When deciding on unity of invention, it is mandatory under Art. 82, R. 44 EPC and under R. 13.1 PCT to determine whether or not a group of inventions claimed in an application forms a single general inventive concept. Thus, when seeking to establish at the search stage whether inventions comply with the requirement of unity, the first question to be considered is whether the inventions are linked by a single general concept.
In W 19/89 the board decided that the application clearly lacked unity of invention, since the four possibilities comprised by claim 1 related to a further development of the state of the art in different directions, namely by employing different classes of dehalogenation agents having no new technical feature in common. Where at least one solution of the underlying technical problem already formed part of the state of the art, the requirement of a "single general inventive concept" implied that the further solutions to that problem proposed in the application must have at least one new element in common, this new element being normally represented by at least one new technical feature. Since the absence of such a unifying new technical feature was admitted by the applicant, the application related to more than one invention.
A useful analysis of the single general concept was made in W 6/90 (OJ 1991, 438). The board found that this concept manifested itself in features common to different teachings expounded individually in the same application. It observed that a teaching for the purposes of patent law encompassed not only the immediate subject-matter representing the solution to the problem as defined in the relevant claim, but also its technical consequences which were expressed as effects. It was noted that any subject-matter was defined by structural features and the relationship between them. The relevant effects, i.e. the outcome or results achieved by the invention as claimed, would usually already be apparent from the problem as stated. A single general concept might therefore be said to be present only if a partial identity existed between the teachings in an application deriving from the structural features of the subject-matters claimed and/or the outcome or results associated with those subject-matters.
Where subject-matters of the same category were concerned, a partial identity, generating unity of invention, could result from the structural features of these subject-matters and/or their associated effects. The absence of such an element common to all the different teachings in the application, and hence a lack of unity, might be established a priori under certain circumstances. A lack of unity might, however, also be established a posteriori between the subjectmatters of different independent claims or in the remaining subjectmatters if the subjectmatter of a linking claim was clearly not novel or inventive visàvis the state of the art. The board gave an example of what was meant by the abstract term "single general concept": a product, a process specially adapted for the manufacture of the said product, and a use of the said product, for example, embodied a single general concept because, on the one hand, the partial identity between the product and its use derived from the structural features of the product and, on the other hand, the partial identity shared by the product and the process specially adapted for its manufacture also derived from the product which was to be considered as the effect or result of this process (see T 119/82, OJ 1984, 217).
The board also noted that the criteria governing unity of invention for the purposes of R. 13.1 PCT elucidated above also applied in principle where the inventive step was based chiefly on the discovery of an unrecognised problem (see T 2/83, OJ 1984, 265). If the common problem, i.e. the effects to be achieved, was itself already known or could be recognised as generally desirable (a mere desideratum) or obvious, there would be no inventive merit in formulating the problem. If the common structural features were to be found only in the prior art portions of the claims, and if these known features did not help to solve the problem of the combined whole, this might also indicate a lack of unity.
In W 38/90 the sole common link between the subjectmatter of claims 1, 2 and 4 was that the respective features all contributed to the realisation of the same device - a viewing assembly for a door. The board observed that this common precharacterising feature was confined to indicating the designation of the invention, as part of the prior art; it was not, however, a specific feature interacting with the remaining characterising features and thereby contributing to the various inventions and their effects. Therefore, this sole link had to be rejected as irrelevant as regards any possible contribution to inventiveness. As already concluded, in the absence of any kind of common relevant feature, no unity "a priori" could be recognised.
In W 32/92 (OJ 1994, 239) the board held that there was no unity of invention where the subject-matters of independent claims, including their effects, had no common inventive feature in those parts of the claims differing from the closest prior art.
In T 861/92 the apparatus claimed was specifically adapted for carrying out only one step of the claimed process. The board came to the conclusion that the subject-matter of the claims satisfied the requirements of Art. 82 EPC 1973 as there was a technical relationship between both inventions.
In W 9/03 the board considered whether the common problem underlying the claims could establish unity between the different inventions. The board held that a common problem could establish the unity of different inventions only under certain conditions, for example in the case of a problem invention. That was not the situation in the case at issue, however. According to the applicant, the common problem was considered that of providing a more secure, faster and easier connection setup. The board started from the assumption that the person skilled in the relevant art would normally consider this problem. The common problem was therefore so broadly formulated that it was "itself already known or could be recognised as generally desirable or obvious" (see W 6/90, OJ 1991, 438); as a result, unity could not thereby be established.