PLAGER, Circuit Judge.

IMS Technology, Inc. ("IMS") appeals from the decision of the United States District Court for the Eastern District of Virginia. The district court granted summary judgment of noninfringement of U.S. Patent No. 4,477,754 ("the § 754 patent") against IMS and in favor of Haas Automation, Inc. and Gene Francis Haas (collectively "Haas"). See IMS Tech., Inc. v. Haas Automation, Inc., No. 98-143-A (E.D.Va. Oct. 2, 1998) (order). IMS argues on appeal that the district court erred in its claim construction and in its infringement analysis. Haas cross-appeals the district court's construction of certain claim limitations. Because the district court erred in its claim construction and infringement analysis and because, under a proper claim construction, there remain genuine issues of material fact concerning infringement with respect to some of the accused devices, we affirm-in-part, vacate-in-part, and remand.

BACKGROUND IMS is the assignee of the § 754 patent, which was originally assigned to Hurco Companies, Inc. ("Hurco"), a manufacturer of machine tools and machine tool controls. IMS is a wholly owned subsidiary of Hurco, formed to license the § 754 patent. Hurco assigned its rights in the § 754 patent to IMS in 1995.

The § 754 patent, which contains both apparatus and method claims, originally issued on October 16, 1984. After discovering prior art during prosecution of Japanese patent applications corresponding to the § 754 patent, Hurco initiated a reexamination proceeding before the United States Patent and Trademark Office ("PTO"). In the reexamination, claim 1, an apparatus claim, and claim 11, a method claim, both at issue in the instant case, were allowed without amendment. One limitation of claim 7, an apparatus claim also at issue in this case, was substantively amended. The PTO issued a reexamination certificate on March 21, 1995.

The § 754 patent relates generally to a control for a machine tool, such as a milling machine, which is used to cut or remove material from an object, referred to as a workpiece, through a machining operation. The type of control at issue is a numerical control ("NC"), which runs a program containing a series of numerical instructions and converts the instructions to electrical control signals. These control signals are applied to, for example, servo motors that control the movement of the machine tool along x, y, and z axes. Typically, a table holding the workpiece moves in the x (left and right) and y (fore and aft) directions, and a spindle holding a tool moves in the z (up and down) direction.

Before invention of the control claimed in the § 754 patent, a programmer created a control program using a standard part programming language such as EIA Standard RS-274-D, which comprises a series of standard codes commonly referred to as G-codes and M-codes. The programmer, usually not the operator of the machine tool, typically worked in a location remote from the machine tool. The programmer created a program by looking at a blueprint of the object to be machined, determining the series of machine tool operations (e.g., movements and cuts) required to make the object, and consulting a handbook of G- and M-codes corresponding to the operations. The completed code was reproduced on a punched paper tape. The machine tool operator fed the punch tape into the machine tool, which executed the program by converting the G- and M-codes into binary code, which was translated into electrical control signals. The coding process was cumbersome and time-consuming.

The invention claimed in the § 754 patent permits interactive programming of the machine tool on the machine shop floor. The machine tool operator himself creates a program by using a keyboard to respond to nested inquiries displayed on a CRT screen. See § 754 patent, col. 5, ll. 52-55; col. 6, ll. 4-10. In general, the program contains data blocks, each of which corresponds to one operational step of the machine tool. When the machine tool operator selects an operation, the control system may prompt him for additional parameters to be included in the data block for that operational step. For example, if the operator selects a "mill" operation, the system will prompt him for the selection of dimensional parameters, e.g., the coordinates for the start and end positions of the operation.

While the operator is creating a program, the program is stored in alterable memory (e.g., random access memory ("RAM")). In the embodiment disclosed in the written description of the § 754 patent, a program may be stored permanently on a tape cassette by means of a tape cassette transport included in the control. See id. at col. 6, ll. 49-52. The written description of the § 754 patent does not specify that programs are stored in any particular storage format. A program previously stored on a tape cassette can be read into alterable memory. See id. at col. 6, ll. 54-58. A microprocessor executes the program by using the information in the data blocks to produce control signals for directing the operation of the machine tool. See id. at col. 11, l. 16 — col. 12, l. 7.

Haas is a corporation which manufactures and sells machine tools with numerical controls. The accused Haas controls also provide interactive programming capability of machine tools on the machine shop floor. Some Haas controls have a floppy disk drive for storing programs. Others have only an RS-232 data port which can be connected to a storage device, such as a personal computer, for storing programs in ASCII format. Haas controls use programming systems known as Quickcode and Conversational Quickcode that assist the machine tool operator in creating a G- and M-code program. In Quickcode, the operator views a split screen. On the right side of the screen, the operator sees a group window that includes a compressed list of short descriptions of G-code operations. By rotating a jog handle clockwise, the operator navigates through the groups. When the operator finds the desired group, he turns the jog handle counterclockwise to see additional operations, called items. The operator selects an item, and the corresponding G-code appears on the left side of the window. The operator can then edit the G-code to change parameter values.

In Conversational Quickcode, the operator can program questions to solicit values corresponding to G-code. After the answers to such questions are placed in the G-code on the screen, the operator can edit the G-code as described above with respect to Quickcode.

The Haas controls store programs in G- and M-code format. During execution of programs, the Haas controls translate G-code into a binary format, which is converted into electrical signals delivered to the machine tool for directing its operation.

