California Court's Bilski Analysis
When does a general purpose computer become a "particular machine" as required by Bilski?
Yesterday, I wrote that a California court had invalidated claims of DealerTrack's U.S. Patent No. 7,181,427. According to Judge Guilford's Order, DealerTrack had asserted claims 1, 3, and 4. The defendants moved for summary judgment arguing that the asserted claims are invalid in view of Bilski. DealerTrack implicitly conceded that its claims do not meet the transformation prong of Bilski's machine-or-transformation test by apparently failing to argue that point in its response to the defendants' summary judgment motion. The Order finds that the claims also do not meet the machine prong. Because the claims meet neither prong, the Order found them to be invalid.
The Order begins its analysis by outlining the current law of the land that a general purpose computer executing a program is patent eligible. Next, the Order indicates that decisions of the Board of Patent Appeals and Interferences (BPAI) and the California decision in Cybersource v. Retail Decisions require more than a general purpose computer, thereby contravening In re Alappat (holding that a general purpose computer becomes a special purpose computer when it executes a program).
According to the Order, "[t]he ‘427 Patent does not specify precisely how the computer hardware and database are 'specially programmed,' and the claimed central processor is nothing more than a general purpose computer that has been programmed in some unspecified manner." (Page 7, lines 1-4.) The Order cites to an earlier Claim Construction Order that indicates that DealerTrack agreed to a broad reading of the relevant devices, clearly now to its detriment.
The Order states that "[t]he Court finds that none of these devices constitutes a 'particular machine' within the meaning of Bilski." (Order, page 6, line 28-page 7, line 1.) However, the Order does not cite to anywhere in Bilski as specifying what the "particular machine" ought to be. It goes on to state:
Under Bilski and the recent decisions interpreting it, the central processor in this case cannot constitute a "particular machine." Further, this Court’s claim construction order defined both the "remote application entry and data device" and "remote funding source terminal device" to include "any device," such as a personal computer or dumb terminal, and these devices clearly cannot constitute particular machines.
(Order, page 7, lines 4-9.)
When does a general purpose computer become a "particular machine" as required by Bilski? Can it become one? No one seems to know yet.
Bilski's algorithm theoretically could be performed by hand, but by the time a human does it, the data wouldn't be any good anymore, and the claim would be technologically useless/impossible. So even though the most interesting claims did not include computational hardware, the CAFC had to inherently imply that adding a GPC is insufficient to create a particular machine because humans can't realistically perform the computations. It seems this is the route Guilford and Patel have taken in their post-Bilski opinions.
The more interesting issue is that the UK House has gone down the path of analytical dissection, requiring that the claims excluding the software must contain something that is independently eligible. Although the EU directives are somewhat of a mess right now (and they're still disputing the meaning of Article 4), I think this may be the direction we head down too. After all, the prohibition on analytical dissection is only mandated in 103, not 101. Dicta in Diehr suggests otherwise, but it's certainly not in the statute. Also, if we were required to look at claims as a whole for purposes of 101, Diehr could have just said that the system has a rubber molder, some computational hardware, and this constitutes a particular machine (adding the algorithm to either the molder, the GPC, or both). But SCOTUS never found a particular machine in Diehr -- they spent the whole opinion finding a transformation, while repeatedly running through PMoT.
In the big picture, some of the biggest litigants simply can't argue 101. Since certain top tier developers have huge software portfolios, attempting to invalidate a software patent through 101 is like a dentist treating a recurring toothache by cutting off the patient's head.