Flotec, Inc. v. Southern Research, Inc.

16 F. Supp. 2d 992, 1998 U.S. Dist. LEXIS 11742, 1998 WL 437416

• Court: District Court, S.D. Indiana
• Decided: June 9, 1998
• Docket: IP 97-1367-C H/G
• Status: Published

Opinion

ENTRY ON PLAINTIFF’S MOTION FOR PRELIMINARY INJUNCTION

HAMILTON, District Judge.

Plaintiff Flotee, Inc. manufactures and sells oxygen regulators for medical uses. In this action Flotee seeks a preliminary injunction against a competitor for alleged misappropriation of its trade secrets and for infringement of the trade dress of its regulators. The court held an evidentiary hearing on Flotee’s motion on March 18-19, 1998, and the parties submitted post-hearing briefs. The court now states its findings of fact and conclusions of law pursuant to Fed. R.Civ.P. 52 and 65. The findings are of course based on the record developed at this time and are subject to revision after a trial on the merits. As explained below, the court concludes that Flotee has not shown a likelihood of succeeding on the merits of its claims. The alleged trade secrets consist of information that is readily ascertainable by reverse engineering Flotec’s product, which has been on the market for several years. In addition, Flotee failed to take reasonable steps to protect the secrecy of the supposedly valuable information, and the weight of the evidence shows that defendants derived their product through a process of reverse engineering from Flotec’s product and a number of competing products rather than by copying Flotee’s technical drawings. Finally, Flotec’s trade dress is not inherently distinctive, and the similarities between Flo-tec’s and defendants’ products are not likely to cause confusion about the source of a product. Accordingly, the court denies Flo-tec’s motion for a preliminary injunction.

Findings of Fact

Plaintiff Flotee, Inc. is an Indiana corporation with its principal place of business in Indiana. Its principal business is the manufacture and sale of high pressure gas regulators that are used primarily for oxygen tanks for medical uses. Flotec’s regulators fit onto the valves of tanks containing oxygen under high pressure. The regulators reduce the pressure of compressed oxygen (at pressures as high as 3000 pounds per square inch) down to much lower pressure (about 50 pounds per square inch) so that the oxygen flow can be adjusted in a way that is safe and reliable for patients who use the oxygen to help them breathe. Flotee manufactured regulators for other companies from about 1983 to 1993. In 1993 Flotee chairman Gilbert Davidson and others in the company designed Flotec’s own version of a regulator and began selling it in the market in early 1994. Flotec’s basic regulator design has been commercially successful, having sold more than 100,000 units in less than four years. Flotee is one of about 30 competitors in the market for similar regulators. Flotee estimates its own market share at roughly 4 to 8 percent of the total. Since 1993, Flotee has continued to refine and improve its products, but the basic concept remains essentially intact. Flotec’s product is not patented. 1

Defendant Southern Research, Inc. (SRI) is a Delaware corporation with its principal place of business in Florida. After dealings with Flotee in late 1995 and early 1996 that are at the core of this dispute, SRI developed in 1996 and early 1997 a competing line of oxygen regulators that are now manufactured by defendant Inovo (a successor to SRI) and distributed by defendant Laing International Corporation. Laing International is a Florida corporation with its principal place of business in Florida. The record does not reflect the citizenship of Inovo.

SRI was a small manufacturing firm with sophisticated machining equipment and personnel skilled in specialized precision manufacturing. In late 1995, SRI was in financial trouble and was looking for new business. In November 1995 SRI vice president Stan Sheridan attended a trade show in Atlanta where he met Gilbert Davidson and others from Flotee. There was some discussion of the possibility of having SRI manufacture key components of Flotec’s oxygen regulators. During the discussions, Flotee chairman Gilbert Davidson learned that SRI’s chairman was LeNoir Zaiser. Although Davidson had known Zaiser for some thirty years, they were casual acquaintances who had seen one another only rarely in recent years.

After the meeting in Atlanta, SRI’s Zaiser, Sheridan, and Brooks Moore visited Flotec’s facility in Indianapolis in December 1995. Davidson and other Flotee personnel met with the SRI personnel and showed them the Flotee plant and product. Zaiser asked for the opportunity to submit quotations for manufacturing several key components. Davidson was doubtful because SRI’s facility in Florida was much farther away than other Flotee vendors, and he thought the logistics would make it difficult to use SRI as a major supplier. Nevertheless, he agreed to give SRI the opportunity to quote prices for components. During the tour, Flotee personnel showed the SRI personnel some of Flotec’s manufacturing techniques. Most important, Davidson gave Zaiser copies of several Flo-tee engineering drawings for the components. The drawings contain a host of dimensions and tolerances for the components. The drawings do not contain any explicit warnings that the information on them is confidential. They have a legend saying: “Property of Flotee All Rights Reserved.” Davidson did not tell Zaiser that Flotee considered the drawings confidential. Davidson also did not ask Zaiser to sign any confidentiality agreement. By way of explaining the omission, Davidson said that he did not want to offend Zaiser. At the hearing Davidson described this omission as an “egregious error.”

SRI took the drawings and product samples back to Florida. SRI personnel then undertook a substantial effort to gather information about prices and quantities for materials and some outside processing that SRI would need to provide a price to Flotee. See Ex. 147. Sometime during the next couple of months, SRI orally submitted a “ballpark” number to Flotee for three key components as a set. Flotee wanted written quotations for separate components, but SRI never provided written separate quotations. In any event, Flotee thought the quoted price for the set was too high.- Flotee never asked SRI to return its technical drawings at any point after the negotiations had stalled. Although there is some conflict and ambiguity in the evidence on this point, the court finds that the negotiations between Flotee and SRI were essentially dead by the end of February 1996. This finding is based primarily on the testimony of Jeff Wyant, a former vice president and general manager of SRI who left the company unhappily and who did not go out of his way to testify favorably to SRI. Wyant testified that the Flotee deal was dead by the end of February, and that in March 1996, Zaiser told Sheridan to buy as many regulators as he could on the market. Tr. 195-96. Zaiser also testified that he told Sheridan to buy as many regulators as he could, and that he did so after SRI had given up “serious hope” of getting a contract to produce parts for Flotee. Flotee has not shown that SRI was planning to build its own regulators while it was still in active negotiations with Flotee. 2

LeNoir Zaiser, the chairman of SRI and Inovo, is the key person involved in developing the SRI regulator. Although his academic credentials are no match for plaintiffs expert’s, Zaiser is an engineer with extensive practical and entrepreneurial experience in some of the most sophisticated and exacting portions of the defense industry. That experience includes, for example, manufacturing components for the guidance systems of Sidewinder air-to-air missiles and other Navy and Army missiles. Tr. 380-81. Zaiser did the engineering on those defense products. He has extensive experience with precision engineering and manufacturing. He also has substantial skill in reverse engineering and bringing products quickly to the stage of efficient production, as shown in part by some of the testimony taken under seal in this case.

After SRI’s Sheridan bought as many regulators as he could in March 1996, Zaiser took them all apart and, in his words, “went to work deciding who had the best idea, best mousetrap.” Tr. 394. Based on this reverse engineering of many regulators, not only Flo-tec’s, Zaiser made a‘number of judgments about design to come up with a good product. Zaiser developed his own design of the flow disk, which he described as “the heart of the unit.” Tr. 397. Zaiser correctly pointed out that many features of regulators are generic: all the regulators have to be nearly the same size and shape to fit on the standard valve on the standard pressure tank; all regulators have to reduce pressure down to about 50 pounds per square inch; and most regulators allow the user to select from about eleven different flow rates for the oxygen. Zaiser did the .principal engineering on the SRI product in the summer of 1996. Zaiser also concluded from his study of competing products that the manufacturing techniques used by Flotee and other competitors were less advanced than SRI’s capabilities, “at least one generation behind in manufacturing technology.” Tr. 400. Zaiser and SRI were able to develop a good, operable regulator very quickly, going from initial design work to manufacturing for production in a period of roughly six to eight months. Flotee argues that very speed shows that SRI must have used Flotec’s drawings, but the court does not draw the same inference. SRI’s reverse engineering process began with very detailed (and legitimately obtained) information about products that SRI knew worked well. The challenge for SRI was not to design a new product from scratch, with all the trial-and-error work that entails, but to understand existing designs, to make improvements from those designs, and then to prepare some sophisticated machine tools and train staff to do the manufacturing more efficiently than other competitors could.