IMS filed suit against Haas alleging that Haas infringes at least apparatus claims 1 and 7 of the § 754 patent and induces infringement and contributes to infringement of at least apparatus claims 1 and 7 and method claim 11 of the § 754 patent. Those claims read as follows:

A programmable microcomputer control apparatus for controlling the relative motion between a tool and a workpiece comprising:

indicator means for providing at an output digital signals indicative of the relative position between the tool and the workpiece;

an alterable memory operable to retain a control program and control parameters;

a microprocessor unit coupled to the output of the indicator means and to the memory and operable to produce control signals dependent upon said indicator means output and said control parameters according to said control program;

control means for directing said control signals from the microprocessor unit to appropriate motion-providing means;

interface means for transferring a control program and control parameters from an external medium into said alterable memory and for recording the control parameter contents of said memory onto an external medium;

data entry means for loading control parameters into said memory through externally accessible data inputs independently of said interface means; and

display means for displaying control parameters, said control program being operable to display control parameter inquiries on the display means, whereby an operator may load control parameters into said memory through said data entry means in response to the inquiries, said apparatus including means to sequentially display data block inquiries and to display, in response to the loading of certain control parameters into said memory relating to the data block inquiries, separate displays of additional control parameter inquiries relating to information used in the data block which was the subject of the previous inquiry, whereby the sequential display of inquiries and direct loading of control parameters as to an operation can be used to make the use of the device simpler and more responsive to the operator.

A programmable microcomputer control apparatus for controlling the relative motion between a tool and a workpiece comprising:

indicator means for providing at an output digital signals indicative of the relative position between the tool and the workpiece;

an alterable memory operable to retain a control program and control parameters;

a processor unit coupled to the output of the indicator means and to the memory and operable to produce control signals dependent upon said indicator means output and said control parameters according to said control program, said control signals including programmed rate signals for controlling the rate of relative motion between the tool and the workpiece;

control means for directing said control signals from the processor unit to appropriate motion-providing means;

interface means for transferring a control program and control parameters from an external medium into said alterable memory and for recording the control parameter contents of said memory onto an external medium;

data entry means for loading control parameters into said memory through externally accessible data inputs independently of said interface means; and

feed rate adjust means externally and manually settable independent of said control parameters, said feed rate adjust means coupled to said processor unit, said processor unit recalculating said rate signals dependent on said feed rate adjust means to vary the rate of relative motion between the tool and the workpiece.

A method for automatically and interactively performing machining operations on a workpiece comprising the steps of:

entering the mode type and dimensional parameters for a machining operation into a microcomputer memory as a data block;

repeating said entering step for data blocks for any further operations and dimensions as necessary to complete processing of the workpiece; and

executing a microcomputer program utilizing said data blocks to direct a machine to perform said operations on a workpiece;

displaying on screen sequentially for observation and response by the operator, a plurality of inquiries regarding mode and dimensional parameters for individual data blocks; and

as to at least some of the individual data blocks, utilizing operator response to initiate and implement subsequent display of additional inquiries for observation and response by the operator to further define the parameters of the data block as to an operation.

§ 754 patent, col. 14, l. 41 — col. 15, l. 13; col. 15, ll. 37-68; col. 16, l. 55 — col. 17, l. 8 (emphasis added).

Upon Haas's motion for claim construction, and after receiving extensive briefing from both parties and conducting a limited hearing on the issue of claim construction, the district court issued a memorandum opinion and order in which it construed many of the terms in the claims of the § 754 patent. See IMS Tech., Inc. v. Haas Automation, Inc., No. CA-97-1043-A (E.D.Va. Sept. 23, 1998). The district court then held a hearing on Haas's motion for summary judgment of noninfringement on October 2, 1998. At that hearing, the district court judge orally amended the claim construction relating to the term "data block." Based on the claim construction, the district court granted Haas's motion for summary judgment of noninfringement.

IMS appeals aspects of the district court's claim construction and its grant of Haas's motion for summary judgment. IMS's appeal focuses on the meaning of two terms: "interface means" and "data block." The district court construed the "interface means" limitation, found in claims 1 and 7, as a means-plus-function limitation in accordance with 35 U.S.C. § 112, ¶ 6 (1994), and determined that the corresponding structure in the written description includes a tape cassette transport. Based on this claim construction, the district court concluded that the Haas controls do not literally infringe claims 1 and 7 as a matter of law because Haas's floppy disk drive and the tape cassette transport disclosed in the § 457 patent are not equivalent structures. The district court also found that Haas's floppy disk drive does not as a matter of law satisfy the "interface means" limitation under the doctrine of equivalents.

Regarding the term "data block," found in claims 1 and 11, the district court stated in its original claim construction order that "data block" was limited to the specific variables and sequence of inquiries set forth in the written description of the § 457 patent. At the summary judgment hearing, the district court added that the term "data block" does not encompass the use of G- and M-codes. The district court concluded that, because the Haas controls do not use the same variables and sequence of inquiries as those disclosed in the written description and rely on the use of G- and M-codes, Haas does not infringe claim 11 either literally or under the doctrine of equivalents.

IMS also appeals the district court's claim construction limiting the scope of claims 1 and 7 to "a control system for machine tools rather than an entire machine tool apparatus."

In its cross-appeal, Haas argues that, if the district court's grant of summary judgment is vacated, we should correct other errors in the district court's claim construction. Specifically, Haas appeals the district court's conclusion that the "indicator means" limitation in claims 1 and 7 covers systems that measure the location of the workpiece and the tool from a fixed and unalterable point. Haas also appeals the district court's construction of the "feed rate adjust means" limitation in claim 7, arguing that the district court construed the original claim language instead of the language as amended during reexamination.

