memorandum: opinion AND ORDER
WILLIAM T. HART, District Judge.Plaintiff Emanuel M. Glaros filed this action against defendants H.H. Robertson Company (“Robertson”) and Inryco, Inc. (“Inryco”), alleging that they had willfully infringed U.S. patent No. 3,535,844 issued to him on October 27, 1970 (the “Glaros patent”) covering a type of structural panel (the “Glaros panel”) used in the construction of steel frame buildings. The defendants have responded by denying the charge of infringement and asserting that the Glaros patent is invalid. During the pendency of this action before Judge Roszkowski, the defendants moved for summary judgment, arguing that the prior art anticipated the Glaros patent, 35 U.S.C. § 102, and, alternatively, rendered it obvious. 35 U.S.C. § 103. Judge Roszkowski denied the motion, as questions of fact existed regarding the scope and content of the prior art, which could only be resolved with the aid of expert testimony. Glaros v. H.H. Robertson Co., 79 C 1803, Slip at p. 5 (N.D.Ill. Sept. 17, 1981). After further discovery and the entry of a final pretrial order on April 22, 1982, the case was set for trial before Judge Roszkowski.
After transfer to this Court’s trial calender, the parties requested further discovery. Discovery reopened for a short period on September 14, 1983, pursuant to an agreed schedule. The parties supplemented the pretrial order and the case was returned to the trial calendar on December 23, 1982. On February 22, 1984, the defendants filed a renewed motion for summary judgment on the issue of obviousness, based on the deposition testimony of plaintiff’s expert, Dr. Larry D. Luttrell. *344That motion is the matter currently before the Court.
In light of the parties’ varying views on the application of the standards for summary judgment contained in Fed.R. Civ.P. 56 to the issue of obviousness, some preliminary discussion is in order. The United States Court of Appeals for the Federal Circuit has repeatedly accepted the propriety of granting summary judgment on the issue of obviousness. E.g., Peterson Mfg. Co. v. Central Purchasing, Inc., 740 F.2d 1541, 1546 (Fed.Cir.1984); Union Carbide Corp. v. American Can Co., 724 F.2d 1567, 1571 (Fed.Cir.1984). Whether an invention is or is not obvious in light of the prior art is a legal question to be determined by the Court. Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1535 (Fed.Cir.1983).
This legal determination is based upon the subsidiary factual inquiries identified in Graham v. John Deere Co., 383 U.S. 1, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966).
Under § 103, the scope and content of the prior art are to be determined; differences between the prior art and the claims at issue are to be ascertained; and the level of ordinary skill in the pertinent art resolved. Against this background, the obviousness or nonobviousness of the subject matter is determined. Such secondary considerations as commercial success, long felt but unsolved needs, failure of others, etc., might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness or nonobviousness, these inquiries may have relevancy.
383 U.S. at 17, 18, 86 S.Ct. at 693, 694. The Federal Circuit has repeatedly reaffirmed the necessity of pursuing these factual inquiries to provide a basis for determining obviousness. See, e.g., Jones v. Hardy, 727 F.2d 1524, 1528 (Fed.Cir.1984). The Federal Circuit has also repeatedly emphasized the importance of the inquiry into secondary considerations, such as the commercial success of the invention and the prior failure of others, as the strongest precaution against judging an invention from the perspective of 20/20 hindsight. See, e.g., Connell v. Sears, Roebuck & Co., 722 F.2d 1542, 1549 (Fed.Cir.1983); W.L. Gore & Associates, Ltd. v. Garlock, Inc., 721 F.2d 1540, 1553 and 1555 (Fed.Cir.1983).
Glaros has chosen to rely almost exclusively upon such secondary considerations in his opposition to the defendants’ motion. Since Glaros does not take issue with the defendant’s presentation of the remaining three predicate inquiries (with one important exception discussed below), the defendants’ evidence on these issues is uncontroverted.1 Glaros does present the opinion of Dr. Luttrell that the Glaros patent was not obvious in light of the then available prior art. But this conclusory opinion as to the legal issue to be decided sheds no light on the underlying factual issues, and so cannot create a material issue of fact precluding summary judgment. Peterson Mfg., 740 F.2d at 1547-48. See also Union Carbide, 724 F.2d at 1573. Therefore, the Court turns to these factual inquiries, allowing the plaintiff the benefit of all favorable inferences, as must be done on a motion for summary judgment. The uncontroverted evidence presented by the parties on these issues reveals the following.
