CAFC: Relative Dimensions Not Explicitly to Scale in Prior Art

from Nystrom v. Trex Co., slip op. 03-1092 (Fed. Cir. June 28, 2004):

"The district court’s acceptance of TREX’s invalidity arguments based on models made from drawings contained in the Zagelmeyer patent was incorrect. The basis of the district court’s summary judgment of invalidity was a model that TREX developed based on that reference, and not on drawing dimensions or a written disclosure of dimensions contained directly in the patent itself. Under the principles set forth in our prior cases [below], the speculative modeling premised on unstated assumptions in prior art patent drawings cannot be the basis for challenging the validity of claims reciting specific dimensions not disclosed directly in such prior art. Thus, we conclude that the district court erred in granting summary judgment of invalidity based on TREX’s models.

Hockerson-Halberstadt indicated our disfavor in reading precise proportions into patent drawings which do not expressly provide such proportions:

The ’792 patent is devoid of any indication that the proportions of the groove and fins are drawn to scale. [The patent owner’s] argument thus hinges on an inference drawn from certain figures about the quantitative relationship between the respective widths of the groove and fins. Under our precedent, however, it is well established that patent drawings do not define the precise proportions of the elements and may not be relied on to show particular sizes if the specification is completely silent on the issue.

Hockerson-Halberstadt, 222 F.3d at 956 (citing In re Wright, 569 F.2d at 1127). In re Wright similarly noted:

We disagree with the [PTO]’s conclusion, reached by a comparison of the relative dimensions of appellant’s and [the] Bauer [references]’s drawing figures, that Bauer “clearly points to the use of a chime length of roughly 1/2 to 1 inch for a whiskey barrel.� This ignores the fact that Bauer does not disclose that his drawings are to scale. Absent any written description in the specification of quantitative values, arguments based on measurement of a drawing are of little value.

569 F.2d at 1127.