DISCUSSION

In reviewing a district court's grant of summary judgment, we must make an independent determination as to whether the standards for summary judgment have been met. See Conroy v. Reebok Int'l, Ltd., 14 F.3d 1570, 1575, 29 USPQ2d 1373, 1377 (1994); Vas-Cath Inc. v. Mahurkar, 935 F.2d 1555, 1560, 19 USPQ2d 1111, 1114 (Fed. Cir. 1991); C.R. Bard, Inc. v. Advanced Cardiovascular Sys., Inc., 911 F.2d 670, 673, 15 USPQ2d 1540, 1542-43 (Fed. Cir. 1990). We view the evidence in a light most favorable to the non-movant, and draw all reasonable inferences in its favor. See SRI Int'l v. Matsushita Elec. Corp., 775 F.2d 1107, 1116, 227 USPQ 577, 581 (Fed. Cir. 1985) (en banc). A motion for summary judgment is properly granted if there is no genuine issue as to any material fact and the moving party is entitled to judgment as a matter of law. See Fed.R.Civ.P. 56(c).

An infringement analysis requires two steps: (1) claim construction to determine the scope and meaning of the asserted claims, and (2) a comparison of the properly construed claims with the allegedly infringing device or method to determine whether the device or method embodies every limitation of the claims. See Cybor Corp. v. FAS Techs., Inc., 138 F.3d 1448, 1454, 46 USPQ2d 1169, 1172 (Fed. Cir. 1998) (en banc). Claim construction is a matter of law over which we exercise independent review. See id. at 1456, 46 USPQ2d at 1174. Whether an accused device or method infringes a claim either literally or under the doctrine of equivalents is a question of fact. See Insituform Techs., Inc. v. Cat Contracting, Inc., 161 F.3d 688, 692, 48 USPQ2d 1610, 1614 (Fed. Cir. 1998). Thus, on appeal from a grant of summary judgment of noninfringement, we must determine whether, after resolving reasonable factual inferences in favor of the patentee, the district court correctly concluded that no reasonable jury could find infringement. See Voice Techs. Group, Inc. v. VMC Sys., Inc., 164 F.3d 605, 612, 49 USPQ2d 1333, 1337 (Fed. Cir. 1999).

An infringement analysis of a claim with limitations drafted pursuant to 35 U.S.C. § 112, ¶ 6 (1994), involves the same two steps — claim construction and a comparison of the accused device or method with the properly construed claims. Limitations contemplated by § 112, ¶ 6, often referred to as means-plus-function or step-plus-function limitations, recite a specified function to be performed rather than the structure, material, or acts for performing that function. Such limitations are "construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof." 35 U.S.C. § 112, ¶ 6 (1994). Claim construction of a § 112, ¶ 6 limitation includes identifying the claimed function and determining the corresponding structure or act disclosed in the specification, both of which are questions of law that this court reviews independently. See Chiuminatta Concrete Concepts, Inc. v. Cardinal Indus., Inc., 145 F.3d 1303, 1308, 46 USPQ2d 1752, 1755-56 (Fed. Cir. 1998).

For literal infringement of a § 112, ¶ 6 limitation, the second step of an infringement analysis begins with determining whether the accused device or method performs an identical function to the one recited in the claim. See Mas-Hamilton Group v. LaGard, Inc., 156 F.3d 1206, 1211-12, 48 USPQ2d 1010, 1015 (Fed. Cir. 1998). If the identical function is performed, the next step is to determine whether the accused device uses the same structure, materials, or acts found in the specification, or their equivalents. See id. at 1212, 48 USPQ2d at 1015. Whether an accused device or method infringes a claim with a § 112, ¶ 6 limitation, i.e., whether it performs the identical function with the same structure, materials, or acts described in the specification or an equivalent thereof, is a question of fact. See Odetics, Inc. v. Storage Tech. Corp., 185 F.3d 1259, 1268-69, 51 USPQ2d 1225, 1230-31 (Fed. Cir. 1999) (citing Palumbo v. Don-Joy Co., 762 F.2d 969, 975, 226 USPQ 5, 8 (Fed. Cir. 1985), and indicating that the holding in Markman v. Westview Instruments, Inc., 52 F.3d 967, 976-79, 34 USPQ2d 1321, 1327-29 (Fed. Cir. 1995), that claim construction is a question of law did not affect the holding in Palumbo that whether an accused device is a § 112, ¶ 6 equivalent is a question of fact). Thus, in order to uphold a summary judgment of noninfringement, we must conclude that no reasonable jury could have found otherwise.

A. Claim Construction

IMS argues that several aspects of the district court's claim construction are incorrect. We will examine each issue in turn using our standard canons of claim construction.

1. "Interface means" in claims 1 and 7

We agree with the district court's ruling, unchallenged by the parties, that the limitation "interface means for transferring a control program and control parameters from an external medium into said alterable memory and for recording the control parameter contents of said memory onto an external medium" in claims 1 and 7 is subject to § 112, ¶ 6. IMS asserts that the district court erred in its construction of the limitation by identifying the tape cassette transport as the relevant corresponding structure found in the specification. IMS further argues that, even if the tape cassette transport is the corresponding structure, the district court erred by concluding as a matter of law that the floppy disk drive in the accused Haas controls cannot be equivalent to the disclosed tape cassette transport.

The "interface means" limitation recites two functions, i.e., "recording" a control program and control parameters from an alterable memory onto an external medium, and "transferring" a control program and control parameters from the external medium into alterable memory. IMS contends that the corresponding structure in the specification, disclosed in the written description, is a tape cassette peripheral interface adapter ("PIA") because that is the only structure that performs the "transferring" function. In making its argument, however, IMS virtually ignores the "recording" function and misapplies the meaning of "transferring" in the context of the claim.

