08 January 2011

NZ IPO Guidelines Try to Render "Embedded Software" Patentable Without Specifying this Legal Term

In April 2010, the parliament of New Zealand voted for a major Patents Reform Bill to tighten the standards of patentability of software-implemented inventions (see related posting). The bill, as drafted by the Select Commerce Committee in July 2010, accepted that "protecting software by patenting is inconsistent with the open source model" and that "computer software should be excluded from patent protection as software patents can stifle innovation and competition" - intensely accompanied by various lobbying organisations. Clause 15 (3A) of the Patents Bill now reads:
A computer program is not a patentable invention.
The Select Commerce Committee, however, considered an exception for "embedded" software but found it too difficult to define it in the bill, whereas the Ministry of Economic Development (MED) decided that machine-controlling software should be treated as a special class that may be patentable. It was then confirmed by Minster Simon Power that, while "further amendment to the bill is neither necessary nor desirable", the NZ Intellectual Property Office (IPONZ) was instructed to develop guidelines to allow inventions that contain embedded software to be patented, since it was accepted by the Committee and the Minister that companies investing in inventions involving "embedded" computer programs should be able to obtain patent protection for these inventions (see related posting).

Apparrently, like other jurisdictions before, also New Zealand's legislation failed to clearly specify the distinction between patentable and unpatentable software in a bill ("to difficult to define"). Now, on 20 December 2010, the instructed executive authority (IPONZ) has issued draft guidelines that should ensure that (only) microprocessor-based products controlled by 'embedded software' are not excluded from patentability. Comments on the proposed guidelines can be sent to patentsbill@med.govt.nz until 11 March 2011.

The draft guidelines clarify New Zealand's approach to statutory exclusions to patentability, based on political considerations according to which only "embedded computer programs" should remain patentable. The draft guidelines observe that there are similarities between certain provisions of the Bill and both the Australian and UK patent legislation. Specifically, the general requirements for patentability are similar to the Australian provisions, being based on a judicially-developed 'manner of manufacture' test, while the concept of an exclusion for computer programs is similar to the UK (or European) scheme.

With respect to the desired exclusion of computer programs, the European Patent Office's approach to patentability, however, has been considered potentially problematical in New Zealand's context especially with respect to the required "technical effect" (this criterion is in fact inappropriate to exclude everything but embedded software), while substantial similarities between UK and New Zealand patent law have been recognised with respect to the statutory framework of the UK Patents Act 1977.

Against this backdrop, the IPONZ considered the four-step Aerotel test of the English Court of Justice (see also case study of UKIPO) as a basis for New Zealand's way of assessing whether or not a claimed invention falls within an excluded area. Due to differences between the proposed law of New Zealand and the UK law, the Aerotel test has not been adopted in its entirety but only on a conceptual level. Especially, the fourth step of the original Aerotel test, asking whether the contribution is actually technical in nature, has been found non-adoptable and thus omitted. The 'adapted' Aerotel test consist of the following three steps:
  1. Properly construe the claim;
  2. Identify the actual contribution; and
  3. Ask whether the contribution falls solely within the excluded subject matter,
Identify the Actual Contribution: Upon identifying the invention's actual contribution "to the stock of human knowledge" the substance of the invention is determined (what is the invention?) based on the properly construed claims. Since the "form" of the claims, i.e. the actual claim wording used by the applicant, is disregarded, the identified actual contribution may differ from what was (meant to be) claimed. For identifying the invention's substantive essence, four considerations have been found useful by the IPONZ:
  1. The problem said to be solved (by the invention);
  2. How the invention works;
  3. What its advantages are (over known products or processes); and
  4. What, if anything, the inventor has added to the stock of human knowledge.
The so determined actual contribution must also observe the Australian "manner of manufacture" test, assessing that the claimed subject matter produces (i.) an end result that is "an artificially created state of affairs" in a "field of economic endeavour" and (ii.) a "physical effect", following Australian case law based on NRDC and Grant). 

