SYSTEM FOR ARTICLE IDENTIFICATION USING MAGNETIC DATA TAG WITH RANDOMLY
DISTRIBUTED MAGNETIC ELEMENTS
FIELD OF THE INVENTION
The present invention relates to systems for article identification. For example, the invention relates to marking systems where data is stored in a marker by the nature and distribution of magnetic materials therein, and the markers are interrogated by remotely generated magnetic fields, and the magnetic response to the interrogation field is sensed with magnetic pickup coils.
BACKGROUND TO THE INVENTION
Magnetic data markers are typically formed from an assembly of magnetic elements, the exact detail of which describes the data to be encoded, fn other types, data is recorded magnetically onto a carefully assembled structure. While the magnetic materials can in principle be of low cost, there may be a significant manufacturing cost and complexity in assembling the magnetic elements reliably into the data carrying configurations.
The inventors have appreciated approaches to reducing the cost of manufacturing data markers.
List of Prior Art:
Random magnetic markers for authentication or tracking:
Multibit magnetic data tags including angle encoding but not specifically random (WO 0108085).
A method of detecting the presence of magnetic elongated particles using electromagnetic fields and harmonic detection techniques (US 5,992,741).
Verifying manufacturing operations by detecting the presence of magnetizable filaments within a critical component of a manufactured assembly by detecting a distortion to an
interrogating electromagnetic waveform caused by the magnetizable filaments (US 6,169,481).
The use of fibers coated with magnetic material (US 4,114,032) and distributed within an article to be authenticated by scanning the surface and counting the number of said fibers within a region on the article.
Authentication verification through the use of random irregularities within a layer of magnetic material, typically a semihard magnetic layer. (US 3,790,754 (claims 15, 32), US 4,806,740, US 4,837,426)
These generally involve contact or near contact measurement of the surface of the semihard magnetic layer, for the purpose of detecting variations in the saturation magnetization.
Authentication verification by forming a random distribution of essentially round particles or flakes of magnetic material (EP 0696 779 A1) or polymer elements containing magnetic metal (EP 0 625 766 A3 also US 5,602,381), reading data relating to the non- uniformity of the distribution of said elements, and thereafter recording onto the same article a magnetic digital representation of the random data.
Random magnetic markers for antitheft:
(US 4,114,032).
EAS with random fibers (US 4,857,891).
Predetermined magnetic data tags:
The earliest published patent we have found on predetermined magnetic data tags is WO 88 09979 (EP 0295085). The initial application broadly describes identification tags which can be differentiated based upon predetermined arrangements of magnetic materials, including angular arrangements, and the properties of said magnetic materials. The patents born from this application have not been maintained.
Detection technology:
Several recent patents describe marking systems comprised of magnetic data tags containing soft magnetic materials, in a pre-formed angular arrangement, which are interrogated by a detector using a rotating magnetic field (EP 0713 195, WO 0008489, GB 2349047, GB 2349048), one of which includes the ability to read more than one tag at a time (GB 2349051). Other art additionally claims the use of coercivity, amplitude and/or magnetic bias for additional data content (WO 0113321 , WO 0126049).
SUMMARY OF THE INVENTION
In a first aspect of the present invention, there is provided a system for article identification comprising a marker containing a plurality of low coercivity, high permeability ferromagnetic elements distributed in a fully or partially random manner, each element exhibiting a magnetization-response which is strongly direction- dependent, and
a reading apparatus comprising:
(a) a means for generating, within an interrogation zone, an interrogating magnetic field whose orientation can be varied in a controlled manner, and
(b) a means for sensing the response of a marker within the interrogation zone using 2 or more pairs of magnetic pickup coils, and (c) a means for recognizing a marker based, at least in part, on the relation of the magnetic response to the sensing direction.
The invention is of advantage in that the system is capable of being more cost effective and yet providing a high degree of marker identification performance.
Preferably, said means of recognizing is achieved by: (a) identifying changes in the magnetization state of some or all of the elements based on the sensed response, and
(b) associating the identified changes in magnetization state with particular elements, and
(c) determining the orientation of each element from the sensed response, and
(d) using the directions of elements so obtained to characterize the marker.