The claimed interface means records data from the alterable memory to an external medium and transfers data from the external medium to the alterable memory. The external medium disclosed in the written description is a tape cassette, and the alterable memory disclosed is a RAM. Thus, the structure corresponding to the "recording" and "transferring" functions must record data residing in the RAM onto the tape cassette and transfer data from the tape cassette to the RAM. IMS correctly states that the written description associates the PIA with these functions: "PIA 77 provides the inerfacing [sic] with a tape cassette which is utilized to load RAM 36. Also, a subsequently inserted cassette may be utilized to record stored RAM program data through PIA 77." § 754 patent, col. 4, ll. 46-48. The written description, however, also associates a tape cassette transport with the functions of transferring data from the tape cassette to the RAM and recording data from the RAM onto the tape cassette: "A tape cassette transport apparatus is shown generally at 225 for receiving a magnetic tape cassette operable to read or write data to or from the RAM memory. With mode switch 203 in the tape auto mode, the data block entries may be recorded onto a tape cassette." § 754 patent, col. 6, ll. 49-52. In view of both of these passages, it is apparent that both the PIA and the tape cassette transport are necessary for transferring data from the tape cassette to the RAM and recording data from the RAM onto the tape cassette.

IMS argues that the "transferring" function is only an "interfacing" function between the RAM and the tape cassette transport. But the claims require the interface means to transfer data to the RAM from an external medium, i.e., a tape cassette, not just from the tape cassette transport. The disclosed PIA performs only part of the "transferring" function, i.e., transferring from the tape cassette transport to the RAM. The tape cassette transport itself transfers data from the tape cassette to the PIA, and therefore both the PIA and the tape cassette transport are part of the disclosed structure corresponding to the "transferring" function of the interface means.

Similarly, the interface means must record data from the RAM onto the tape cassette. The tape cassette transport receives data from the RAM via the PIA and completes the recording function by recording data onto the tape cassette. The PIA and the tape cassette transport together perform this function and thus are both part of the disclosed structure corresponding to the claimed interface means.

IMS also argues that, under the doctrine of claim differentiation, the interface means cannot be limited to a means for "reading" and "writing" because claim 2, which depends from claim 1, places that additional limitation on the interface means: "said interface means includes means for reading from and writing onto a magnetic stored information input." IMS further argues that the "interface means" cannot be limited to a tape cassette transport because claim 3, which depends from claim 2, specifically claims a tape cassette transport as the means for reading and writing.

We reject IMS's argument that the doctrine of claim differentiation requires that the corresponding structure of the interface means of claim 1 be limited to the disclosed PIA. The scope of claim 3 is clearly narrower than that of claim 1 because claim 3 covers only a tape cassette transport, whereas claim 1 covers a tape cassette transport and its equivalents in accordance with § 112, ¶ 6. See Laitram Corp. v. Rexnord, Inc., 939 F.2d 1533, 1538, 19 USPQ2d 1367, 1371 (Fed. Cir. 1991) (holding that claim differentiation is maintained when the disclosed structure corresponding to an independent § 112, ¶ 6 claim is recited in a dependent claim). The scope of claim 2 is also narrower than that of claim 1, at least because it limits the external medium to a magnetic stored information input.

Furthermore, the proper claim construction does not give the same meaning to "recording" in claim 1 and to "writing" in claim 2 as IMS contends. Rather, the "reading" and "writing" functions in claim 2 are the parts of the "transferring" and "recording" functions of the "interface means" that are performed by the disclosed tape cassette transport, rather than the PIA. In any event, it is permissible for claim 1 and claim 2 to have similar scope after each is correctly construed in light of the structures disclosed in the written description, because the judicially-created doctrine of claim differentiation cannot override the statutory mandate of § 112, ¶ 6. See id. (noting that claim differentiation is a guide, not a rigid rule).

In sum, we conclude that the proper construction of the "interface means" limitation covers the disclosed structure, which includes the PIA and tape cassette transport, and its equivalents in accordance with § 112, ¶ 6. IMS argues alternatively that, even under this claim construction, the district court erred by finding that, as a matter of law, Haas's floppy disk drive could not be an equivalent to the disclosed tape cassette transport. We shall address this issue infra in connection with our infringement analysis.

2. "Data block" in claims 1 and 11

The district court construed the term "data block," which appears in several steps of method claim 11 and in the display means limitation of claim 1, as limited to the specific set of variables disclosed in the written description and the disclosed sequence of inquiries regarding those variables. The district court also held that a "data block," as properly construed, cannot include or refer to any instructions written in a known part programming language, such as G- and M-codes. We agree with IMS that the district court erred in construing the term "data block" so narrowly.

The parties agree that the claim limitations in which the term appears are subject to § 112, ¶ 6. They dispute, however, whether the meaning of the term "data block" itself is affected by § 112, ¶ 6. Haas contends that the district court correctly held that § 112, ¶ 6 applies to limit the meaning of the term "data block" to the variables and sequence disclosed in the written description. We hold that this is an erroneous application of § 112, ¶ 6.

Section 112, ¶ 6 does not limit all terms in a means-plus-function or step-plus-function clause to what is disclosed in the written description and equivalents thereof; § 112, ¶ 6 applies only to interpretation of the means or step that performs a recited function when a claim recites insufficient structure or acts for performing the function. See O.I. Corp. v. Tekmar Co., 115 F.3d 1576, 1581, 42 USPQ2d 1777, 1780 (Fed. Cir. 1997). In claim 1, the display means includes a "means to sequentially display data block inquiries." The recited function consists of sequentially displaying data block inquiries, and the claim recites no structure supporting the means for performing that function. Therefore, in accordance with § 112, ¶ 6, the means is construed to cover the disclosed structure, i.e., a CRT, and its equivalents. The "data block" is not the means that causes the sequential display and is therefore not subject to construction under § 112, ¶ 6. See id. (holding the term "passage" within the clause "means for passing the analyte slug through a passage" is not subject to § 112, ¶ 6). Similarly, in claim 11, assuming that the several limitations in which the term "data block" appears are subject to § 112, ¶ 6, the "data block" is not part of the steps that perform the recited functions. We therefore must construe the term "data block" according to our standard claim construction methodology without application of § 112, ¶ 6. See id.