THE GLAROS PATENT
The Glaros patent contains three claims relevant to this action. First, the Glaros patent claims a unitized structural panel composed of two sheets of some rigid material (normally sheet metal) or “skins,” forming “a box-like configuration” and filled with an insulating material, or “insulant,” which adheres to these exterior skins (see Figure 1). Both side edges of each *345panel must form a weathertight interlocking joint when combined with adjoining panels. To provide such an interlock, the Glaros panel utilizes a “double tongue and groove” joint in which each side edge contains both a “tongue” and a “groove” that engage the corresponding tongue and groove of the adjoining panel. The Glaros tongue is a “boxed tongue.” Rather than a tongue made solely of a single thickness of the sheet metal by making a single bend in the skin, this boxed tongue is made with a double bend, so that it consists of two thicknesses of the sheet metal skin reenforced by the insulant enclosed within the “box.” This claim, Claim No. 1, comprises the basic invention. By virtue of its unitized construction this panel allows for one step installation onto the supporting framework of a building. Its double tongue and groove interlocking joint forms a weather-tight joint between panels.
The Glaros patent lists as its second claim a panel with a particular version of this double tongue and groove interlock. In this version, the front and back skins are “offset” so that the tongue on each side edge protrudes beyond the sides of the groove on that same side (see Figures 2 and 3). The purpose of this modification is twofold. First, a fastener, such as a bolt or screw, may be inserted through the back skin and the elongated tongue that results from the offset. When interlocked with the corresponding groove of the panel that has been fastened to the building’s frame, the protruding tongue of the adjoining panel hides the fastener, providing a smooth front. Second, and perhaps more important, the offset eliminates any direct contact between the front and back sheets of the structural panel, improving the insulating properties of the panel. Since the front and back sheets of the panel are ordinarily made from some type of sheet metal, which freely conducts heat or cold, any direct contact allows heat or cold to pass more easily through a wall of panels.
The third relevant claim of the Glaros patent is a panel, as described in Claim No. 1 and filled with rigid foam insulant. Use of rigid foam insulant allows one step fabrication of the Glaros panel by applying liquid insulant to the metal skins. The liquid then expands to fill the interior of the panel.
THE PRIOR ART
The defendants rest their obviousness argument on two alternative panel designs, which became part of the prior art in 1936 and 1947, respectively. They suggest that either of these panels, when modified to account for subsequent advances in metal-forming technology and the invention of rigid foam insulants, rendered the Glaros patent obvious. The later of these enhancements in the art occurred in 1963; roughly five years prior to the date Glaros “invented” his panels.
The first panel design is disclosed in U.S. Patent 2,047,154 issued during 1936 to Carl J. Pimsner (alternatively the “Pimsner patent” or the “Pimsner panel”). Pimsner’s panel was designed to insulate devices such as industrial ovens, driers or refrigerators which maintain a high temperature differential between their interior and exterior.2 The Pimsner patent discloses a panel composed of two sheet metal skins joined by narrow metal cross members to form a box-like configuration filled with a loose insulant such as rock wool or fiberglass. The Pimsner panel side edges are formed in an offset double tongue and grove interlock configuration (see Figure 4). The tongue portions of the Pimsner panel interlock are not “boxed tongues” as are Glaros’, but rather tongues composed of a single thickness of sheet metals. The front and back skins of the Pimsner panel are also offset to provide both a space through *346which a fastener may be inserted and a gap to prevent direct heat exchange between the front and back skins. However, the offset is created by recessing rather than protruding the tongue of each side edge.