According to a second aspect of the present invention, there is provided a system for article identification comprised of:
a marker containing a plurality of low coercivity, high permeability ferromagnetic elements distributed so as to be substantially contained within a common marker plane yet either fully or partially random within said marker plane, each element exhibiting a magnetization-response which is strongly direction-dependent and, a reading apparatus comprising:
(a) a means for generating, within an interrogation zone, a interrogating magnetic field whose orientation can be varied in a controlled manner, and
(b) a means for sensing the response of a marker within the interrogation zone using 2 or more pairs of magnetic pickup coils, and
(c) a means for recognizing said marker plane based upon the relation of the magnetic response to the sensing direction, and (d) a means of rotating the interrogation field vector substantially within the marker plane, and (e) a means of sensing the in-plane magnetic response of said ferromagnetic elements to the directional variation of the magnetic field within the marker plane using 2 or more pairs of magnetic pickup coils, and (f) a means of recognizing a marker based, at least in part, on the dependence of the magnetic response on the sensing direction within the marker plane.
The present invention is concerned with a type of tag which requires much less control in manufacture, thus enabling simpler, faster and lower cost manufacture of magnetic data tags. This simplification in tag construction is made possible by an advanced detector technology which can sense the magnetic response of the marker in detail, particularly as regards the angular dependence of the response.
Markers of the invention contain a random or partially random distribution of magnetic elements, as may be obtained by sprinkling the elements onto a surface and securing them in place. Reading systems of the invention interrogate the markers with a rotating magnetic field, and sense the magnetic response of the marker as a function of the angle of the interrogation field. In one embodiment, the reading system recognizes from the magnetic response the presence of a plurality of magnetic elements of a particular type, which are distributed at different angles within the plane of the marker. In another embodiment, detailed angular distribution of the magnetic response of the marker is recorded, and used to subsequently identify the marker as unique from other markers with different angular magnetic response.
These random tag techniques can be applied to most types of passive magnetic data tags, including tags containing detectable elements made from soft magnetic materials, including amorphous metal materials, where said elements are in the form of ribbons, wires, fibers, foils,- or combinations thereof. In cases where some magnetic elements are well characterized as magnetically dipolar, the angles between the magnetic dipole axes can be determined from the detected angular magnetic response. In cases where the elements are non-dipolar, such as fibers which are bent in the tag plane, the angular magnetic response can still be measured, and used to characterize the marker.
DESCRIPTION OF THE DIAGRAMS
Embodiments of the invention will now be described, by way of example only, with reference to the following drawings in which:
Figure 1 : Magnetic data tags (1) containing soft magnetic materials distributed in random or partially random manner.
A) straight fibers of substantially constant length (2) and random angle and position B) straight fibers of random length (3), random -angle and random position, C) straight ribbons of substantially constant length (4), random angle and random position,
D) straight and non-straight fibers of random length (5) , random angle and random position
Figure 2: Magnetic data tags (1) containing soft magnetic materials distributed in a random or partially random manner, plus one or more soft magnetic elements in a predetermined location and orientation.
A) straight fibers of substantially constant length (2) and random angle and position plus one ribbon of predetermined length (6), position and orientation B) straight fibers of random length (3), random angle and random position plus 2 ribbons each of predetermined lengths (6), positions and orientations
C) straight ribbons of substantially constant length (4), random angle and random position, plus one piece of foil of predetermined size (7), position and orientation
D) straight and non-straight fibers of random length (5), random angle and partially random position plus one ribbon of predetermined length (6), position and orientation
Figure 3: Reading apparatus with drive coils for generating a magnetic interrogation field which can rotate in 3-dimensions, plus pick-up coils for sensing the magnetic response of a marker within the interrogation zone.
Figure 4: Reading apparatus with drive coils for generating a magnetic interrogation field which can rotate principally in 2 dimensions, plus pick-up coils for sensing the magnetic response of a marker within the interrogation zone.
Figure 5: An example of the magnetic response from a marker as in Figure 1A, from which the angles between fibers may be determined.