To construe the term "data block," we look first to the claim language, the written description, and the prosecution history. See Zelinski v. Brunswick Corp., 185 F.3d 1311, 1315, 51 USPQ2d 1590, 1592-93 (Fed. Cir. 1999). The ordinary and customary meaning of "data block" in the context of programmable machine tools is a computer data structure containing the information needed by a machine tool to perform a single machining operation. Haas contends that the ordinary meaning does not apply because the patentee gave the term "data block" a special meaning by describing the preferred embodiment in the patent as follows: "The sequence of inquiries on the CRT screen for a data block follows the sequence: data block number, machine mode, control mode, X dimension, Y dimension, Z dimension, feed rate, pack rate and tool number." § 754 patent, col. 6, ll. 4-7. We find nothing in the written description, however, that indicates this is the patentee's specialized definition of "data block." Rather, the written description merely describes the preferred embodiment, and to limit "data block" to the sequence of variables disclosed would be to impermissibly read a particular embodiment into the claim. See Comark Communications, Inc. v. Harris Corp., 156 F.3d 1182, 1186-87, 48 USPQ2d 1001, 1005 (Fed. Cir. 1998).

Haas also argues that the following statement, made by the applicant during prosecution of the § 754 patent, limits the meaning of the term "data block" to the specific variables and display sequence set forth in the written description: "particular attention should be paid to the term `data block' as used in the specification and claims when considering the scope of these claims." Joint App. at 319. This statement alone cannot narrow the scope of the claims. Haas contends further that a brochure submitted to the Examiner by the applicant limits the term "data block" to the specific variables and display sequence disclosed in the written description. This argument is also without merit, for it is clear from the prosecution history that the brochure was intended only to assist the Examiner in understanding the preferred embodiment. In sum, we hold that nothing in the written description or prosecution history is sufficient to overcome the ordinary meaning and to require that the meaning of "data block" be limited to the specific variables and display sequence disclosed in the written description as the preferred embodiment.

Another issue raised by Haas is whether other statements made during the prosecution and reexamination of the § 754 patent limit a "data block" to a format that in no way relies on G- and M-codes. In distinguishing over a prior art reference, the applicant stated that the claims were "intended to focus upon a novel interactive system in which the operator was sequentially asked questions and via the display the operator in response simply answered the question rather than have to worry about how to program the device." Joint App. at 319. The applicant made two more similar statements in responding to later office actions. IMS argues that these statements were intended to convey the interactive nature of creating a parts program using the claimed invention and do not disclaim the use of prior art programming languages such as G- and M-codes so long as a program is created interactively according to the claims.

During reexamination, the patentee distinguished several other prior art references by emphasizing the interactive programming technique of the invention, including the use of nested inquiries to prompt the user to enter additional information based on the user's response to a previous inquiry. No statements regarding the prior art clearly disclaim the use of G- and M-codes as Haas contends. In the last response to the PTO during reexamination, the patentee contrasted the invention with the cumbersome prior art programming process:

The present invention utilizes an interactive display which operates in a question and answer format without resorting to the M and G codes of the machine tool. . . . The interactive processing of the data blocks enables the machine tool operator to perform the tasks of both the programmer and machinist on the shop floor.

Joint App. at 932.

We agree with IMS that this statement, taken in context, as well as the other statements made during prosecution and reexamination, do not disclaim the use of G- and M-codes in the claimed invention. The purpose of the statements was to emphasize the interactive nature of the invention as an improvement over the prior art programming method in which a user had to create a program line-by-line using only G- and M-codes. Thus, the claimed invention does not require the creation of a G- and M-code program; at the same time, however, the invention does not preclude the creation of a G- and M-code program so long as it is created using the claimed interactive inquiry process.

Haas further argues, in support of its contention that a data block cannot be stored in G- and M-code format, that the written description never mentions the use of G- and M-codes. While it is true that the embodiment disclosed in the written description does not purport to store data blocks in G- and M-code format, the use of data blocks in the invention as claimed is independent of the storage format. The claimed invention is an apparatus and method for interactively creating a program for controlling the machine tool. That program contains data blocks, each of which, according to the customary meaning of the term, includes the information needed by a machine tool to perform one machining operation. It is irrelevant to the claimed invention whether a data block is stored as a line of G- and M-code, in a binary format, or in any other format.

In sum, we conclude that the proper construction of the term "data block" is a computer data structure containing the information needed by a machine tool to perform a single machining operation. As properly construed, a "data block" is not limited to the specific set of variables and display sequence disclosed in the written description and does not preclude the use of G- and M-codes.

3. "Control apparatus" in claims 1 and 7

IMS contends that the district court erred when, based on the preambles of claims 1 and 7, it limited those claims to a "control system for machine tools rather than an entire machine tool apparatus." To the extent that the district court's claim interpretation precludes a finding of infringement by a machine tool apparatus that includes the claimed control-related limitations, we agree that the district court improperly limited the claims.

"[A] claim preamble has the import that the claim as a whole suggests for it." Bell Communications Research, Inc. v. Vitalink Communications Corp., 55 F.3d 615, 620, 34 USPQ2d 1816, 1820 (Fed. Cir. 1995). If the preamble adds no limitations to those in the body of the claim, the preamble is not itself a claim limitation and is irrelevant to proper construction of the claim. See Pitney Bowes, Inc. v. Hewlett-Packard Co., 182 F.3d 1298, 1305, 51 USPQ2d 1161, 1165-66 (Fed. Cir. 1999). That is the case here. The phrase "control apparatus" in the preamble merely gives a descriptive name to the set of limitations in the body of the claim that completely set forth the invention. Its use does not limit the claims, as Haas contends, to a control apparatus that is separate from the machine tool. The claim is infringed by any apparatus encompassing all of the limitations in the body of the claim. Such an infringing apparatus may be a machine tool apparatus that includes the claimed control features or a control apparatus that is separate from and communicates with a machine tool apparatus.