The second panel noted by the defendants is the Fenestra Type C Building Panel (the “Fenestra panel”). This unpatented design became part of the prior art when produced and marketed by the Detroit Steel Products Company in 1947. The Fenestra panel consists of two sheet metal skins formed into a box-like configuration by pressing the sheets together at the panel’s side edges (see Figures 7(A)-(C)). Although the panel is filled with loose insulant, as is the Pimsner panel, it does not require cross members to hold together the front and back skins because the skins themselves are joined at the side edges. The side edges each possess a tongue and groove, again forming a double tongue and groove interlock with the adjoining panels. Since the front and back skins of the Fenestra panel are pressed together to form a single side edge, the tongue is a boxed tongue. The front and back sheet metal skins are not offset, so that the tongue and groove are even. However, one method illustrated for installing Fenestra panels is by inserting a fastener through the back side of the groove (see Figure 7(D)). Apparently then, there is sufficient clearance without any offset to allow attachment of a fastener, although the fastener must pass through two thicknesses of sheet metal. Despite the lack of any offset, the Fenestra panel design addresses the problem of heat transfer between the front and back skins, by inserting a strip of asbestos felt between the two pieces of sheet metal before they are pressed together.
The Fenestra panel also embodies an advance in metalworking design that took place in approximately 1940. In that year the American Iron & Steel Institute published its light gauge cold-formed design manual (the “Steel Institute manual”). This manual allowed design engineers to ascertain for a given metal form, such as a boxed tongue, a numerical measure of its strength. Hence, without performing tests, a design engineer could compare the relative strengths of a single thickness sheet metal tongue of the sort embodied in Pimsner’s panel to the type of boxed tongue utilized in the Glaros panel. Further, the sheet metalforming art developed considerably between the time Pimsner’s patent issued and the time of Glaros’ invention. Dr. Luttrell questioned Pimsner’s ability to form a boxed tongue, but admitted that by the 1960’s, such configurations could easily 'be formed from sheet steel. The Fenestra panel is formed into such a boxed tongue configuration.
The last prior art reference advanced by the defendants is U.S. Patent No. 3,290,845 issued to Marvin K. Snyder on December 13, 1966 (the “Snyder patent” or “Snyder panel”). The Snyder patent also discloses a panel comprised of two sheet metal skins placed in a box-like configuration and filled with insulant. Snyder utilizes the same rigid foam insulant claimed by Glaros to form a unitized panel. The side edges of the Snyder panel do not resemble those of Glaros and Pimsner, and will not be discussed except to note that they further evince advances in the metalforming art (see Figure 6).
THE DIFFERENCES
(a) Pimsner
The Glaros panel differs from the Pimsner panel in several respects. First, the outer skins of the Glaros panel form a unitized panel with the rigid foam insulant rather than a hollow shell filled with loose insulating material. However, rigid foam had not been invented in 1936. Although Dr. Luttrell had reservations as to the ease of fabrication, he agreed that the loose insulant utilized by Pimsner could be replaced with a rigid foam insulant on any date after rigid foam filler had been disclosed by Snyder in 1963.
Second, the side edges of the Pimsner panel utilize a single thickness tongue rather than a boxed tongue. Again, in 1936 it is doubtful that Pimsner could have fabri*347cated boxed tongues had he desired to or been aware of their increased strength. However, by 1963 the ability to fabricate such boxed tongues had been developed and the Steel Institute’s manual described the increased strength of a boxed tongue in relation to a single thickness tongue.
Third, although the Pimsner panel utilized an “offset” as did the Glaros panel, the offsets are not the same. Pimsner’s offset provides a space for installing a hidden fastener and also reduces heat transfer between the panels’ skins, as does the Glaros offset. Yet filling the Pimsner panel with rigid foam rather than loose insulant poses difficulties for engaging the tongues and grooves in the Pimsner offset configuration. Dr. Luttrell explained that the Pimsner panel could be interlocked to an adjoining panel only when the nonboxed side of a groove could push past or through the loose insulant along its corresponding tongue to engage its corresponding tongue. Rigid foam insulant cannot be displaced as can loose insulant, so that it impedes any interlock.
