Mr. Lam (“Plaintiff”), the owner of Taiwanese patent 420783 (“’783 patent”) titled “Commandless programmable controller” granted in 2001 sued Nyquest Technology (“Defendant”) for infringement. The IPC Court as both the first and second instance court held that the ‘783 patent was invalid. The Supreme Court vacated and remanded for the first time in July 2022.[1] In November 2023, the IPC Court reviewed the case again and ruled in favor of the Defendant.[2] Tsai, Lee & Chen had reported the judgement.[3] The Defendant appealed again. In November 2024, the Supreme Court made a judgement to vacate and remand back to the IPC Court for the second time.[4]

  The Supreme Court’s analytical elaboration to define the inventive step requirement is a highlight in the judgement. The Court stressed that, in the comparison with the prior art to determine whether an invention meets the requirement of inventive step, some examination principles are to be followed including: (1) examination of the invention as a whole rather than on only the specific or partial technical features, (2) comparison against multiple references or common knowledge in combination, and (3) opinion to be made for each claim in the manner one after another.[5] Based on the “technical problem the invention to be solved” and the “technical means to solve the problem,” one should proceed to follow the steps below:

  1. Define the scope of the claimed invention.

  2. Identify the technical content revealed in the relevant prior art.

  3. Ascertain the level of technical knowledge of a person skilled in the art within the relevant technical field.

  4. Determine the differences between the claimed invention and the content disclosed in the relevant prior art.

  5. Ascertain whether a person skilled in the art, in view of the content disclosed in the relevant prior art and the existing technology or knowledge before the filing date, could have easily achieved the claimed invention.[6]

  It is critical to note that individual components or steps of the invention cannot be simply disassembled and then mechanically combined and compared against the prior art. Instead, it is important to first identify the "closest prior art" (primary reference), which refers to a single prior art document that provides the optimal referential basis on which the invention was created. This is the prior art reference from which the inventor, starting with the disclosed technical information, would most likely have been led to achieve the invention. The differences between this prior art reference and the claimed invention should then be analyzed. This approach helps avoid a mechanical assembly or combination of prior art, which could lead to an erroneous hindsight bias in determining inventive step.

  Furthermore, the Court adopted the EPO’s could-would method. Namely, in the Step Five above whether the claimed invention can be easily achieved by a person skilled in the art, the Court deemed it necessary to find that skilled person “would” have been prompted to make an improvement over the prior art and merely that the skilled person “could have made” the invention is not sufficient. In other words, the key element to determining inventive step lies not only in whether success is theoretically possible but also in whether, in a given case, there exists incentives, support with specific facts, or motivation to drive the skilled person to engage the research activities which finally lead to an inventive success.

  In the present case, the Court found the lower court erred in jumping straight to a conclusion before having a comprehensive review of facts. Specifically, the lower court did not reason why the plaintiff’s arguments were not accepted. In the review of validity issue, the plaintiff contended that a mechanical selection of features from a reference to combine with other two references to reject the inventiveness of the claimed invention was wrong. Some combinations are scientifically illogical and hence conceptually dissuaded. The lower court’s acquiescence on the arbitrary combinations of references was equivalent to turning a blind eye to the “teaching away” factor. The Court opined that the lower court’s failure to start with citation of a primary reference and the absence of reasons to negate the plaintiff arguments were wrong. The laws were erroneously applied and the reasons were insufficiently provided in the lower court’s judgement.

  As per Taiwan’s evidence rule, the court shall investigate the evidence presented by the parties unless the court deems it unnecessary.[7] An “unnecessary” situation happens in only one of the two circumstances. Firstly, the evidence is irrelevant to the facts; or, secondly, the court has already established an opinion on the matter. In other word, when the presented evidence is relevant to the facts, it is unnecessary and then bypassed investigation. In the present case, the defendant submitted an expert witness report for supporting its non-inventive assertion. The plaintiff questioned the report for lack of substantial evidentiary value. As the challenge was relevant to the pending facts, that the lower court failed to conduct an investigation was therefore illegal.

  The Court also stressed the important role of secondary factors by circumstantial evidence which the lower court should not put a light weight on in the evaluation of inventive step. According to the evidence rule, when necessary for finding out the truth, the court may conduct evidence investigation on its own discretion.[8] Categorically, the admissible evidence includes both the direct evidence and the circumstantial ones, the later meaning the evidence to corroborate another fact that can be further derived to reach the asserted and pending fact. Here, the time of inventive step decision when the patent was filed far departs from the time of an infringement lawsuit. The difference in time potentially leads to subject biased mistake. The court should consult the objective facts for better making a determination. Secondary factors - long-felt needs, failure of others, unexpected results, and commercial success - stem from economic and motivational aspects of the inventions. If they are to provide explanatory and objective information for the evaluation of inventiveness, the court should consider those factors in a pursuit to reenact the then research and/or commercial reality at the time of filing.

  The plaintiff argued that his claimed invention was granted to patent in the US, the UK, Mainland China, Japan, etc. in addition to Taiwan. The ‘783 patent has been licensed to many stock exchange-listed corporates including Realtek Semiconductor, ELAN Microelectronics, Sunplus Technology, etc. The long-felt technical barrier were resolved and commercial success was attained since the claimed IC chips were invented and sold overseas in volume of billions, the plaintiff insisted. As these stories were supported by paper documents, the Court demands that the lower court enquire said licensee companies regarding the technical impact of the ‘783 patent to their business and the details of business negotiations for the license agreements, in order to find out more information for the secondary factors to re-evaluate the inventive step.

  To summarize, the Court concluded that the lower court failed to follow examination principles for inventive step and the fundamental rules of evidence. Due to the clear errors in the lower court’s judgement, the case was remanded for further proceedings.

  The judgement emphasized several rules for inventive step determination. Specifically, the judgement survey through the concepts and principles of the five-step examination, the primary reference citation, the could-would methods, and secondary factors in detail. The judgement serves meaningful value for inventive step analysis in detail for any upcoming cases.