The prosecution history is consistent with this interpretation. In a restriction requirement, the PTO recognized that the claimed invention, a machine tool control, was part of a machine tool. In sum, the applicant's choice to elect claims directed to machine tool control rather than machine control structure does not limit the scope of the claims to a control apparatus that is separate from the machine tool itself.

B. Infringement Analysis

In light of the proper claim construction, the key infringement issue before us is whether the accused Haas systems contain the claimed "interface means." As noted, the district court based its finding of noninfringement of claims 1 and 7 on its conclusion that the Haas floppy disk drive is not equivalent, under either § 112, ¶ 6 or the doctrine of equivalents, to the disclosed tape cassette transport.

To determine whether the Haas systems literally contain the "interface means" of claims 1 and 7, it is necessary to first determine whether the accused devices perform the identical functions recited in the claims — recording data from memory onto an external medium and transferring data from the external medium to memory. There is no dispute that the Haas systems with a floppy disk drive perform these functions. Since the Haas systems do not contain the same structure as the disclosed PIA and tape cassette transport, the only question is whether the Haas systems contain a structure that is an equivalent of the disclosed structure. Assuming that the Haas systems include an interface device equivalent to the PIA, a fact uncontested by Haas, the determinative issue under § 112, ¶ 6 is whether the floppy disk drive in the Haas system is equivalent to the disclosed tape cassette transport.

This court has on several occasions compared statutory equivalence under § 112, ¶ 6 and the judicial doctrine of equivalents. See, e.g., Odetics, Inc. v. Storage Tech. Corp., 185 F.3d 1259, 1267, 51 USPQ2d 1225, 1229-30 (Fed. Cir. 1999); Al-Site Corp. v. VSI Int'l, Inc., 174 F.3d 1308, 1319-21, 50 USPQ2d 1161, 1167-68 (Fed. Cir. 1999); Chiuminatta Concrete Concepts, Inc. v. Cardinal Indus., Inc., 145 F.3d 1303, 1310, 46 USPQ2d 1752, 1757-58 (Fed. Cir. 1998); Cybor Corp. v. FAS Techs., 138 F.3d 1448, 1467, 46 USPQ2d 1169, 1184 (Fed. Cir. 1998) (en banc) (Mayer, C.J., concurring); Alpex Computer Corp. v. Nintendo Co., 102 F.3d 1214, 1222, 40 USPQ2d 1667, 1673-74 (Fed. Cir. 1996); Valmont Indus., Inc. v. Reinke Mfg. Co., 983 F.2d 1039, 1042-44, 25 USPQ2d 1451, 1453-55 (Fed. Cir. 1993). While acknowledging that there are differences between § 112, ¶ 6 and the doctrine of equivalents, this court on several occasions has indicated that the tests for equivalence under § 112, ¶ 6 and the doctrine of equivalents are "closely related," involving "similar analyses of insubstantiality of the differences." Chiuminatta, 145 F.3d at 1310, 46 USPQ2d at 1757-58; see also Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17, 28, 41 USPQ2d 1865, 1870 (1997) (stating that application of § 112, ¶ 6 "is an application of the doctrine of equivalents in a restrictive role. . . ."); Valmont, 983 F.2d at 1043, 25 USPQ2d at 1455 ("The word `equivalent' in section 112 invokes the familiar concept of an insubstantial change which adds nothing of significance."). Thus, a reduced version of the well-known tripartite test for the doctrine of equivalents has been applied in the § 112, ¶ 6 context to determine if the differences are insubstantial, i.e., after determining that the accused device performs the identical function, as required by statute, whether it performs the function in substantially the same way to achieve substantially the same result. See Odetics, 185 F.3d at 1267, 51 USPQ2d at 1229-30; see also Dawn Equip. Co. v. Kentucky Farms Inc., 140 F.3d 1009, 1019-20, 46 USPQ2d 1109, 1116 (Fed. Cir. 1998) (Plager, J., additional views) (suggesting use of the tripartite test "to resolve the question of insubstantial changes" under § 112, ¶ 6). Evidence of known interchangeability between structure in the accused device and the disclosed structure has also been considered an important factor. See Al-Site, 174 F.3d at 1316, 50 USPQ2d at 1165; Chiuminatta, 145 F.3d at 1309, 46 USPQ2d at 1757 (citing Graver Tank Mfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 609, 85 USPQ 328, 331 (1950)).

In light of the similarity of the tests for equivalence under § 112, ¶ 6 and the doctrine of equivalents, the context of the invention should be considered when performing a § 112, ¶ 6 equivalence analysis just as it is in a doctrine of equivalents determination. See Texas Instruments, Inc. v. ITC, 805 F.2d 1558, 1563, 231 USPQ 833, 835 (Fed. Cir. 1986) ("It has long been recognized that the range of permissible equivalents depends upon the extent and nature of the invention. . . ."); cf. Warner-Jenkinson, 520 U.S. at 40, 41 USPQ2d at 1875 (noting that in a doctrine of equivalents determination, "[a]n analysis of the role played by each element in the context of the specific patent claim will thus inform the inquiry as to whether a substitute element matches the function, way, and result of the claimed element, or whether the substitute element plays a role substantially different from the claimed element"). As a result, two structures that are equivalent in one environment may not be equivalent in another. More particularly, when in a claimed "means" limitation the disclosed physical structure is of little or no importance to the claimed invention, there may be a broader range of equivalent structures than if the physical characteristics of the structure are critical in performing the claimed function in the context of the claimed invention. Thus, a rigid comparison of physical structures in a vacuum may be inappropriate in a particular case. Indeed, the statute requires two structures to be equivalent, but it does not require them to be "structurally equivalent," i.e., it does not mandate an equivalency comparison that necessarily focuses heavily or exclusively on physical structure.