In contrast to the Pimsner offset, the Glaros offset poses no such difficulty when filled with a rigid insulant. By protruding the tongue rather than the groove of the side edge, Glaros alleviates this problem. In the Pimsner design, that portion of the tongue over which the side of the groove must pass to engage the interlock is an interior surface edge of the panel. In the Glaros design, this edge has become an exterior surface, so that it does not contact the insulant filling the interior of the panel,
(b) Fenestra
The Fenestra panel utilizes loose insulant as does Pimsner, since rigid foam insulant was not yet available when it was designed. Due to the absence of any offset, filling the Fenestra panel with rigid foam poses no new difficulties as to the side edge interlock. The Fenestra tongues are boxed, as are Glaros’. The Fenestra panel also utilizes a hidden fastener. The Fenestra panel differs chiefly from the Glaros panel by virtue of the lack of any offset. As noted above, the Fenestra design deals with the heat transfer problem in a different manner.
THE LEVEL OF ORDINARY SKILL
Dr. Luttrell testified that during the early 1960’s the “vast majority of people in the industry, the light gauge cold-formed industry, ... probably were sort of half artists and half engineers.” See deposition transcript pp. 172-177. He described artists as those “with some understanding of sheet metal work, though not a great deal of engineering abilities to deal with it.” However, by 1963 structural panels had become “a fairly widely engineered product,” so that the ordinary “structural type engineers working on this problem” were graduate engineers with bachelor’s degrees and three or four years of experience. Although “relatively few engineers really understood the business of calculating these strengths [of various tongues and grooves],” and so might not know the full extent of various changes, one with ordinary skill would have appreciated the increase in strength achieved by “boxing” a tongue and groove interlock.
SECONDARY CONSIDERATIONS
To avoid creation of a material issue of fact precluding summary judgment, the defendants have not contested Glaros’ evidence regarding their failure to design a successful structural panel and the subsequent commercial success of their allegedly infringing panel. That evidence demonstrates the following. While defendant Inryeo began “exploring” foam insulated structural panels as early as 1962, Inryco seriously pursued development only after 1974. Since Glaros filed his patent application during 1968, evidence of any developmental failures or success by Inryco is irrelevant.
Beginning in 1961, defendant Robertson undertook a significant program to develop a unitized structural panel. Robertson engineers unsuccessfully pursued a number of alternative panel designs, until November of 1967, when they selected one for a pilot manufacturing program. Robertson *348engineers ultimately choose a panel whose side edge possessed a double tongue and groove interlock. This double tongue and groove consisted of two single thickness tongues with no offset to reduce heat transfer.
In February of 1968, Robertson personnel visited a trade show in Chicago where Glaros was exhibiting his panel design. Soon after, a copy of Glaros’ brochure found its way to Robertson’s development team. Robertson then scrapped its plans for pilot production of the previously selected panel. Instead, Robertson began to produce a panel whose side edge contained an offset boxed tongue and groove design (it is unclear whether this interlock is a single or double tongue and groove). In 1970 and again later during the 1970’s, Robertson modified the side edge configuration of its panels, moving closer to Glaros’ design. Each of these designs has enjoyed commercial success.
The Glaros panel enjoyed a lesser degree of commercial success than did Robertson’s panel. In 1970, annual sales of the Glaros panel exceeded one million dollars and were increasing. Wheeling-Pittsburgh Steel Company (W-P) marketed the panel jointly with Glaros Products, Inc. until 1972, when W-P acquired exclusive rights to the Glaros panel. W-P sold roughly four to five million dollars worth of Glaros’ panels from 1972 through 1974, when it stopped selling panels with Glaros’ design due to unspecified problems. Glaros reacquired the rights to his patent in August of 1976. Since then he has sold no panel embodying his design and had granted no licenses to others.
OBVIOUSNESS
Under the patent statute, ‘ideas’ are not patentable; claimed structures and methods are. Analysis properly begins with the claims, for they measure and define the invention. Further, each claim must be considered as defining a separate invention (citations omitted).