Figure 6: An example of the magnetic response from a marker as in Figure 2A, from which the angles between each fiber and the ribbon on the marker may be determined.
The peak amplitudes of some fiber signals (601) and the ribbon signal (602) are indicated.
Figure 7: An example of the magnetic response from a marker as in Figure 1 B, from which the angles between fibers may be determined.
Figure 8: An example of the characteristic magnetic response from a marker as in Figure 1 D, showing peaks of width and distribution as may be used to characterize the magnetic response.
DESCRIPTION OF EMBODIMENTS OF THE INVENTION
Random magnetic tags
Markers of the current invention are formed of a substrate onto which soft magnetic materials are distributed in a random or partially random manner. It will be understood throughout that a key feature of these soft magnetic elements is that they are low coercivity, high permeability ferromagnetic elements, and that they exhibit a magnetization-response which is strongly direction-dependent.
Typical markers of the invention are shown in Figure 1. Such tags can be conveniently formed by sprinkling lengths of soft magnetic fiber (2,3,5) or ribbon (4) onto a web of paper or polymer film (1) which has been coated with pressure sensitive or other adhesive, and thereafter laminating a covering film or paper on top to bond the magnetic elements fixedly in place. During the fiber distribution process, physical masks may be used to restrict the magnetic elements to localized regions of the web, such that upon subsequent cutting of individual tags, it is not necessary to cut the fibers. This permits a more uniform distribution of fiber lengths in the final markers. Alternately, the fibers may be applied randomly across a large area of the web. In each case, the density of random elements is chosen to provide a suitable average number of elements (2,3,4,5) per marker, typically in the range of 5 to 50. For a given detector resolution, the average number of fibers per tag can be optimized to produce the smallest fraction of duplicate tags. For the 3-D detector described herein, this fraction can typically be 10 or fewer duplicate tags per million. Punching, die cutting, shear cutting or other techniques may be used to form individual markers, which may be formed with another layer of adhesive on a release liner for ease of handling and application.
It will be noted that by using more than one differentiable type of magnetic element, such as 2 types of magnetic fibers with different coercivities or hysteresis shapes, distributed in a random or partially random manner, a greater variety of distinguishable markers can be formed.
In addition to those elements (2,3,4,5) distributed in a random or partially random manner, one or more predetermined soft magnetic elements may be included on the marker as reference elements (6,7) (Figure 2). These reference elements (6,7) can be distinguished from the random elements (2,3,4,5) by the reading system based upon features of the sensed magnetic response including amplitude, shape, coercivity or combinations thereof. For example, for a marker containing a random collection of 20 mm long, 30 micrometer diameter amorphous magnetic fibers, a 35mm long, 25 micron thick by 0.7 millimeter wide strip of amorphous magnetic material can form a reference element which could be distinguished from the fibers on the basis of peak amplitude (601 , 602). This reference element can be used to establish a baseline for angular measurements of the magnetic response of the marker, greatly simplifying the process of coding and searching for a matching magnetic response. The reference element can also be used to sense the presence of a marker within the interrogation zone for the purpose of triggering a tag measurement. As another example, two amorphous magnetic strips are located at 45 degrees to one another within the marker plane, to form a more identifiable reference, and to provide a baseline for angular measurements of the random fibers. As a third example, a length of permalloy wire can be used as a reference element, differentiable from the randomly distributed fiber on the basis of the size, shape of magnetic response and the coercivity as deduced by the reading apparatus.
Such reference elements can be added to the markers during manufacture by laying one or more strips or wires of soft magnetic material onto the web from which the markers are to be cut.
Another format of marker can be formed where a predetermined number of known magnetic elements are distributed in a random manner within the marker. The marker
data is simply the number of soft magnetic elements, as can be determined by the reader apparatus.
Other techniques may be used for making magnetic data tags, whereby the magnetic elements are formed directly onto a web by patterned electroplating or electroless plating, where the patterning is achieved by photolithography or by printing process.