In some cases, an analysis of insubstantial differences in the context of the invention results in a finding of equivalence under § 112, ¶ 6 even though two structures arguably would not be considered equivalent structures in other contexts, e.g., if performing functions other than the claimed function. See Odetics, 185 F.3d at 1269-71, 51 USPQ2d at 1231-32 (reinstating jury verdict of infringement when there was evidence that a "bin array" with a cam and cam follower mechanism performed a rotary function in the same way as a "rotary means" with a gear mechanism by receiving force); Al-Site, 174 F.3d at 1315-17, 50 USPQ2d at 1164-65 (affirming jury verdict of infringement based on expert testimony of known interchangeability of glue and rivet as a "fastening means" on hanger tag for glasses). But see Odetics, 185 F.3d at 1277-79, 51 USPQ2d at 1237-38 (Lourie, J., dissenting) (criticizing majority for focusing exclusively on function and not on structure). In other cases, in which the specific physical features of the structure corresponding to the "means" limitation may have more relevance to the claimed invention, a finding of noninfringement results. See Chiuminatta, 145 F.3d at 1309-10, 46 USPQ2d at 1757 (finding wheels and skid plate not equivalent for supporting surface of concrete, particularly since there was no allegation that one skilled in the art recognized the interchangeability of structures for performing claimed function).

Turning to the case at hand and the issue of whether Haas's floppy disk drive can be a § 112, ¶ 6 equivalent of the disclosed tape cassette transport, we consider the substantiality of their differences in the context of the claimed invention. The invention is directed to an apparatus that permits interactive programming of a machine tool. The transferring and recording functions of the claimed "interface means" merely provide a way of storing programs created using the inventive programming apparatus and process. This does not appear to be a case in which any physical characteristics of the interface means, such as the specific format of recorded data and the mechanism for accessing data, are important to the invention. IMS has provided some evidence of structural similarities between a floppy disk drive and a tape cassette transport, and, while there are admittedly physical differences, there is at least an issue of fact as to whether those differences are substantial in light of the role played by the "interface means" in the claimed invention. One way to address that question is to ask whether the structures perform the same function in substantially the same way to achieve substantially the same result. IMS has also supplied evidence that one of ordinary skill in the art would have recognized the interchangeability of a floppy disk drive and a tape cassette transport for performing the transferring and recording functions in the claimed invention. Such evidence should be considered in a § 112, ¶ 6 equivalence determination.

In light of the evidence presented by IMS regarding the insubstantiality of the differences between a floppy disk drive and a tape cassette transport in the context of the claimed invention, we conclude that there are genuine issues of material fact regarding literal infringement under § 112, ¶ 6 of claims 1 and 7. Furthermore, because of the similarities in analysis of equivalence under § 112, ¶ 6 and the doctrine of equivalents, there also remain genuine factual issues regarding infringement of claims 1 and 7 under the doctrine of equivalents. Accordingly, with respect to the Haas systems containing floppy disk drives, we vacate the grant of summary judgment of noninfringement of claims 1 and 7 and remand for further proceedings consistent with this opinion. On remand, the district court is instructed to apply the correct construction of the terms "data block" in claim 1 and "control apparatus" in claims 1 and 7 as discussed above.

Haas also manufactures and sells systems that include only an RS-232 port for connection to a storage device, but do not include the storage device itself. These systems do not perform the functions of transferring and recording data because there is no external medium. Therefore, the Haas RS-232 systems do not contain the claimed "interface means," and we affirm the district court's grant of summary judgment that these systems do not directly infringe claims 1 and 7, either literally or under the doctrine of equivalents. This holding does not, however, preclude a finding that Haas induces infringement or contributorily infringes if, for example, a Haas customer connects a Haas system to an external storage device. Thus, with respect to the RS-232 systems, we vacate summary judgment regarding Haas's liability for inducement of infringement and contributory infringement of claims 1 and 7 and remand for further proceedings consistent with this opinion.

Finally, the only disputed claim language in claim 11 is the term "data block," which the district court construed incorrectly, as explained above. Because the district court granted summary judgment of noninfringement with respect to claim 11 based on an incorrect construction of the term "data block", we vacate the summary judgment of no literal infringement and noninfringement under the doctrine of equivalents with respect to claim 11.

C. Haas's Cross-Appeal

Because we vacate the district court's grant of summary judgment, we now address Haas's cross-appeal regarding claim construction. Haas contends that the district court erroneously construed the limitation that reads "indicator means for providing at an output digital signals indicative of the relative position between the tool and the workpiece." The district court correctly concluded that this claim limitation is subject to § 112, ¶ 6. The claimed function is providing output signals indicative of the relative position between the tool and the workpiece, and the corresponding structure in the written description is hardware circuitry — an encoder and an up/down counter as shown in Figure 4 for each axis of movement. The dispute between the parties centers on the meaning of the phrase "indicative of the relative position between the tool and the workpiece." Haas contends that a signal that outputs absolute position, measured with reference to a fixed point, cannot be a signal indicative of relative position within the meaning of the claims. The district court rejected Haas's proposed claim construction and held that an indicator that indirectly measures the distance between the tool and the workpiece by indicating the distance of each to a fixed point does indeed indicate the "relative position."

