T 1005/02 (Fingerprinting magnetic media/WASHINGTON UNIVERSITY) 20-10-2005

Reasons for the Decision

1. The appeal complies with the requirements of Articles 106 to 108 and Rules 1(1) and 64 EPC and is thus admissible.

2. The appeal is allowable on the basis of auxiliary request 6 only. The higher-ranking requests do not meet the requirements of the EPC.

Novelty concerning main request and auxiliary requests 1 to 4

3. The subject-matters of the respective claims 1 of the main request and auxiliary requests 1 to 4 do not meet the requirement of novelty (Articles 52(1) and 54 EPC).

3.1 The generic subject-matter of these claims is a "device for fingerprinting an object for the later determination of its identity", which essentially means that the device is suitable for "labelling" the object with a kind of "fingerprint", i.e. a signature unique to the object (see the WO-application, page 8, lines 17 to 20 and page 9, lines 13 to 24). This function of the device includes the detection of the fingerprint on the object and the recording of the characteristics of the detected fingerprint for later verification, authenti cation, etc.

3.2 The claims, however, also refer expressly and in considerable detail, to features of the object which becomes labelled with its fingerprint using the device. These in-use features may define a potentially new use of a possibly old device and may thus camouflage lack of novelty regarding the device itself.

Abstracting, for a moment, from the properties of the object when in use with the device, a "common denominator" of the subject-matters of the claims under consideration can be formulated as follows:

The device has:

- (a) means for determining a noise signal from the object comprising the fingerprint for said object and

- (b) means for recording the noise signal for later comparison with the later determination of said noise signal to thereby identify said object.

To this common subject-matter, the individual requests add the following features:

Claims 1 of the main and auxiliary requests 1 to 4:

- (c) the noise signal is a remanent noise for a medium portion of the object.

Claims 1 of auxiliary requests 2 and 3, in addition:

- (d) said magnetic medium portion has a microstructure

- (e) said remanent noise corresponds to said microstructure.

Claim 1 of auxiliary request 2A, in addition to claim 1 of auxiliary request 2:

- (f) the microstructure comprises particles or grains being thousands of Angstroms or less in dimension.

Claim 1 of auxiliary request 2B, in addition to claim 1 of auxiliary request 2:

- (g) the microstructure comprises particles or grains being hundreds to thousands of Angstroms in dimension.

Claim 1 of auxiliary requests 4, in addition to claim 1 of auxiliary request 3:

- (h) said magnetic medium portion comprising a plurality of microcrystalline structures that are arranged in a random pattern

- (i) the remanent noise is determined from a region of said magnetic medium portion which is substantially uniformly magnetized.

3.3 The prior art document D1 anticipates a device for fingerprinting an object for the later determination of its identity having features (a) and (b) (see above):

Figure 2 shows "a sensor or read head 28 of a standard variety", moved relative to a section 24 of a region 20 of a magnetic material such as the region 12 or the region 16 in figure 1 (see document D1, column 8, lines 55 to 62 and column 9, lines 23 to 28). Regions 12 and 16 contain a fingerprint ("a detectable randomly varying magnetic characteristic unique to the object", see document D1, column 7, line 61 to column 8, line 28) which can be used for authenticating and identifying the object.

3.4 Claim features (c),(d),(e),(f),(g), and (h) (see above) merely define the object and the remanent noise produced when the device is in use. They do not define a functionality of the device which allows to distinguish it from possibly other devices. These claim features have thus no direct relevance to novelty of the device itself.

3.5 The microstructure of the magnetic medium as defined in feature (f) might only contribute to the patentability of the device if it imposes a particular functionality on the device, for example, if it requires, as it was argued by the appellant, a novel and inventive design of the device enabling it to detect the noise signal from such a magnetic medium. However, the Board rejects this line of argument for the following reasons:

The fingerprint signal is the detected, so-called "remanent noise", which arises from the permanent microstructure of the magnetic medium and which is "characteristic of that permanent microstructure after practically any magnetic history" (see the WO-publication, page 2, lines 25 to 28). Thus, the "remanent noise" itself is related only to intrinsic features of the object and the magnetic history of the medium.

Moreover, the invention can be implemented by conventional recording heads as used in prior art read or read/write devices (see the WO-publication, for example page 8, line 33 to page 9, line 10). The fact that the detection of the remanent noise does not require any recording head or read device specially adapted in this respect was confirmed by the appellant's representative and by Prof. Indeck at the oral proceedings before the Board. The microstructure of the object and the characteristics of the remanent noise do thus not define any relevant difference between the claimed device and the prior art of document D1.

3.6 Finally, document D1 also anticipates feature (i) as defined in point 3.2 above. Column 8, lines 17 to 27 specifically indicates that magnetic material even in an erased state can be used for identification and authentication. Plot B of figure 3D (see also the corresponding text in document D1, column 10, lines 45 to 58) gives an example of such a fingerprint. It shows a continuously but randomly varying waveform of residual noise which is sensed when no signal is recorded (see plot A), and which "can be recorded and used in subsequent verification of the medium" (loc.cit.). Plot A clearly shows that the region from which the fingerprint is taken is uniformly magnetized.