Because a substantial number of independent magnetic elements can be distributed quickly and effectively in a bulk process, the manufacture of random magnetic tags offers cost savings over more controlled manufacturing processes. Because tag cost can be the dominant expense in implementing a data tag solution, the low tag cost is a significant advantage. In addition, the randomness of the process makes it very difficult to artificially re-create any one of the random or partially random markers, and even more difficult to re-create a sequence thereof. This feature provides strong anticounterfeit protection for security applications.
Note that the system performance can be enhanced by presorting the tags. One possible presort is to reject unsuitable tags. This may be achieved by interrogating each tag after manufacture and discarding any tags which do not generate a suitable number of well-defined individual peaks in the magnetic response. In other applications, it may be desirable search a database of tags within a given manufacturing lot and to additionally discard tags which the system considers to be a close match or duplicate to another tag within the lot.
Alternate presorting schemes are possible wherein tags are interrogated and those which generate a magnetic response containing a number of clearly defined peaks above a threshold, not more than some maximum number and not less than some minimum number are selected and physically separated into bins based upon the number of such peaks, the ratios of angular spacings between some or all the peaks, and/or other criteria. Tags from each bin can subsequently be applied to groups of articles, using one bin per group, such that the articles can subsequently be unambiguously identified with their group by reading the tag and interpreting the magnetic response using the same or similar criteria as was used to separate the«.bins of tags.
Reader features and descriptions
A detailed description of the reader technology is provided in WO 9935610, and only a survey will be provided here.
Readers of the invention have been made in different geometries, including a 2-D reader and a single-sided reader which can read a tag which has been oriented within a predefined reading plane near the reader, and a 3-D aperture reader which can read a tag placed at any orientation within the interrogation volume. Key to reading the data from the tags is that the soft magnetic elements have a low coercivity, high permeability and highly directional magnetic response. When placed within the interrogation volume, dipolar elements of this description remain near magnetic saturation at all times except when the interrogation field rotates past the angle perpendicular to the dipolar axis, at which point the direction of magnetization of the element abruptly reverses. This flux reversal is sensed by the pick-up coil sets and may be used to determine the orientation of the element, in addition to the relative magnetic moment, coercivity and hysteresis shape. A 20cm aperture 3-D reader has been constructed on this basis, with an intrinsic angular resolution of less than 4 degrees.
Figures 3A and 3B illustrate the receiver coil arrangement for the 3-D reader. Figure 3 A illustrates the three sets of orthogonal coils used to couple with the tag magnetic elements within the interrogation zone. For the y-direction, 143, the receiver coil set comprises 4 coils, 144, 145, 146,and 147. Inner coils 145 and 146 extend along the x- direction as illustrated by dimension 148. The outer coils 144 and 146 are wound on a second co-axial former. The coils are connected in series in _the electrical sense illustrated and 'balanced' by small mechanical re-alignments to achieve zero sensitivity to a uniform magnetic field. A second receiver coil set as illustrated is sensitive to tag- generated field in the z-direction, 149. This coil set is identical to the coils 144, 145, 146 and 147, rotated through 90 degrees as shown. The third coil sensitive to tag-generated field in the x-directioπ, 150, comprises two solenoid coils, 151 and 152. The inner coil 151 is wound on the first former, 142, and the outer coil 152 is wound on the second coaxial former. Figure 38 illustrates all the coils wound on the inner former, 142, and the outer former, 153.
Figure 3C illustrates the three orthogonal transmit coils configuration. The coils are wound on a cylindrical former, 154. A uniform magnetic field in the y-directioπ, 155, is produced by four coils, 156, 157, 158 and 159. Coils 156 and 158 comprise a 'modified Helmholz' arrangement. Coils 157 and 159 comprise a second modified 'Helmholz' arrangement, with a magnetic axis 25 degrees offset from coils 157 and 159. The two 'modified Helmholz' coils sets have magnetic axes 12.5 degrees on either side of the y- direction, 155. The four coils are connected in series in the sense illustrated. A second transmit coil set generates a uniform field in the z-direction, 160. This set comprises four identical coils oriented in an orthogonal direction as illustrated. The final transmitter coil consists of a long solenoid coil, 162, wound on the coil former, 136. This generates uniform field in the x-direction.