SRI began manufacturing and selling its own line of regulators at the end of 1996 or beginning of 1997. These products first came to the attention of Flotee in early 1997, when an SRI advertisement appeared for a product called “Millenium.” The evidence before the court shows that SRI and its successor Inovo produce their regulators at costs considerably lower than Flotec’s. When Flotee still manufactured its principal components in-house, Flotee produced roughly 3000 regulators per month with a total of 30 to 40 workers doing manufacturing, assembly, and testing. Inovo uses different manufacturing equipment and produces about 5000 regulators per month with a staff of only 9 doing manufacturing, assembly, and testing. These differences in manufacturing processes allow Inovo to price its regulators below the price of Flotee regulators. The SRI/Inovo regulators have met with considerable commercial success.

The regulators in question here are made primarily from machined aluminum and brass. They are about 6 inches long and just under 1.5 inches in diameter. Each side offered evidence from engineers showing that they had closely examined the other side’s product and had taken very detailed and very precise measurements of that product. The sizes and shapes of the various parts are readily accessible to an engineer with a few hours to spare using some standard precision measuring devices, including micrometers, calipers, optical comparators, bore gauges, depth gauges, thread gauges, ring gauges, roll pins, gauge pins, and profilometers. With those tools, each side was able to measure the other’s product dimensions to the nearest one-thousandth of an inch (and finishes on highly polished surfaces to even finer dimensions).

The parties also presented evidence concerning the effects of anodizing aluminum on the dimensions of precision-machined aluminum parts. Anodizing is the process by which the surface of a piece of aluminum is oxidized to create a protective coating. As set forth in a treatise that both sides treat as authoritative, the anodizing process adds some thickness to a part. The added thickness is only one-third the total thickness of the anodized layer because the surface of the aluminum beneath the layer retreats as a result of the process. See Ex. 138. 3 If anodizing had a significant effect on dimensions, that effect would add marginally to the difficulty of reverse engineering from measurements of a finished, anodized product back to machining dimensions for unanodized parts because the person doing the reverse engineering would need to allow for the change in dimensions. In this case, however, Flotec’s drawings show that the anodized layer on Flotec’s product is about 0.0002 inches thick, meaning that the change in dimension on each surface is between 0.000060 and 0.000070 inches. In addition, anodizing has an even smaller effect on the inside dimensions of very small holes in a piece of aluminum because the anodizing process cannot force the anodizing material into those small holes, the precise dimensions of which are important in these oxygen regulators. The court finds that the effects of anodizing on the Flotee products at issue here, where the most precise key measurements are to a tolerance of 0.001 inches, do not make reverse engineering significantly more difficult.

When Flotee solicited price quotations from SRI, Flotee provided copies of its detailed engineering drawings of the relevant components. Ex. 7-17. Those drawings are not marked as being confidential but contain the legend: “Property of Flotee All Rights Reserved.” The drawings show several hundred dimensions. The drawings also identify some “critical” dimensions and tolerances and some “major” dimensions and tolerances. The legends on the drawings show that most dimensions expressed in thousandths of an inch (“.XXX”) have tolerances of plus or minus 0.005 inches. For several critical or major dimensions, however, the drawings show finer tolerances of plus or minus 0.001 or 0.002 inches.

Flotee has argued that it also provided confidential trade secrets to Zaiser and others during the visit to the Flotee facility in the form of oral statements about manufacturing “know-how.” Flotee has not identified any such specific information, however, so Flotee has not shown that any such oral “know-how” amounts to a trade secret.

Flotee has a number of measures in place to protect the confidentiality of some of its manufacturing and business information. Flotee keeps its technical drawings in a secure and locked area. Computers used for technical drawings have access limited by passwords. The Flotee facility is protected by a burglar alarm. All employees are required to sign agreements acknowledging the confidential nature of Flotec’s technical information and drawings and promising to protect that information. Since at least several years before 1995, Flotee has also used written confidentiality agreements with at least some suppliers who manufacture parts to its specifications. Flotee has also placed legends on some technical drawings labeling them as confidential. As noted, however, none of the drawings that Flotee provided to SRI were labeled confidential.

Flotee called several witnesses who are active in the machine shop industry. These witnesses testified that, in their experience, machine shops treat the technical drawings of customers and potential customers as confidential information and use that information only for the customers’ purposes. All of these witnesses, however, are also familiar with written confidentiality agreements, which are quite common between machine shops and their customers to protect the customers’ technical information, especially when the technical information concerns products that are still in development and not yet on the market. In fact, Flotee has often used written confidentiality agreements with potential suppliers and manufacturers.

For purposes of Flotec’s claim of trade dress infringement, the most critical evidence is the numerous samples of regulators manufactured by these parties and by several other competitors. To avoid repetition, the relevant facts are set forth below at page 35-1010 as part of the court’s analysis of the likelihood of success on the trade dress claim. Other factual matters are mentioned in the conclusions of law, and the substance of a finding of fact or a conclusion of law, rather than its label, should govern its treatment.

Conclusions of Law

The court has federal question jurisdiction over the subject matter of this case because the Lanham Act claim for infringement of trade dress arises under federal law. The court also has jurisdiction over the parties. The parties have briefed the case on the assumption that Indiana law governs the questions of trade secret law, and the court applies Indiana law to that claim, which falls within the court’s supplemental jurisdiction. Indiana has enacted the Uniform Trade Secret Act substantially as drafted by the National Conference of Commissioners on Uniform State Laws. Indiana courts treat decisions interpreting the uniform act in other jurisdictions as relevant in construing the Indiana act. See Amoco Production Co. v. Laird, 622 N.E.2d 912, 917-18 (Ind.1993).

Standard for a Preliminary Injunction

To obtain a preliminary injunction, the moving party must show a likelihood of success on the merits of its claims, as well as a substantial and immediate risk that it will suffer irreparable harm if injunctive relief is not granted. If the moving party can show these two things, the court must then weigh the harm to the non-moving party if an injunction is granted erroneously against the harm to the moving party if an injunction is denied erroneously. The court must also consider how its decision to grant or deny an injunction would affect the public interest, which generally refers to the interests of persons who are not parties to the lawsuit. See, e.g., Computer Care v. Service Systems Enterprises, Inc., 982 F.2d 1063, 1067 n. 2 (7th Cir.1992); Abbott Laboratories v. Mead Johnson & Co., 971 F.2d 6, 11-12 (7th Cir.1992). 4

In cases of trademark infringement, unfair competition, trade dress infringement, and trademark and/or trade dress dilution, courts generally presume that the harm from a wrong will be irreparable, so that a finding of likelihood of success on the merits can be enough to support (but not always require) issuance of a preliminary injunction. See, e.g., International Kennel Club of Chicago v. Mighty Star, 846 F.2d at 1092 (trademark). The same is generally true in trade secret cases. See, e.g., PepsiCo, Inc. v. Redmond, 54 F.3d 1262, 1267 & n. 3 (7th Cir.1995) (affirming preliminary injunction to prevent misappropriation of trade secrets); Ackerman v. Kimball Int’l, Inc., 634 N.E.2d 778, 785-86 (Ind.App.1994) (same), aff'd in relevant part, 652 N.E.2d 507, 510-11 (Ind.1995); see also Norlund v. Faust, 675 N.E.2d 1142, 1149-50 (Ind.App.1997) (affirming preliminary injunction because violation of non-competition covenant would cause irreparable harm). The court considers first plaintiff Flotec’s likelihood of success on the merits of its claims for misappropriation of trade secrets and for infringement of its trade dress.

I. Trade Secrets

Indiana law prohibits “misappropriation” of “trade secrets” and authorizes injunctive relief and damages as remedies. See Ind. Code §§ 24-2-3-3 & -4. The critical concepts in the trade secret act are contained in the statutory definitions. “Trade secret” is defined to mean:

information, including a formula, pattern, compilation, program, device, method, technique, or process, that:

(1) derives independent economic value, actual or potential, from not being generally known to, and not being readily ascertainable by proper means by, other persons who can obtain economic value from its disclosure or use; and

(2) is the subject of efforts that are reasonable under the circumstances to maintain its secrecy.

Ind.Code § 24-2-3-2. “Misappropriation” of a trade secret includes the disclosure or use of a trade secret by a person who used “improper means” to acquire knowledge of the trade secret. Id. The statute defines “improper means” to include “theft, bribery, misrepresentation, breach or inducement of a breach of a duty to maintain secrecy, or espionage through electronic or other means.” Id.

