafting a patent specification to ensure a patent applicant is not admitting that a given problem or solution was known in the art, drafting a patent application in such a way that a problem with conventional approaches can be identified and a specific solution to that problem, with a disclosed benefit, is recited in the claims may set up a patent application for better success domestically and abroad.
The information you obtain at this site is not, nor is it intended to be, legal advice. You should consult an attorney for advice regarding your individual situation.
Author: James Watson