Jones v. Hardy, 727 F.2d 1524, 1528 (Fed.Cir.1984).
The defendants maintain that Glaros’ design is no more than an update of the Pimsner panel. The defendants note that the invention claimed by Glaros consists of: (i) two sheet metal skins unitized with rigid insulant filler; (ii) double tongue and groove side edge interlocks; (iii) with boxed tongues; (iv) an offset between the tongue and the groove to reduce heat transfer; and (v) allow installation of a hidden fastener. They contend that the Pimsner patent disclosed a panel with (i) double tongue and groove side edge interlocks with (ii) an offset to reduce heat transfer, (iii) which would allow installation of a hidden fastener. Take Pimsner’s design, box the tongues and substitute rigid foam for loose insulant, they argue, and you have the Glaros design.3
Glaros all but concedes that in 1963, substituting foam for loose insulation involved no inventiveness. After the Snyder patent issued in 1963, it was obvious to apply rigid foam to two sheet metal skins *349to form a unitized structural panel. Hence Claim No. 3 is invalid as obvious.
Boxing Pimsner’s single thickness tongues to increase the strength of the side edge interlock would also have been obvious during this period. Thus, Claim No. 1 of the Glaros patent is also invalid, for it discloses only the boxed double tongue and groove interlock on a unitized structural panel. As the Fenestra panel illustrated, boxed double tongue and groove interlocks were well known in the structural panel art long before the 1960’s. The Steel Institute manual disclosed widely both the existence of such a configuration and its relative utility, as opposed to that of a single thickness tongue and groove.
The result differs as to Claim No. 2 of the Glaros patent. That claim discloses all five of the elements set forth by the defendants, including the offset created by Glaros’ protruding tongue configuration. The defendants are mistaken when they assert that the panel described in Claim No. 2 of Glaros is simply Pimsner with boxed tongues and rigid foam insulant. It is more.
To establish obviousness, the prior art must contain more than merely the individual features of the claimed invention. Rather, the prior art must “suggest the desirability and thus the obviousness, of making the combination.” Lindemann Maschinenfabrik GmBH v. American Hoist & Derrick Co., 730 F.2d 1452, 1462 (Fed.Cir.1984). Dr. Luttrell stressed the problem created by filling the Pimsner panel with solid rather than loose insulant. Since rigid foam insulant impedes the Pimsner tongues and grooves from engaging, it is not clear that one with ordinary skill in the panel art would have attempted to enhance Pimsner’s design in this manner. Where a design “was contrary to the expectations and expectations of the art, the structure effectuating it would not have been obvious to those skilled in the art.” Schenck v. Nortron Corp., 713 F.2d 782, 785 (Fed.Cir.1983). In light of the difficulties in engaging Pimsner’s. offset if filled with rigid foam, combining an offset tongue and groove interlock with rigid foam insulant may have been counter-intuitive and thus nonobvious to those in the structural panel art.
Robertson’s own failure to develop the design they now suggest was obvious suggests that these enhancements to Pimsner’s design were nonobvious. Robertson’s failure cannot be given undue weight, for Robertson’s engineers may not have run across the Pimsner design. However, the standard for determining obviousness is objective, focusing upon the inventor, possessing ordinary skill in the art, “working in his shop with the prior art references — which he is presumed to know — hanging on the walls around him.” Union Carbide, 724 F.2d at 1576, quoting In re Winslow, 365 F.2d 1017, 1020 (CCPA 1966). Although Robertson’s failure is thus at best equivocal evidence of nonobviousness, on the defendants’ motion for summary judgment Glaros is entitled to the most favorable inference from Robertson’s failure.