Flotee claims that the technical drawings of several regulator components that it provided to SRI contained product dimensions and tolerances that amount to trade secrets, that SRI had an implied duty to keep the drawings confidential and not to use the information for its own purposes, and that defendants misappropriated the trade secrets by using them to design and manufacture a ■competing regulator. Defendants’ arguments against Flotec’s trade secret claim can be summarized as follows: (1) any arguably valuable information disclosed to SRI had been disclosed by the product itself because the information was readily available to any interested engineer through the proper means of reverse engineering; (2) any information provided to SRI was not the subject of reasonable efforts to maintain its secrecy because Flotee disclosed the information to SRI and to other vendors without any explicit indication that Flotee considered the information confidential; and (3) SRI did not use any “improper means” to determine any information about Flotec’s products. In response, Flotee argues: (1) reverse engineering would not have disclosed tolerances for machining dimensions; (2) despite the absence of any explicit promise of confidentiality, other circumstances show that SRI should have known that the information on Flotee’s drawings should be kept confidential; and (3) the similarities between Flotec’s drawings and SRI’s drawings and between Flotec’s product and SRI’s product are so substantial that SRI obviously copied the confidential information provided.

These arguments must be evaluated against a backdrop of some important principles of trade secret law. First and foremost, trade secret law protects only in formation that is kept secret. The owner of the secret must take reasonable, though not extravagant, measures to protect its secrecy. See Rockwell Graphic Systems, Inc. v. DEV Industries, Inc., 925 F.2d 174, 178-80 (7th Cir.1991); Ackerman v. Kimball International, 634 N.E.2d at 784. The owner may disclose information in confidence to employees or others without losing the legal protection of the trade secret, but a disclosure outside a confidential relationship destroys the legal protection. Thomas v. Union Carbide Agricultural Products Co., 473 U.S. 568, 584, 105 S.Ct. 3325, 87 L.Ed.2d 409 (1985); Smith v. Snap-On Tools Corp., 833 F.2d 578, 580 (5th Cir.1987); Eli Lilly & Co. v. EPA, 615 F.Supp. 811, 820 (S.D.Ind.1985); Restatement (3d) of Unfair Competition § 39 comment f (1995). Second, trade secret law does not protect information that is publicly available, including information that can be discerned with reasonable effort by inspecting a product available for purchase on the market. See, e.g., Roboserve, Ltd. v. Tom’s Foods, Inc., 940 F.2d 1441, 1454-55 (11th Cir.1991) (sale of plaintiffs product destroyed any reasonable expectation of secrecy as to unpatented parts); Skoog v. McCray Refrigerator Co., 211 F.2d 254, 257 (7th Cir.1954) (public display of product in plaintiffs store amounted to full public disclosure so that information was not protected); Restatement (3d) of Unfair Competition § 43 (1995) (independent discovery and analysis of publicly available products or information are not improper means of acquiring information). Third, as a corollary to the second point, the process known as “reverse engineering,” in which a skilled person studies a product and figures out how to produce it, is permissible and even encouraged under trade secret law. The Supreme Court has explained that trade secret law “does not offer protection against discovery by fair and honest means, such as by independent invention, accidental disclosure, or by so-called reverse engineering, that is by starting with the known product and working backward to divine the process which aided in its development or manufacture.” Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470, 476, 94 S.Ct. 1879, 40 L.Ed.2d 315 (1974); accord, Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 155-58, 109 S.Ct. 971, 103 L.Ed.2d 118 (1989) (trade secret law does not protect information that can be derived by reverse engineering from products and information publicly available). Against this background, the court turns to the specific issues here.

1. Information Disclosed by Flotec’s Product: To qualify as a trade secret, information must derive independent economic value from not being generally known to, and not being readily ascertainable by proper means by, other persons who can obtain economic value from its disclosure and use. Ind.Code § 24r-2-3-2. Flotee has not shown that any of the information it provided to SRI satisfies this criterion. The regulator in question is a relatively simple mechanical device. It is not patented, and its functions can readily be understood by engineers familiar with gas flow and devices to regulate it. Most important, Flotec’s regulator was already on the market when Flotee gave its drawings to SRI. SRI’s Zaiser was able to measure the dimensions of the Flotee product from the finished product on the market in a relatively brief time. Similarly, Flotec’s Gilbert and Brian Davidson were able to perform complete measurements of the SRI product in a few hours. The fact that the relevant information was so readily available undercuts the possibility of trade secret protection. See Bonito Boats, 489 U.S. at 159-60, 109 S.Ct. 971 (state statute purporting to prohibit reverse engineering of publicly available but unpatented boat design was preempted by federal patent law); Amoco Production Co. v. Laird, 622 N.E.2d at 918 (the “first and foremost consideration is whether the ... information is readily accessible to a reasonably diligent competitor”), quoting Surgidev Corp. v. Eye Technology, Inc., 648 F.Supp. 661, 682 (D.Minn.1986), aff'd, 828 F.2d 452 (8th Cir.1987); see also Coenco, Inc. v. Coenco Sales, Inc., 940 F.2d 1176, 1179 (8th Cir.1991) (machine did not embody trade secrets because information was readily available from published patents, or from reverse engineering or “simple observation”); SI Handling Systems, Inc. v. Heisley, 753 F.2d 1244, 1255 (3d Cir.1985) (stating general principles and finding that key information was not readily available from reverse engineering); Fisher Stoves, Inc. v. All Nighter Stove Works, Inc., 626 F.2d 193, 196 (1st Cir.1980) (defendant was free to copy plaintiffs functional innovations that were not protected by patent or trademark; imitation is fact of business life); Skoog v. McCray Refrigerator, 211 F.2d at 257 (characteristics of product that had been displayed to public were not protectable as trade secrets); Restatement (3d) of Unfair Competition § 43 (1995).

Whether information is “readily ascertainable by proper means” is a matter of degree, of course. The Supreme Court of Indiana provided a useful illustration in Amoco Production Co. v. Laird, where the information in question was the locations of potential oil fields. Beginning with some information in the public domain, highly skilled employees of Amoco Production had invested nearly seven months in exploratory work, and the company then spent $150,000 to have an outside contractor conduct an airborne microwave radar survey of a 13,000 square mile area. The court explained:

While some tools leading to Amoco’s site discoveries were easily accessible within the public domain, such as U.S. Geological Survey data and the existence of microwave radar technology, we find that, taken together, the integration of pertinent site information and resultant productions as to potential oil reserves constitutes a unique compilation of information not previously known in the marketplace. We thus agree with the trial court’s conclusion that the information generated by Amoco, later appearing on Clendenning’s map, was not readily ascertainable.

622 N.E.2d at 920; accord, Rockwell Graphic Systems, Inc. v. DEV Industries, Inc., 91 F.3d 914, 917 n. 4 (7th Cir.1996) (undisputed evidence showed it would take eight years to reverse engineer the product at issue). In this case, by contrast, the skilled engineers for each party were able to take complete sets of dimensions from the other’s product within a few hours.

Flotee points out that it invested years of effort and a great deal of money to develop its design. If Flotee had still been on the brink of marketing its product when it gave its drawings to SRI, that investment might have supported an argument that its design was a trade secret because it was not yet “readily ascertainable by proper means.” After Flotee sold its product on the market, however, that design was made readily available for potential competitors to inspect, copy, and improve upon if they could. See Roboserve, Ltd., 940 F.2d at 1454-55 (after plaintiffs product was available for sale, trade secret law would not protect any un-patented part or combination of parts in the product); Fisher Stoves, Inc., 626 F.2d at 196 (competitors were free to imitate functional aspects of plaintiffs design; absent patent or trademark protection, loss of sales to competitors’ imitations was a fact of business life); cf. Sikes v. McGraw-Edison Co., 665 F.2d 731, 734 (5th Cir.1982) (defendant’s ability to reverse engineer a design not yet on the market was not a defense to trade secret claim; by copying secret design before plaintiff had sold product, defendant gained unfair competitive lead in the market). Reverse engineering requires more, of course, than merely measuring dimensions of a product, but those dimensions from the Flotee regulator and other competing products were “readily ascertainable” and enabled SRI to design and bring its product to market within a period of roughly six to eight months.