With respect to the required "physical effect", the guidelines explain that the mere "presence of any physical interaction [...] would not be sufficient to render something patentable, so that, e.g., upon performing an arbitrary method on or with a physical apparatus, "the physical changes must be of consequence in that they are central to the contribution (invention) and not peripheral", following Australian case law based on Invention Pathways (see earlier posting). Further,
the normal interactions between a computer and computer program would not be considered to be an invention if it did not produce a "physical effect" [...]. For the purpose of the granting of a patent for an invention which contributed to the stock of human knowledge, something more is needed for it to be a "manner of manufacture" if the writing of the data to the computer memory is the only "physical effect".
By integrating the Australian "manner of manufacture" approach into the second Aerotel step, the
fourth Aerotel step, asking for the technical character of the invention's contribution (manifesting the European idea of a technical invention), could be omitted, so that after determining whether a "physical effect" exists, there is no perceived need anymore for a further step of checking technical character.

Solely falling within excluded subject matter: The draft guidelines explain that the third Aerotel step is intended to make sure that a computer program that falls within the exclusion from patentability cannot be rendered patentable by simply adding some non-excluded claim language, e.g. by structurally reciting the program as a component of an apparatus (wouldn't that be embedded software?).

In fact, the IPONZ recognised that the UK Court of Appeal considered the "solely within excluded area" criterion comparable to the "as such" criterion of Art. 52(3) EPC and ("the third step is merely an expression of the "as such" qualification; see Aerotel/Macrossan, § 45). However, it is the central purpose of the EPC's "as such" criterion to, generally speaking, differentiate between abstract and conceptual computer algorithms and concrete implementations in a technical environment. That is, according to the EPC approach, a computer program can be guided out of the excluded "as such" area by adding technical details (which of course doesn't mean that the teaching is novel and inventive), as recently acknowledged in the G 3/08 opinion of the Enlarged Board of Appeal (see related posting). Similarly, one should expect that a computer program can be prevented from "solely [i.e. exclusively] falling within the excluded area" by adding appropriate non-excluded features, e.g. by sufficiently specifying a computer program as a machine-controlling (embedded) software.

The IPONZ's related explanation under § 31 of the guidelines, according to which "it may be that the contribution does not fall wholly [solely] within the scope of the exclusion to computer programs, but may still be excluded for example, as a mere scheme or plan within the business context", appears somewhat odd to me, since a software-based business method would already be excluded as lacking a "physical effect" according to the adopted "manner of manufacture" test.

Under §§ 27 to 29, the guidelines suggest a number of questions whose answers should provide a better sense of whether or not a computer program's actual contribution as a whole (solely) falls under the exclusion of clause 15 (3A):
  1. Is the artifact or process new and non-obvious merely because there is a computer program? If yes, then it is excluded.
  2. Would the artifact or process still be new and non-obvious in principle even if the same decisions and commands were undertaken a different way? If no, then it is excluded.
  3. Is the computer program merely a tool adapted to achieve efficiency or something similar? If yes, then it is not likely to be excluded.
Conclusion: In August 2010, I posted that "New Zealand allows patents on embedded software but fails to provide a definition". Honestly, I still cannot see how the IPONZ''s draft guidelines could possibly execute this political demand as an exception of the general legal provision that computer programs are not patentable in New Zealand.

After all, claims on embedded software should be considered patentable according to the IPONZ's draft guidelines, since their actual contribution normally goes beyond that of a plain computer program, they exhibit a "physical effect" and they do not as a whole fall within the computer program exclusion of clause 15 (3A). My current feeling is, however, that there might exist numerous tangible technical software implementations that cannot be regarded as pure embedded software in the technical sense of the term but would nevertheless be patantable based on the suggested scheme.

As to the extend of excluded computer programs, Australian patent attorney Mark Summerfield commented on his Patentability blog that
this will most likely exclude computer-implemented business methods and 'pure' data processing applications (which generally lack a 'physical effect'), as well as the automation or computerisation of known processes or devices (which lack an 'actual contribution' beyond the software implementation). 

However, it is not clear whether this approach will exclude novel processing of data that represents physical artifacts (e.g. image processing of MRI or X-ray scan data) where the only 'physical effect' may be in the quality or content of a computer display.  We find this uncertainty to be of some concern.
Further, Jonathan Lucas, a New Zealand patent attorney, asked
The guidelines were intended to cover the examination of patent applications for embedded software.  How can they successfully do this when they don’t use the term “embedded” once?

IPONZ guidelines have been ignored by the courts in the past because they have no official status. How much influence will the guidelines have in a court of law.