The defendants argue that the problem of engaging Pimsner’s tongue and groove joints that arises when rigid foam insulant is substituted for Pimsner’s loose insulant is easily solved by either of two methods. Foam insulant that impeded engagement could be routed out during installation, or could be blocked out from the critical area during the initial “foaming” of the panel (see Figure 8). Either solution is obvious, they contend. Even if it would be obvious to modify a Pimsner panel as suggested and the block or route out excess foam insulant, the result would not be the invention claimed by Glaros. A claimed invention is nonobvious if it “achieved more than a combination which any of the prior art references suggested, expressly or by implication.”4 In re Sernaker, 702 F.2d 989, 994 (Fed.Cir.1983). By protruding the *350tongue rather than the groove of the panel’s side edge to form his offset, Glaros eliminated the need to block or route out unwanted foam insulant. The metal skin that forms one side of the Glaros tongue blocks any foam insulant from impeding the tongue and groove interlock. Accordingly, the Glaros panel is not the functional equivalent of a Pimsner panel with boxed tongues and foam insulant. Glaros’ decision to protrude the tongue rather than recess it, has another function in addition to those achieved by Pimsner’s offset.5 There is no evidence in this record that this change in the offset was obvious in light of the prior art. To the contrary, Dr. Luttrell testified that the problem of rigid foam impeding Pimsner’s interlock would require “cleverness,” “ingenuity,” “another tack,” all of which denote a level of inventiveness or nonobviousness. See deposition transcript pp. 145, 233, 242-43, and 249-50.
IT IS THEREFORE ORDERED that the defendants’ motion for summary judgment is granted in part and denied in part. Claims 1 and 3 of United States patent number 3,535,844 issued on October 27, 1970 to plaintiff Emanuel M. Glaros are declared to be invalid as obvious in light of the prior art. The validity of Claim No. 2 is reserved for trial.
*351
*352
. The defendants are to be complimented on a very thorough and lucid presentation of a complex and technical set of facts. Although lengthy briefs, as were submitted here, often become repetitive and unhelpful, the defendants have avoided these pitfalls with their clear and straightforward presentation.
. In his surreply, Glaros belatedly raises the issue of whether the Pimsner patent is a part of the relevant prior art. See Lindemann Maschinenfabrik GmBH v. American Hoist & Derrick Co., 730 F.2d 1452, 1460 (Fed.Cir.1984). However, Glaros has failed to direct the Court to any evidence that the Pimsner patent is not relevant to "the particular problem with which the inventor was involved.” Union Carbide, 724 F.2d at 1572. To the contrary, Dr. Luttrell discussed at length the application of Pimsner’s teaching to the problem faced by Glaros.
. Glaros argues that the analysis advanced by the defendants improperly focuses upon various elements of the claimed invention rather than the obviousness or nonobviousness of the invention as a whole. Glaros correctly defines the focus of the ultimate determination. Nonetheless, that determination cannot be made without some examination of individual differences between the prior art and the claimed invention. Such an inquiry is a proper aid for determining obviousness and is mandated by Graham v. John Deere, supra. Jones v. Hardy, 727 F.2d 1524, 1528 (Fed.Cir.1984); Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1537 (Fed.Cir. 1983).
In his surreply, Glaros also incorrectly attempts to characterize the defendants’ argument as an impermissible “obvious to try" theory. See, e.g., Jones v. Hardy, 727 F.2d at 1530. The prior art cited by defendants clearly provides a motivation for making these enhancements to Pimsner as well as identifying the possible modifications. For example, the Steel Institute's manual does not simply identify boxed tongues in the abstract, but also teaches that this configuration adds strength to a tongue and groove interlock. The Snyder Patent does not simply disclose the existence of rigid foam as an insulant, but also teaches the utility of rigid foam in forming such a unitized panel.
. This is not to say that some synergy is required. Although synergy is evidence of nonobviousness, the absence of some synergy is not dispositive. Kansas Jack, Inc. v. Kuhn, 719 F.2d 1144, 1150 (Fed.Cir.1983).
. The Fenestra panel provides no support for the defendants’ argument on this point. Although its side edge does allow sufficient space to insert a hidden fastener, there is no real offset as in the Pimsner and Glaros panels, as evidenced by the different approach to the heat transfer problem taken in the Fenestra panel. Nonetheless, the Fenestra panel does indicate that the sheet metal forming a boxed tongue can be utilized to confine insulant away from the interlock.