In response to these points, Flotee argues that measurement of finished products does not fully disclose the relevant information for two reasons. First, measurement of a finished product does not disclose information about the tolerances of machining dimensions, that is, the accuracy required for specific dimensions in the manufacturing process. Second, the aluminum components of the finished Flotee products are anodized, meaning that a thin layer of aluminum oxide coats the part and adds a small amount to its dimensions, so that a measurement of an anodized part will differ slightly from the machining dimension before the part is anodized. The evidence on these points does not support a finding that Flotec’s information could be protected as trade secrets.

(a) Tolerances and. Other Dimensions: Even where a product is on the market, courts have recognized that tolerances are not readily available by merely measuring the marketed product. See Smith v. BIC Corp., 869 F.2d 194, 200-01 (3d Cir.1989) (holding that district court erroneously denied protective order based in part on unre-butted affidavit that tolerances on design drawings could not be derived by reverse engineering); SI Handling Systems, Inc. v. Heisley, 753 F.2d at 1256 (recognizing tolerances as “trade secrets par excellence, because they cannot be obtained by even the most precise measurements”); Williams v. Curtiss-Wright Corp., 681 F.2d 161, 163-64 (3d Cir.1982) (tolerances for replacement parts could not be reverse engineered); National Rejectors, Inc. v. Trieman, 409 S.W.2d 1, 21-24 (Mo.1966) (describing detailed and conflicting testimony about value of tolerance information, but holding information was not protected because plaintiff had not kept it confidential); Schulenburg v. Signatrol, Inc., 33 Ill.2d 379, 212 N.E.2d 865, 867 (1965) (defendants, who were ex-employees of plaintiff, held liable for copying drawings, including tolerances and even mistakes on drawings to establish competing business); see also Rockwell Graphic Systems v. DEV Industries, 925 F.2d at 175-76 (discussing difference between detailed technical drawings with tolerance information used to manufacture parts and less detailed drawings used merely to assemble product). It should be noted, however, that tolerances for the plaintiffs product are most important when the defendant is trying to compete in an aftermarket or replacement parts market, so that the defendant’s products will actually need to fit together physically with the plaintiffs or others’ products. See, e.g., Williams v. Curtiss-Wright Corp., 681 F.2d at 164 (tolerances were critical for replacement parts for jet engines); Ecolaire Inc. v. Crissman, 542 F.Supp. 196, 203 & n. 54 (E.D.Pa.1982) (tolerances needed to ensure that defendant’s reverse-engineered part would be interchangeable with plaintiffs parts). In this case there is no issue of interchangeability. Flotec’s and defendants’ products are each complete and work independently, without any need for their parts to be interchangeable.

To determine tolerances accurately from measurements of finished products, it would be necessary to measure a statistically significant sample of finished products and to determine the range within which a particular dimension falls. Determining tolerances on Flotec’s regulators through such an extensive review of a large sample of finished regulators would require a substantial investment of time and expense. See Amoco Production Co. v. Laird, 622 N.E.2d at 919 (where duplication or acquisition of alleged trade secret information requires substantial investment of time, expense, or effort, such information may be found “not being readily ascertainable” to as to qualify for trade secret protection); see also Williams v. Curtiss-Wright Corp., 681 F.2d at 163-64 (testimony showed that tolerances could not be derived by measuring finished components of jet engines). The evidence shows that there are relatively standard tolerances for machining, however, and Flotee needed to identify specific tolerance information that was economically valuable to Flotee because the information was not readily ascertainable by its competitors.

The court asked Flotec’s expert witness, Dr. Allen T. MacDonald, to identify features disclosed on the Flotee drawings given to SRI that satisfied two criteria: first, the dimension or tolerance or other characteristic of the product was not disclosed by the final product and thus was not available by reverse engineering, and second, the particular dimension is critical to the functioning of the device. Tr. 278. The second criterion keeps the focus on features of the product that actually add value by being secret, as distinct from minor, rather generic details like the design of a control knob or the precise radius of a curved edge. See Tr. 226, 278 (SRI used same dimension as Flotee for detents for control knob, which would be difficult to reverse engineer, but that technology is well known). From among hundreds of dimensions and tolerances on the Flotee drawings given to SRI, Dr. MacDonald identified four features that he contends satisfy both criteria: (1) the tolerances for an internal diameter of 0.935 inches shown on the far right-center portion of Flotec’s drawing, Exhibit 7, and in zone C-3 of SRI’s drawing, Exhibit 41; (2) the “helieoil” that Flotee uses in the thread at the end of the yoke that is used to attach the regulator to a standard fitting on an oxygen bottle; (3) the hole that is 0.024 inches in diameter as shown in the top center portion of Exhibit 7 as: “# 73 (.024) DRILL” and in zone A-3 of SRI’s drawing, Exhibit 41; (4) a depth dimension of 0.178 inches with a tolerance of plus or minus 0.002 inches shown in zone A-3 of SRI’s drawing, Exhibit 41. 5

The court concludes that none of this information qualifies as protected trade secret information. First, the internal diameter of 0.935 inches with a tolerance of plus or minus 0.001 inches on Flotec’s drawing (Exhibit 7) does not correspond precisely with the SRI dimension on SRI’s drawing. The SRI drawing also shows an internal diameter of 0.935 inches, but with a tolerance of plus 0.002 inches and minus 0.000 inches, meaning that the nominal dimension for the SRI product would have to be 0.936 inches to correspond with the Flotee drawing. Where one one-thousandth of an inch is supposed to make a big difference, this difference in the dimensions undermines the view that this dimension with its precise tolerance is a Flotee trade secret. The difference supports the view that SRI made its own judgment about this dimension and its tolerances, and that its judgment was (slightly) different from Flo-tec’s judgment. Also, looking at Flotec’s device (e.gEx. 19a) and comparing it to the drawing, the internal diameter of 0.935 inches accommodates a piston with an O-ring, and the Flotee drawing requires the surface to be microfinished. It is readily apparent, even to a layperson familiar with the evidence here, that this particular internal diameter of the cylinder is likely to require a very fine tolerance, and the diameter is in fact marked as a “critical” dimension on Flotec’s drawings. There is no tolerance “stack-up” problem with that particular dimension making it more difficult to reverse engineer. The dimension is easy to measure, and the fact that the tolerance is relatively fine on what is obviously a critical dimension for operation of the device is not a trade secret.

The “helicoil” is a metal insert into a threaded hole. Flotee uses the insert to protect the relatively soft aluminum yoke of the regulator from the harder threaded stainless steel T-handle that holds the regulator on the oxygen tank. The size of the hole is easy to measure, however, and the correct size of the thread insert can be determined from a catalog. See Tr. 220-21. In addition, the helicoil insert is readily visible when the T-handle is unscrewed. Finally, the court does not see how this relatively minor and .generic part adds substantial value to the design of Flotec’s regulator. Flo-tec’s own comparison of competitors’ products shows that many use the same standard insert. See Ex. 33 (identifying products with same “Dome thread insert”).

The 0.024 inch hole, by contrast, is a critical dimension of both Flotec’s and SRI’s regulators. That small hole connects the high pressure end of the regulator with the low pressure end. The problem is that the evidence as a whole shows that the size of that hole is readily ascertainable from the finished product. Because the hole is so small, the anodizing process does not affect its dimensions enough to alter measurements in the thousandths of an inch. See Tr. 409. More important, the size 0.024 inches corresponds with the standard # 73 drill size. By way of comparison, a # 72 drill is .025 inches and a #74 drill is 0.0225 inches. Tr. 365. Using pin gauges, as both Flotee and SRI did in measuring one another’s devices, the critical dimension of 0.024 inches is relatively easy to measure. On Exhibit 28, Flotec’s Brian Davidson recorded measurements of the size of this hole on five Flotee regulators after the parts had been anodized. His measurements show a range of 0.0221 inches up to 0.0234 inches. He attributes the deviations from the machining dimension of 0.024 inches to the effects of anodizing around the edge of the hole. Tr. 319. Even allowing for some effects of anodizing in the range of a few ten-thousandths of an inch, however, it would not take a great leap to conclude that the hole was produced with a standard # 73 drill. Under Flotec’s view of the effects of anodizing, any knowledgeable engineer would know to allow for modest anodizing effects, and that allowance would exclude the #74 drill with a nominal machining dimension of 0.0225 inches. Thus, although the size of this hole is critical to the function of the regulators, it is readily ascertainable by proper means and cannot qualify as a trade secret. 6

The 0.178 inch dimension identified by Dr. MacDonald on SRI’s drawing (Exhibit 41) cannot qualify as a trade secret of Flotec’s because it has no counterpart in the Flotee drawings. Similarly, the longitudinal dimension of 1.237 inches plus or minus 0.002 inches identified by Dr. MacDonald on SRI’s drawing, see Tr. 280, also does not correspond to Flotec’s dimension. The comparable dimension on Flotec’s drawing is 1.280 inches, with a tolerance of plus or minus 0.002 inches. The longitudinal dimension arguably could involve a “stack-up” of tolerances (unlike the internal diameter of 0.935 inches discussed above), but the fact that SRI developed and used a significantly different measurement undermines a claim for trade secret misappropriation based on use of the Flotee dimension.