3.7 In the appellant's view document D1 diverted the skilled person from the present concept of using the magnetic microstructure for fingerprinting by recording a discontinuous signal along the magnetic medium portion for enhancing the magnetic fingerprint characteristics. Such an enhancing signal hid and masked the remanent noise produced by the microscopic structure of magnetic medium and made detection impossible.

The Board does not share this view: recording an enhancing signal, although a preferred embodiment, is not essential to the technique of document D1. A recordable magnetic material in an erased state may be used for identification and authentication (loc.cit.). Plot B of figure 3D shows an unenhanced randomly varying waveform, which can be sensed and recorded for subsequent verification of the medium (see also document D1, column 10, lines 45 to 58).

3.8 The circumstance that the micro- or even submicrostructure of the magnetic material should expressly be exploited for fingerprinting does not justify a different assessment, even if the appellant's argument is accepted that the "enhancing signal" in document D1 extends over a macroscopic region of the magnetic medium.

The enhancing signal amplifies and fixes the detectability of the randomly varying characteristic of the magnetic medium and is thus a kind of spatial noise filter sensitive only at the signal peaks (see document D1, column 8, line 55 to column 9, line 2). Although the enhancing signal may extend over a macroscopic distance of a couple of inches this filter probes the magnetic material with a resolution in the order of the peak width, i.e. about 100 micrometers for example (see column 12, line 58 to column 13, line 4).

3.9 In summary, the subject-matters of claims 1 of the main request and auxiliary requests 1 to 4 are anticipated by document D1 and do thus not meet the requirement of novelty (Articles 52(1) and 54 EPC).

Added subject-matter in claims of the main request and auxili ary requests 2, 2A, 2B, and 5

4. Article 123(2) EPC prohibits amendments of the application which add new subject-matter to the application as filed. The main request and auxiliary requests 2, 2A, 2B, with respect to claim 9, and auxiliary request 5, with respect to claim 12, add the new feature that the speed of a first read of the magnetic medium portion is recorded in association with the fingerprint for later compensating for differences between speeds in subsequent reads. The main support for this feature can be found, according to the appellant, in the embodiment disclosed in the WO-application on page 23, line 30 to page 25, line 5.

According to this text portion, the speed at which a credit card moves past the read head can be measured and used for adjusting the sampling rate to match that speed of the magnetic medium. However, this does not mean that the speed should be recorded on the magnetic medium in association with the fingerprint.

The text in the cited portion also indicates that two transitions or fiduciary marks are placed on the card which are a fixed distance D apart. The time it takes for the card to be pulled from the first to the second transition "defines the velocity that the card is being pulled through the reader" (loc.cit.). Even when it is assumed that this distance D and the time are recorded on the card it still makes a difference to recording the "speed (…) in association with said fingerprint".

It might be an absolutely obvious next step to record the speed itself or to calculate the speed from the distance and the time used, but obvious does not mean the direct and unambiguous disclosure of the feature.

For these reasons, the main request and auxiliary requests 2, 2A, 2B, and 5, in respect of claim 9 and claim 12, respectively, are considered not to comply with Article 123(2) EPC.

It follows that none of the main request and auxiliary requests 1 to 5 is allowable.

Auxiliary request 6

5. The claims of auxiliary request 6 and the invention related thereto meet the requirements of the EPC.

5.1 The amendments of the claims do not add new subject-matter to the application as filed. In particular, the features added in connection with the definition of the fingerprint as taken from "a region () of said magnetic medium portion () located between two successive ones of said magnetic transitions" forming a data signal recorded on the magnetic medium are supported by figures 24 to 26 and the accompanying text.

5.2 The amended claims comply with the requirements of Article 84 and Rule 29 EPC, except for a number of inconsistencies relative to the present content of the description. The description has still to be amended in order to assure full conformance with the claims.

5.3 Moreover, taking into account the experimental results reported on in the research paper cited on page 2, lines 28 ff. of the WO-publication, the Board sees no reasons for doubts that the invention to which the amended claims relate is disclosed in manner sufficiently clear and complete to be carried out by the skilled person (Article 83 EPC).

5.4 Turning next to patentability of the invention, it is first to be noted that the priority does not fully cover the subject-matter of the new claims: Although the priority document indicates as an alternative that "the fingerprint can be obtained from the region between two recorded magnetic transitions" (see priority document page 7, lines 23 to 25), it does neither contain a text portion nor a figure from which the precise location for the fingerprint, namely "between two successive ones of magnetic transitions" of the data signal, can be derived.

Therefore, the Board considers document D5 as prior art in respect of the subject-matter of the auxiliary request 6.