The written description supports the district court's construction. In the disclosed embodiment, the machine tool operator sets a "table zero" point by moving the table containing the workpiece to a desired location and pushing a button. At this point, the microprocessor software reads the digital outputs of the counters corresponding to the x and y axes and records those outputs as the zero reference point. The counter outputs themselves are arbitrary 8-bit words and are not physically reset to zero. Similarly, the operator can perform a "tool calibration" procedure by lowering the spindle to a desired zero reference point and pushing a button, at which point the microprocessor software saves the digital output of the z axis counter as a zero reference point. As the machine tool performs an operation, the encoders output "clicks" when they detect movement of the workpiece along the x and y axes and the tool along the z axis. The counters count the number of clicks in the positive and negative directions and output a digital signal representing the number of clicks. At specified time intervals, the microprocessor calculates the actual positions of the tool and workpiece by comparing the counter outputs with those saved as the zero reference point.

The digital signal outputs of the counters are "indicative of the relative position between the tool and the workpiece" because they represent the distance and direction the tool and the workpiece have moved from a zero reference point and can therefore be used to determine the relative position between the tool and the workpiece. Thus, in light of the written description, "indicative of the relative position" simply means that the signals can be used to determine the relative position between the tool and the workpiece. Although it may be common for the operator to perform a "table zero" operation when the tool is directly above the workpiece, thus using the workpiece to establish a zero reference point, nothing about the digital signals from the counters directly reflects the position of the tool relative to the workpiece as Haas incorrectly states. Only when compared to the counter values read when the workpiece and tool were at the zero reference point can the counter outputs be used to determine a position, relative or otherwise. Similarly, signals representing the absolute positions of the tool and workpiece with respect to a fixed point, such as in the Haas system, are relevant only as compared to the fixed reference point. Nothing in the claims or written description precludes a signal that directly indicates the distance of the tool and workpiece from a fixed reference position from being a signal "indicative of the relative position between the tool and the workpiece."

We are not persuaded by Haas's argument that IMS surrendered coverage of absolute positioning systems during reexamination of the § 754 patent when explaining the pertinency of several prior art references. While the patentee stated that the Eaton reference, U.S. Patent No. 3,821,525, discloses sensors indicating position and a Japanese Kokai reference discloses a position register storing absolute position, the patentee did not distinguish these structures from the structure disclosed in the § 754 patent. Indeed, it is not clear in either reference that positions are measured with respect to a fixed point established by the machine tool, as in the Haas systems, rather than a fixed point established by the operator, as in the disclosed embodiment of the § 754 patent. Both references distinguish "absolute" positioning from "incremental" or "point-to-point" positioning, but those terms appear to refer to the reference point for entering successive operational commands; "absolute" positions are entered as coordinates from a fixed reference point, whereas "incremental" positions are entered as coordinates with respect to the position after the last operation. In light of the ambiguity of the patentee's statements and the subject matter actually disclosed in the references, we cannot say that the patentee clearly disavowed coverage of absolute positioning systems during reexamination. See York Prods., Inc. v. Central Tractor Farm Family Ctr., 99 F.3d 1568, 1575, 40 USPQ2d 1619, 1624 (Fed. Cir. 1996) ("Unless altering claim language to escape an examiner rejection, a patent applicant only limits claims during prosecution by clearly disavowing claim coverage.").

Haas also appeals the district court's construction of the "feed rate adjust means" in claim 7. Although in its claim construction order the district court quoted language from claim 7 as originally issued instead of from the claim as amended in reexamination, the district court correctly construed this claim limitation. The claimed function of the "feed rate adjust means" is to allow the operator to adjust manually the feed rate independently of the control parameters in memory. The "recalculating" function is performed by the processor unit, not the "feed rate adjust means," and thus is not subject to construction in accordance with § 112, ¶ 6. Therefore, it is unnecessary to identify structure corresponding to the recalculation function or to discuss the method of recalculation. The term "recalculating" possesses its ordinary meaning.

SUMMARY

We hold that the district court made several errors with respect to claim construction. First, under a proper claim construction, the "interface means" limitation in claims 1 and 7 should be construed to cover the corresponding structure disclosed in the written description, including a PIA and a tape cassette transport, and equivalents thereof in accordance with § 112, ¶ 6. Next, the proper construction of the term "data block" in claims 1 and 11 is a computer data structure containing the information needed by a machine tool to perform a single machining operation. As properly construed, a "data block" is not limited to the specific set of variables and display sequence disclosed in the written description and does not preclude the use of G- and M-codes. Finally, the use of the term "control apparatus" in claims 1 and 7 does not preclude infringement of those claims by a machine tool apparatus that includes a control system. On remand, the district court is instructed to apply the construction of these terms as has been explained above.

In view of the proper claim construction, there are genuine issues of material fact as to whether a floppy disk drive is an equivalent of the disclosed tape cassette transport. Accordingly, we vacate the district court's grant of summary judgment of noninfringement of claims 1 and 7 with respect to the Haas systems containing floppy disk drives and remand for further proceedings consistent with this opinion.

We affirm the grant of summary judgment that the Haas systems that include only an RS-232 port do not directly infringe claims 1 and 7, either literally or under the doctrine of equivalents. Our holding, however, does not foreclose a finding that Haas induces infringement or contributorily infringes. Thus, with respect to the RS-232 systems, we vacate summary judgment regarding Haas's liability for inducement of infringement and contributory infringement of claims 1 and 7 and remand for further proceedings consistent with this opinion.

Because the district court erred in its construction of the term "data block," we vacate the district court's summary judgment of noninfringement of claim 11 and remand for further proceedings consistent with this opinion.

Finally, we affirm the district court regarding the claim construction issues raised in Haas's cross-appeal.

CONCLUSION

The decision of the district court is affirmed-in-part and vacated-in-part, and the case is remanded for further proceedings consistent with this opinion.

AFFIRMED-IN-PART, VACATED-IN-PART, AND REMANDED.

COSTS

Each party shall bear its own costs.