(b) Anodizing: The fact that Flotee anodizes the aluminum components of its finished products does not have a material effect on the trade secret issues here. Flotee argues that, because the anodizing process adds material to a machined piece of aluminum, reverse engineering from the finished product cannot reliably be used to determine machining dimensions. The question, however, is one of degree. The evidence shows that the anodizing process for Flotec’s components produces a layer about 0.0002 inches thick, and that the actual addition to any dimension is about one-third of that amount, or between 0.000060 and 0.000070 inches. That effect is not sufficient to interfere with the ability to reverse engineer the machining dimensions from the finished, anodized product. Even if the anodizing added the full 0.0002 inches, the effect would still be minimal and would not interfere significantly with reverse engineering. Accordingly, based on the evidence presented at the hearing, the court concludes that Flotee has not shown that the information contained on the drawings it gave to SRI qualifies as a trade secret. For that reason, Flotee is unlikely to succeed on the merits of its trade secret claim.

2. Reasonable Measures to Protect Secrecy: Even if any of the information that Flotee disclosed to SRI qualified as a trade secret, the weight of the evidence presented here shows that Flotec’s disclosure of that information to SRI was outside the scope of any confidential relationship, so that Flotec’s disclosure destroyed the secrecy of all the information. That disclosure and loss of secrecy provide an independent reason for concluding that Flotee is unlikely to prevail on the merits of its trade secret claim.

As noted above, Flotee takes some substantial measures to protect the confidentiality of some of its manufacturing and business information. Flotee keeps its technical drawings in a secure and locked area. Computers used for technical drawings have access limited by passwords. The Flotee facility is protected by a burglar alarm. All employees are required to sign agreements acknowledging the confidential nature of Flo-tec’s technical information and drawings and promising to protect that information. For at least several years before giving drawings to SRI in late 1995, Flotee had used written confidentiality agreements with at least some suppliers who manufactured parts to its specifications. Flotee has also placed legends on some technical drawings labeling them as confidential. See, e.g., Exhibit 105 (Flotee copyright application with key dimensions removed from drawings).

The basic problem for Flotee in this case is that it did not take even the most elementary steps to protect the technical drawings it provided to SRI. The drawings did not bear a legend that they were confidential. Flotee did not ever tell SRI that it considered the information confidential, let alone obtain a confidentiality agreement when it provided the information to SRI, and Flotee never sought assurances from SRI that it would keep the information confidential. There also is no evidence that SRI ever gave Flotee any indication that it believed the information was confidential or that it intended to keep the information confidential. Even after the negotiations broke down in early 1996, Flotee never asked SRI to return the drawings. These circumstances strongly indicate that Flotee failed to take reasonable steps to maintain the secrecy of the allegedly valuable information and thus weigh heavily against the claim that Flotee is entitled to trade secret protection for any of the information it disclosed to SRI. See Titcomb v. Norton Co., 208 F.Supp. 9, 18-19 (D.Conn.1959) (plaintiffs product information was no longer secret after it consented to providing samples for trial and use in actual manufacturing operations, without restriction as to disclosure of the product dimensions, easily obtainable by any draftsman from the products themselves), aff'd, 307 F.2d 253 (2d Cir.1962); Van Products Co. v. General Welding & Fabricating Co., 419 Pa. 248, 213 A.2d 769, 778-79 (1965) (manufacturer’s know-how was not trade secret protected from use by former employee where product had been sold on the open market and its features had been the subject of advertising material and extensive coverage in trade press); cf. Trandes Corp. v. Guy F. Atkinson Co., 996 F.2d 655, 664 (4th Cir.1993) (jury could reasonably find that secrecy was not lost where plaintiffs only disclosures of secret object-code version of software were made in confidence under licensing agreements or in licensing negotiations); Laser Industries, Ltd. v. Eder Instrument Co., 573 F.Supp. 987, 991 (N.D.Ill.1983) (denying summary judgment for defendant on issue of secrecy where, among other things, the drawings that plaintiff disclosed to defendant, a potential outside manufacturer, were marked confidential); see also Atlantic Wool Combing Co. v. Norfolk Mills, Inc., 357 F.2d 866, 870 (1st Cir.1966) (finding misappropriation of trade secrets where plaintiff had developed new design for equipment and then obtained explicit agreement from machine shop that manufactured equipment to keep designs confidential and not to manufacture equipment for other competitors, and where another competitor then induced machine shop owner to breach promise of confidentiality); Booth v. Stutz Motor Car Co. of America, 56 F.2d 962, 964-66 (7th Cir.1932) (applying Indiana common law, disclosure of proposed but unmarketed automobile designs and blueprints to manufacturer was made with explicit demands for and assurances of confidentiality, as shown by correspondence of parties); Affiliated Hospital Products, Inc. v. Baldwin, 57 Ill.App.3d 800, 15 Ill.Dec. 528, 373 N.E.2d 1000, 1006 (1978) (plaintiffs drawings were not marked confidential, but defendant had signed agreement prohibiting disclosure of information gained during visit to plaintiffs plant, putting him on notice that plaintiff considered its manufacturing process secret).

To cope with this problem in its trade secret claim, Flotee argues that the court should find that, when SRI accepted Flotec’s technical drawings so that it could prepare a bid to manufacture components for Flotee, SRI had an implied duty to maintain Flotec’s information in confidence and not to use that information for its own purposes. To support this theory of an implied duty of confidence, Flotee relies on cases that have recognized implied duties of confidentiality in some business negotiations. Flotee has also tried to prove that there is a custom in the machine shop industry under which a machine shop is expected to and obligated to protect the confidentiality of any information turned over to it by a customer or prospective customer.

The authors of the Restatement of Unfair Competition have framed the relevant principles so as to impose a duty to protect the confidentiality of a trade secret if:

(a) the person made an express promise of confidentiality prior to the disclosure of the trade secret; or

(b) the trade secret was disclosed to the person under circumstances in which the relationship between the parties to the disclosure or the other facts surrounding the disclosure justify the conclusions that, at the time of the disclosure.

(1) the person knew or had reason to know that the disclosure was intended to be in confidence, and

(2) the other party to the disclosure was reasonable in inferring that the person consented to an obligation of confidentiality.

Restatement (3d) of Unfair Competition § 41 (1995); accord, Smith v. Dravo Corp., 203 F.2d 369, 376 (7th Cir.1953) (applying Pennsylvania law, finding that disclosure of confidential details about recently marketed product was protected under implied duty of confidence where plaintiff made disclosure to enable defendant to appraise value of business for possible purchase); Hoeltke v. C.M. Kemp Mfg. Co., 80 F.2d 912, 923 (4th Cir.1935) (finding that obligation of confidentiality was implied in negotiations over possible patent license for invention still in pre-mar-keting development stage, despite absence of explicit promise of confidentiality). 7

Flotee relies on Phillips v. Frey, 20 F.3d 623 (5th Cir.1994), in which the Fifth Circuit held under Texas law that a disclosure of a secret manufacturing process occurred within a “confidential relationship” when a manufacturer made the disclosure during negotiations for the sale of its business. There was no trade secret in the design of the product at issue, and the product could be reverse engineered. Id. at 629. There was no explicit promise of confidentiality, and the Fifth Circuit explained that the owner of a trade secret “may not unilaterally create a confidential relationship without the knowledge or consent of the party to whom he discloses the secret.” Id. at 632, citing Smith v. Snap-On Tools, Inc., 833 F.2d 578, 579-80 (5th Cir.1988). In the absence of any explicit promise of confidentiality, the issue in Phillips, as it is here, is “whether the recipient of the information knew or should have known that the information was a trade secret and the disclosure was made in confidence.” Phillips, 20 F.3d at 632, citing Snap-On Tools, 833 F.2d at 580, and Restatement of Torts § 757 comment j (1939). In Phillips the Fifth Circuit reviewed only for plain error because the jury rendered a verdict in favor of the plaintiff and the defendant had failed to move for judgment as a matter of law. 20 F.3d at 627. The Fifth Circuit concluded that the jury could have reasonably concluded that the disclosure in that case was made in a confidential relationship, particularly in view of evidence tending to show the defendant had not been sincere in its interest in buying the business and had used the negotiations merely as a guise for obtaining the secret manufacturing process. Id. at 632.