5.5 Auxiliary request 6 seeks protection for a device and a method for fingerprinting an object (independent claims 1 and 6), for a device and a method for authenticating an object (independent claims 4 and 5), and for an object having its fingerprint recorded thereon for the later verification of its identity (independent claim 2), all these claims refer to the precise location from which the fingerprint is derived. This common feature defines directly the structure of the object and steps of the methods, which are the subject-matters of independent claims 2, 5 and 6. But it also defines, although implicitly, the functionality to be implemented in the devices of claims 1 and 4.

This functionality is distinguished from a mere suitability for detecting the noise signal from the microstructure of the magnetic medium (see point 3.5 supra) since the respective means of the claimed devices must be arranged to detect the two successive ones of the magnetic transitions, to locate the magnetic media portion between said transitions, to determine the remanent noise resulting from this small intermediate region with a resolution sufficient for a significant identification signal and to record said signal for fingerprinting and authenticating (in this context see also decision T 170/02-Controlling access/NEWS DATACOM, not published in OJ EPO; points 8 to 10 of the reasons).

In order to assess novelty and inventive step with regard to all of the claimed categories of the invention, it is sufficient to consider said common feature concerning the precise location from which the fingerprint is derived.

5.6 None of the prior art documents discloses the detection and processing of a fingerprint from a location between successive ones of magnetic transitions of the recorded data signal.

The closest prior art, document D1, clearly indicates the importance of the precise location from which the fingerprint is taken (see for example column 11, line 52 to column 12, line 6) and suggests as an alternative to use a start mark or indicia on the object at which the signal recording can be started. However, this indicia is not part of the data signal recorded on the magnetic medium. Neither does the document mention a stop mark, nor a recording length which fits between two successive ones of the magnetic transitions of the recorded data signal.

Even if the enhancing signal in document D1, which is indeed formed of magnetic signal transitions (see signal S in figure 3C), is considered as a data signal recorded on the object, the difference remains in respect of the claimed invention that the fingerprint is taken at the peak positions (51). The signal portions (52, 53) in-between are not used for fingerprinting.

Document D5 clearly discloses the use of the remanent noise from the microstructure of the magnetic medium for fingerprinting but does not address the question from which location, relative to a data signal recorded on the magnetic medium, the fingerprint should be taken.

The document US-A-4 837 426 (published in 1989), although closely related to document D1, does not go beyond what is disclosed in document D1. The rest of the prior art documents cited against the application are even more remote from the claimed invention.

Therefore, novelty is acknowledged with respect to all claims of auxiliary request 6.

5.7 According to the practice of the boards of appeal the examination of inventive step follows the so-called problem-solution approach. The Board considers documents D1 and D5 as equally appropriate starting

points for assessing inventive step in the light of the prior art. It follows from decision T 967/97 - Chipkarte/OVD KINEGRAM AG (not published in OJ EPO), that in case of a positive judgement regarding inventive step all fairly promising possibilities should be explored in applying the approach.

5.8 Regarding first document D1 as the closest prior art:

The substantial difference between the claimed invention and document D1 resides in the precise location from which the fingerprint is taken, i.e. the region between two successive ones of the magnetic transitions of the recorded data signal, as already pointed out above.

According to document D1, this location can be determined in a variety of ways: physically, or electronically, e.g. by writing the enhancing signal in a predetermined frame (see column 19, lines 3 to 14).

The claimed invention uses a region between successive magnetic transitions of a data signal recorded on the magnetic medium. In substance, therefore, the problem solved by invention is merely to provide an alternative location of the region from which the fingerprint is to be taken.

There is however no hint given in the prior art to the alternative proposed by the invention.

Document D1 itself rather discourages the skilled person to think of the recorded data signal as the

right place for locating the fingerprint since it uses, as a preferred embodiment, an enhancing signal consuming a considerable length of the magnetic medium and which is thus too long as to fit into a bit cell or between two successive magnetic transitions of the recorded data signal.

5.9 Starting from document D5 as closest prior art, the situation is similar.

Document D5 expressly indicates that the remanent noise resulting from the magnetic microstructure is reproducible and can be used as fingerprint information. It points to the only small size of the region required for taking the fingerprint (see for example page 4095, left-hand column, "Magnetic Fingerprinting", second paragraph).

Nevertheless, the document does not give any concrete information regarding the placement of the fingerprint in relation to the data signal recorded. It rather reports results for waveforms having lengths between 85 micrometers and 4300 micrometers and proposes as optimum a length between 430 and 4300 micrometers. Waveforms of such a length would certainly not fit into a bit cell of a standard data signal as normally recorded on a disk.

5.10 The rest of the prior art documents cited against the application is even more remote than documents D1 and D5. It is not necessary, therefore, to go in any further detail regarding those documents.

5.11 The Board concludes that the invention to which the claims of auxiliary request 6 relate meets the requirement of inventive step (Article 52(1) and 56 EPC).