Phillips shows that an explicit promise of confidentiality is not always necessary if the recipient of the information knew or should have known that the information was a trade secret and that the owner of the secret expected the recipient to keep the information secret. That principle is included in § 41(b) of the Restatement (3d) of Unfair Competition. There is no evidence here that SRI actually knew that Flotee considered any information provided to it to be a trade secret. Thus, the question under § 41 in this case is whether SRI should have known that Flotee considered the information to be a trade secret. That question must be answered based on all the circumstances as they appeared to SRI when the information was disclosed. See Restatement (3d) of Unfair Competition § 41(b); see also Rockwell Graphic Systems v. DEV Industries, 925 F.2d at 179-80 (only in extreme ease can court decide on motion for summary judgment whether plaintiff took “reasonable precautions” to protect alleged trade secrets).

What were the circumstances as they appeared to SRI? SRI knew that Flotee had not ever indicated in any way — on the drawings themselves, or in discussions between the parties — that it considered any information to be secret. SRI knew that the Flotee components shown in the drawings had been on the market for a couple of years and were readily susceptible to reverse engineering by any skilled company interested in trying to compete. SRI also knew that Flotee had provided the drawings for a limited purpose — to enable SRI to provide a price quotation. That is the fact most helpful to Flo-tec’s position on this issue. See, e.g., Smith v. Dravo Corp., 203 F.2d at 376 (applying Pennsylvania law), but see Van Products Co. v. General Welding & Fabricating Co., 213 A.2d at 779-80 (disagreeing with Smith as statement of Pennsylvania law); Den-Tal-Ez, Inc. v. Siemens Capital Corp., 389 Pa.Super. 219, 566 A.2d 1214, 1226-30 (1989) (reviewing Pennsylvania courts’ rejection of the Smith analysis; shifting from conduct of defendant in confidential relationship with plaintiff “to a close analysis of whether the information was truly a trade secret”). SRI could also reasonably assume, however, that Flotee knew how to ask that information be considered confidential if it really thought the company’s crown jewels were at risk. And even under Flotec’s theory and evidence, only a few dimensions could be protected, and the vast majority of the dimensions on the Flotee drawings would not be trade secrets.

The evidence showed a widespread but not uniform practice in the machine shop industry of keeping a customer’s information confidential. The evidence does not support, however, a uniform custom of keeping such information confidential when it concerns components of a product already on the market and where the customer does not ask for any promise of confidentiality. The witnesses who described the supposed custom all had experience with explicit confidentiality agreements in the industry. See also Rockwell Graphic Systems v. DEV Industries, 925 F.2d at 177 (denying summary judgment for defendant on secrecy issue where, among other things, machine shops signed confidentiality agreements to receive plaintiffs detailed technical drawings needed to manufacture parts). The witnesses who testified that they keep customer information confidential all viewed that practice as beneficial to their businesses. The specific problem here, however, is whether it was reasonable for Flotee to rely on this asserted custom as the sole means to protect the confidentiality if information it claims is secret and vital to its business success. Where virtually all of the valuable information is disclosed by the finished product itself (so that an engineer could measure its dimensions carefully, as occurred in this case), and where that product is already on the market, that expectation of confidentiality was not reasonable. The difference between information about products that have reached the market and those that have not was recognized in the eases cited above. That difference was also suggested in the testimony of James Young, whose business now manufactures most of the major components of Flo-tec’s regulators. Flotee insisted that he sign a confidentiality agreement, but Mr. Young understood that the agreement was intended primarily to protect the information about some experimental work and new product ideas. Tr. 29, 33-34. To the extent that Flotee was relying on a supposed custom of the machine shop industry to protect the confidentiality of the drawings it gave to SRI, Flotee did not take reasonable steps to protect the confidentiality of the information on those drawings.

The second question under § 41(b) of the Restatement (3d) is whether Flotee reasonably inferred that SRI had consented to an obligation of confidentiality. There simply is no evidence here that SRI gave any indication that it agreed to keep Flotec’s information confidential. Nor does the evidence show that Flotee could reasonably have inferred such consent from SRI’s silence on the subject.

Viewing all these .circumstances together, the court does not believe SRI should have figured out that Flotee thought that a few nuggets of information on the drawings amounted to trade secrets. There was neither any explicit nor any implied duty of confidentiality here. Because Flotee made its disclosure outside a confidential relationship, it lost any trade secret rights it might have had in that information and is unlikely to succeed on its trade secret claim.

3. Proper or Improper Means: Flo-tee provided its supposedly confidential drawings to SRI without any warnings, claims, or agreements about confidentiality. Even if Flotee had insisted on a confidentiality agreement, however, or if there were other evidence that Flotee had taken reasonable steps to protect its secrecy, the court still would not be persuaded that SRI used any improper means in developing its regulator. Both Zaiser and Wyant testified about the process of reverse engineering that SRI used to develop its own regulator. Wyant obviously has no great affection for SRI. His testimony on SRI’s reverse engineering efforts is highly credible and corroborates Zaiser’s testimony. As noted above, “reverse engineering” is entirely lawful and legitimate. Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. at 155, 161, 109 S.Ct. 971; Kewanee Oil, 416 U.S. at 476, 94 S.Ct. 1879; Rockwell Graphic Systems v. DEV Industries, 925 F.2d at 178; American Can Co. v. Mansukhani, 742 F.2d 314, 329, 334 & n. 24 (7th Cir.1984).

Flotee identified 12 to 15 dimensions on the SRI drawing and product that it says are identical to dimensions on Flotec’s own drawing and product. But that is 12 to 15 dimensions out of several hundred. In addition, Flotee did not provide SRI with a complete set of drawings for its regulators, but only drawings for several major components. The court is persuaded that SRI developed its product primarily through the process of reverse engineering from Flotec’s product and other competitors’ products, and then added its own modifications and improvements to come up with its own design. It is possible that SRI referred to Flotec’s drawings at some point in that process, but if it did so, any use of that information was minor and certainly did not involve information that might be deemed so critical as to qualify as trade secrets (even if Flotee had taken reasonable measures to protect the information). Both Flotec’s Brian Davidson and SRI’s Le-Noir Zaiser identified a number of features that are different between Flotec’s and SRI’s products. See Tr. 291-303 (B. Davidson, describing advantages he sees in Flotee product over defendants’ product); Tr. 402-05 (Zaiser identifying features unique to Flotee product that he did not use in SRI product). Although completely mechanical copying, absent any improvements, is relatively unusual even where trade secret rights are being violated, see, e.g., American Can Co. v. Mansukhani, 742 F.2d at 329, the evidence of significant differences between the competing products here tends to corroborate Zaiser’s and Wyant’s testimony about SRI’s reverse engineering efforts and its exercise of independent engineering judgments in developing its new product.

Flotec’s witnesses viewed the speed with which SRI developed its competing product as evidence that it had misappropriated trade secrets. Dr. MacDonald was especially skeptical about SRI’s ability to bring the product to market so quickly. Flotec’s evidence contrasted SRI’s efforts of a few months with Flotec’s own substantial investment in development costs. That difference does not support Flotee’s case. SRI was starting from products on the market that it knew were operable and commercially successful. A person using reverse engineering does not need to reinvent the wheel and follow every blind alley and detour on the development path of the person who first developed the product. The fact that Flotee worked longer and harder to develop its product than SRI worked to develop its product does not mean that Flotec’s information is protected. See National Rejectors v. Trieman, 409 S.W.2d at 19-22. Although Dr. MacDonald provided valuable testimony in this case, the court is not persuaded by his opinion that SRI did not produce evidence of a sufficiently extensive design effort on its part. See, e.g., Tr. 233-35, 248. Dr. MacDonald is a very knowledgeable engineer and an experienced witness, but he does not have experience in reverse engineering a product to bring it to market, as SRI’s Zaiser does. Dr. MacDonald’s view on how long it should have taken SRI to do its reverse engineering and how many drawings it should have accumulated has, in the words of the Third Circuit, “the smell of the lamp about it.” See SI Handling Systems, Inc. v. Heisley, 753 F.2d at 1263 (affirming district court’s rejection of expert testimony that defendants could have reverse-engineered the plaintiffs system and could have done so relatively quickly, especially where no comparable product was on the market); Boeing Co. v. Sierracin Corp., 108 Wash.2d 38, 738 P.2d 665, 675 (1987) (affirming trial court’s rejection of expert’s testimony in part because expert had no experience with necessary reverse engineering technique). Zaiser is an experienced engineer and entrepreneur who was operating under intense financial pressure to bring one of two reverse-engineered products to market as quickly as he could. He accomplished the task quickly. That speed does not show that SRI misappropriated trade secrets from Flotee. The speed with which Zaiser accomplished the task is consistent with the evidence showing that he worked from the designs of several successful regulators and added some changes and improvements.

In summary, Flotee has not shown that it is likely to succeed on the merits of its claim for misappropriation of trade secrets. It has not shown the existence of truly valuable information that is not disclosed by the product itself; it has not shown that it exercised reasonable measures to protect the alleged trade secrets; and it has not shown that SRI used improper means to obtain information about Flotec’s product.

II. Trade Dress

Flotee has a separate claim for trade dress infringement under § 43(a)(1) of the Lanham Act, 15 U.S.C. § 1125(a)(1), which prohibits the use of a “word, term, name, symbol, or device or any combination thereof, or any false designation of origin ... which is likely to cause confusion, or to cause mistake, or to deceive” as to the affiliation of the user with another person or as to the origin of the goods. Trade dress also can be protected under § 43(a). The term “trade dress” refers to the total image of a product. It includes features such as size, shape, color or color combinations, texture, graphics, or even particular sales techniques. Computer Care, 982 F.2d at 1067; Roulo v. Russ Berrie & Co., 886 F.2d 931, 935 (7th Cir.1989). A party may obtain relief for trade dress infringement if it can prove (1) its trade dress either has acquired secondary meaning or is inherently distinctive, and (2) that the similarity of the defendant’s trade dress causes a likelihood of consumer confusion about the source or affiliation of the products. Dorr-Oliver, Inc. v. Fluid-Quip, Inc., 94 F.3d 376, 380 (7th Cir.1996). To be inherently distinctive, the trade dress must be “sufficiently distinctive to allow consumers to identify the product from the trade dress.” Computer Care, 982 F.2d at 1069, quoting AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531, 1536 (11th Cir.1986).

In evaluating the likelihood of confusion, the Seventh Circuit uses the same factors for trade dress infringement that it does for trademark infringement. See Dorr-Oliver, 94 F.3d at 380-81. These are the similarity of the trade dresses; the similarity of products; the area and manner of concurrent use; the degree of consumer care; the strength of the plaintiffs trade dress; actual confusion; and a defendant’s intent to palm off its products as plaintiffs. No single factor is dispositive, and the weight of particular factors will vary from case to case. Id. at 381; Schwinn Bicycle Co. v. Ross Bicycles, Inc., 870 F.2d 1176, 1187-88 (7th Cir.1989); International Kennel Club of Chicago v. Mighty Star, 846 F.2d at 1087.

In this case, the analysis of the distinctiveness of Flotee’s trade dress and the likelihood of confusion overlap to a considerable extent. The most important factors on likelihood of confusion here are the degree of similarity of the trade dresses and the strength or weakness of plaintiff Flotec’s trade dress. When Flotec’s product (or a particular configuration of its product) is put side-by-side with SRI’s product, the products do indeed appear very similar. The problem for Flotee is that the relevant comparison is in the market as a whole. SRI has come forward with samples of several other competitors’ regulators. See Exs. 140—144 and 150—154. These other regulators also bear striking resemblances to Flotec’s regulator. See Mana Products, Inc. v. Columbia Cosmetics Mfg., Inc., 65 F.3d 1063, 1069-71 (2d Cir.1995) (black compact packages not distinctive because many other cosmetic sellers used similar packages).

Flotee argues that the following specific features of its regulators are non-functional and are components of its “distinctive” trade dress: (1) the top radius or “dome” where the T-handle screws in to hold the regulator on an oxygen tank; (2) the knob style, shape, and color; (3) the machining radii around the tank window; (4) a single-barb hose fitting colored green; (5) the size, shape, location, and edge contours of the window to view the flow setting; (6) the silver T-handle; and (7) the overall green color.

Taking these features in reverse order, there is nothing remotely distinctive about Flotec’s use of the color green for its oxygen regulators. The color green is widely associated with oxygen, and many competitors col- or their regulators green. No competitor is entitled to exclusive use of the color green, or even a particular shade of green, in this market. Moreover, Flotee itself produces regulators in at least eight other colors, so even Flotee does not use color to indicate source to its customers. Flotec’s T-handle also is not distinctive. Many competitors use silver T-handles. Every product sample admitted into evidence has a silver T-handle, and nearly all the competing products described by Flotee in Exhibit 33 use the same size T-handle. Similarly, the flowmeter windows are not at all distinctive among the many competitors, and neither are single-barb hose fittings, even single-barb hose fittings colored green.

The knob shape and color are essentially identical as between the Flotee samples admitted into evidence and SRI’s product. But nearly all competitors use black knobs, see Ex. 33, and Flotee itself uses several different shapes of black knobs to give customers a choice. Only Flotee and SRI appear to use the same shape of knob with a slight recess on the outer surface. That similarity is trivial in the overall picture, and that detail of one particular style of knob does not identify a regulator as a Flotee regulator.

The dome shape and machining radii for the end of the Flotee regulator that fits over the oxygen tank are also not distinctive, as a comparison of the competing products shows. See Exs. 140, 141,143,144,150,151, and Ex. 33 (noting nearly identical dome features on Allied Spectrum, Ambu, Penox, Precision, Spiracle, and Veriflo Sierra). In addition, as Gilbert Davidson testified, Flotee designed that shape to be functional, to be easier on the more fragile hands of elderly patients who are the principal users of many Flotee products. See Qualitex Co. v. Jacobson Products Co., 514 U.S. 159, 165, 115 S.Ct. 1300, 131 L.Ed.2d 248 (1995) (in general terms, “ ‘a product feature is functional,’ ... ‘if it is essential to the use or purpose of the article or if it affects the cost or quality of the article,’ that is, if exclusive use of the feature would put competitors at a significant non-reputation-related disadvantage”), quoting Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 850, n. 10, 102 S.Ct. 2182, 72 L.Ed.2d 606 (1982); Schwinn Bicycle Co. v. Ross Bicycles, 870 F.2d at 1189 (“A feature is functional if it is one that is costly to design around or do without, rather than one that is costly to have.”). For these reasons, Flotec’s trade dress is not inherently distinctive, and these factors also weigh against a finding of likelihood of confusion, as discussed below.

Turning to the additional factors relevant in determining likelihood of consumer confusion, the products in this case are essentially identical in function. They compete directly with one another. The area and manner of concurrent use are also essentially identical. With respect to consumers’ degree of care, both Flotee and SRI distribute their products through sales representatives and advertisements. The products are relatively high value (selling for approximately $40 to $60 each), and they are obviously so important that they are purchased with some degree of care. Oxygen regulators are not impulse purchases. Cf. AHP Subsidiary Holding Co. v. Stuart Hale Co., 1 F.3d 611, 616-17 (7th Cir.1993) (expressing skepticism toward district court finding that consumers exercise high degree of care in purchasing cans of non-stick cooking spray in grocery store). The market is competitive and competitors try to establish name recognition of their products.

Flotee has come forward with evidence of a handful of instances in which its distributors wondered whether Flotee was manufacturing defendants’ product and allowing defendants to put their own labels on the product. This evidence was about distributors trying to make sure that Flotee was not invading their exclusive territories. This evidence actually tends to support defendants’ position, for these instances show that the distributors saw the different brand names on the products and were trying to determine whether they in fact came from two different sources, as the different names tended to show. These few instances do not show that Flo-tec’s trade dress is so distinctive as to warrant protection or that the similarities in appearance between Flotec’s and defendants’ products are likely to cause substantial consumer confusion about the origin of defendants’ products.

Flotee has argued that this court should disregard the appearance of other competitors’ products on the theory that SRI is improperly trying to use a “jus tertii ” defense. On this point, Flotec cites Marshak v. Sheppard, 666 F.Supp. 590, 599 (S.D.N.Y.1987) (trademark case), and Visa International Service Ass’n v. Bankcard Holders of America, 211 U.S.P.Q. 28, 40, 1981 WL 40539 (N.D.Cal.1981) (trademark and unfair competition). Those eases indicate that a defendant cannot defend a trademark infringement claim on the theory that the plaintiffs rights to the mark may be junior to the rights of some third person. Flotec’s argument is off target, at least where the issue concerns the distinctiveness of features of trade dress. Limitations on a jus tertii defense do not prevent courts from considering relevant evidence as to whether plaintiffs trade dress is distinctive or not. The other competitors’ products are relevant here because they show the extent to which the features shared by the Flotee and SRI products are distinctive or not. In fact, Flotee itself acknowledged the relevance of such comparisons when it prepared and offered Exhibit 33, which is its own comparison of the appearances of numerous competing products, including those that do not even use the basic “unibody” design shared by Flotee, SRI, and a number of other products. See Mana Products v. Columbia Cosmetics, 65 F.3d at 1069-70 (plaintiffs use of color black for cosmetic cases was not distinctive because many competitors used black).

Although Flotec’s trade dress is not inherently distinctive, Flotee also argues that its trade dress has acquired secondary meaning, that is, that there is “a mental association in buyers’ minds between the alleged mark [or trade dress] and a single source of the product.” Echo Travel, Inc. v. Travel Associates, Inc., 870 F.2d 1264, 1266 (7th Cir.1989), quoting 1 McCarthy, Trademarks & Unfair Competition § 15:1 at 657 (2d ed.1984). To determine whether a trademark has acquired secondary meaning, courts consider direct evidence, such as direct consumer testimony and consumer surveys. Courts also consider circumstantial evidence, such as the length and manner of exclusive use, the amount and manner of the owner’s advertising, the amount of sales and number of customers, the product’s established place in the market, and evidence of intentional copying. See Spraying Systems Co. v. Delavan, Inc., 975 F.2d 387, 393 (7th Cir.1992); Echo Travel, Inc., 870 F.2d at 1267. Flotee has not come forward with any consumer survey tending to show confusion about source. Flotee has offered evidence of a handful of incidents in which its distributors thought that Flotee was manufacturing defendants’ product and allowing defendants to put their own labels on it. As discussed above, however, that evidence is minimal and tends to show that the distributors recognized the different brand names and just wanted to make sure that the regulators with defendants’ brand names in fact came from a source other than Flotee. Flotec’s product has been on the market only since early 1994, so any period of “exclusive” use of its trade dress has been minimal at best. Flotee also does some advertising, but its advertising does not emphasize its overall trade dress or any particular feature, and Flotee has not identified any features of its trade dress that are so distinctive as to enable consumers to identify the source of the product. The Flo-tee regulator is solidly established in the market, but with a market share of no more than 8 percent. The evidence of intentional copying here is minimal in view of the many similar generic, undistinetive, and functional characteristics of the competing regulators. Flotee has not shown a likelihood of prevailing on its trade dress claim based on any secondary meaning attached to its trade dress.

III. Balance of Harms

For the reasons explained above, plaintiff Flotee has not shown that it has a reasonable likelihood of prevailing on the merits of its trade secret or its Lanham Act claims. Because of the nature of the wrongs alleged, the court assumes for purposes of argument that harm to plaintiff would be irreparable to the extent that plaintiff could prove liability under either theory. That harm would include lost sales and some possible harm to goodwill. 8 At the same time, the court must note that a preliminary injunction would also cause substantial harm to defendants. The relief plaintiff seeks would effectively shut down a significant new competitor for a substantial period of time. If the court were to issue such an order erroneously, it would cause substantial disruptions for defendants, their employees, their customers, and the end-users of their products. Thus, the balance of harms in this case does not tip decidedly in either direction. If the court is wrong in denying a preliminary injunction to plaintiff, the result will be irreparable harm to plaintiff. If the court were to grant a preliminary injunction erroneously, the result would be substantial, and perhaps irreparable, harm to defendants. Under these circumstances, the balance of harms does not favor the grant of injunctive relief, and the court should not grant injunctive relief where the plaintiff has not shown it is likely to succeed on the merits of its claims.

For the reasons set forth above, plaintiff Flotec’s motion for preliminary injunction is hereby DENIED.

So ordered.

1 . LeNoir Zaiser, the chairman of defendant SRI, testified that Flotec's Gilbert Davidson had told him that the flow disk in the Flotee regulator is patented. Tr. 402. SRI developed its own flow disk for its competing regulator. There is no claim for patent infringement here.

2 . Gilbert Davidson testified that he visited SRI in Florida and took additional drawings, additional sample components, and a complete regulator. He also testified that during that visit, he saw 10 or 12 boxes with disassembled regulators manufactured by several different companies, including Flotee. Davidson said he asked Zaiser: “You’re not getting into the regulator business, are you?" According to Davidson, Zaiser did not say no but said that SRI made components. Tr. 70-71. Davidson's testimony did not date that visit specifically, and the evidence shows that he stayed in touch with SRI’s Zaiser and Sheridan as late as 1997. There is no indication that the later contacts involved serious negotiations. The preponderance of the evidence shows that SRI's efforts to collect competing regulators and develop its own product occurred after it was clear, as a practical matter, that the possible deal between SRI and Flotee was dead.

3 . In a post-hearing brief, Flotee incorrectly indicated that Exhibit 138 had not been admitted into evidence. It was not admitted when first offered, but on the second day of the hearing it was admitted without objection. See Tr. 378.

4 . The Seventh Circuit has sometimes phrased the standard for the first element of a preliminary injunction in terms of a "better than negligible" likelihood of success on the merits. See, e.g., International Kennel Club of Chicago, Inc. v. Mighty Star, Inc., 846 F.2d 1079, 1084 (7th Cir.1988); Roland Mach. Co. v. Dresser Indus., Inc., 749 F.2d 380, 386-87 (7th Cir.1984). Where the moving party faces a threat of immediate and disastrous harm, and where the balance of harms lips decidedly in favor of the moving parly, the court has the discretion to prevent irreparable harm and to preserve the status quo even if it appears at the early stage of the lawsuit that the moving parly is not likely to prevail on the merits. See Abbott Laboratories v. Mead Johnson, 971 F.2d at 12 & n. 2 (explaining “sliding scale” approach to preliminary injunctive relief). Such cases are rare, and this is not one of them. A preliminary injunction would not preserve the status quo in this case and would inflict substantial harm on defendants. Because the balance of harms is not lopsided in this case, the court has used the term “likelihood” to describe the standard on the merits element.

5 . To protect the alleged secret character of the drawings, the court has kept the drawings themselves under seal. The details set forth in the texl, however, were the subject of detailed testimony in open court.

6 . The evidence does not show, however, that use of either a # 72 or # 74 drill would interfere with the proper operation of the device. In other words, the evidence does not show any tolerance information for this critical dimension. See Tr. 221 (Dr. MacDonald).

7 . Flotee also cites Biodynamic Technologies, Inc. v. Chattanooga Corp., 644 F.Supp. 607 (S.D.Fla.1986), in which the plaintiff and the defendant were negotiating a possible licensing agreement and, during the negotiations, plaintiff provided to defendant a sample of its newly developed product. The negotiations broke down and the defendant then developed and sold a very similar competing product. The district court granted summary judgment for plaintiff on a trade secret claim on the basis that the licensing negotiations created a confidential relationship that obligated the defendant not to use the information against the plaintiff's interests. 644 F.Supp. at 610-12. Distinguishing that case from this one, however, where no explicit demands for or promises of confidentiality were made, the Biodynamic court found as an undisputed fact that the defendant made "express promises of trust and confidentiality.” Id. at 612.

8 . Defendants have argued that Flotee is not entitled to a preliminary injunction because it waited too long to file its motion. Flotee has explained that it waited to file its motion for a preliminary injunction until it obtained evidence from Jeff Wyant, the former SRI vice president. The court does not view Wyant's testimony as compelling support for Flotec's case. But the court is not penalizing Flotee for failing to file a motion for preliminary injunction before it thought it had the evidence to support its motion.