Figure 3D illustrates the 3-D Reader antenna. The transmit coils on former 154 are located co-axially with receiver coil tube, 142. The interrogation volume, 6, is defined by a further 190mm ID co-axial tube (not shown) that is used to define a mechanical constraint on possible tag positioning on the antenna.
Figure 4A illustrates the components comprising the single-sided reader antenna. The antenna comprises two receiver antennae and two orthogonal transmit antennae. The transmit antennae comprise two orthogonal coils, 109 and 1 10, wound on a mu-metal or ferrite plate, 1 11. The coils, 109 and 1 10, are wound orthogonally as shown. The receive antennae comprises two identical printed circuit boards (PCB), 112 and 1 13, mounted on either side of the mu-metal/coil assembly.
Figure 4B illustrates the PCB's copper patterns. Each PCB is double sided with copper track patterns on both sides, 114 and 115. On one side, 1 14, a 'figure of 8' loop, comprising turns 116 and 1 17 connected in the sense illustrated, 118, forms a quadropole. The loops, 116 and 1 17, are connected to have an emf induced by an a.c. magnetic field in the direction, 119. The PCB side 115, comprises turns 120 and 121 , identical to side 1 14, but the pattern is rotated through 90 degrees. The 'figure of 8' antenna comprising loops 116 and 1 17 on side 1 14 on one PCB, 112, is connected to the identical loops on the second PCB, 1 13. The connection sense are in such a manner that the that the generated field from either transmit antenna, 109 or 1 10,
induces zero net voltage in that receive antenna channel, comprising coils on either side of the transmit coils. The receive antenna 'figure of 8' configuration (loops 116 and 117) provides rejection of uniform H-field interference (from far field sources). It can be seen that the loops 1 16 and 117 are sensitive to a magnetic dipole above the board in the direction 119, whilst being insensitive to a magnetic dipole in an orthogonal direction. Loops 120 and 121 are not sensitive to a magnetic dipole in the direction 1 19.
Tags are placed with their magnetic elements parallel to the plane of the antenna defined by the PCB plane, 112.
Figures 5 through 8 illustrate the response of various tag types when interrogated by the reading apparatus. For markers containing a number of dipolar elements within a plane, the angles between the dipole elements can be established and used to characterize the marker (Fig. 5). For markers which contain magnetic elements of various sizes and magnetic moments, successive thresholds can be used to select only the more prominent features of the magnetic response for coding, searching and matching (Fig. 6,7).
For markers containing elements which are not purely dipolar, the element orientation is less well defined, and cannot be measured to the same accuracy. Such markers can still be successfully characterized however, by rotating the interrogation field within the marker plane and recording the response detected perpendicular to the interrogation field. A typical response is shown in Figure 8, for a marker containing a distribution of both straight and non-straight amorphous fibers, showing a spectrum of peaks and valleys, the detailed shape being unique to that marker. Thus, a reproducible magnetic signature can be recorded to characterize markers which contain elements which are not purely dipolar.
When the marker plane is controlled, rotating the interrogation field within the marker plane is straightforward. In the 3D aperture reader, where the marker may be at any orientation, more complex interrogation field trajectories are possible. For example, a spiral scan may be used to sweep the magnetic field vector over a broad range of angles in 3 dimensions. The magnetic response from this spiral scan can be used to determine the plane of the tag as the plane common to all the detected elements. With the tag
plane identified, a second interrogation scan may then be performed within the tag plane to more accurately establish the angles between the dipole elements or to unambiguously record the magnetic signature from a marker which contains non-dipolar elements.
Data handling methods
While the use of random elements can offer a significant simplification and cost saving to the manufacture of data tags, the requirements for extracting the data are more onerous. One method is to measure each marker after manufacture, but prior to use, and to store a coded version of the magnetic signature in a computer database for subsequent searching and matching. During use, tags are interrogated by a compatible reader system and the magnetic response so obtained is compared with those in the database, in search of the best match.
Efficient storage, searching and matching of signatures from random markers requires compression of the data into a form which retains the key information. This may conveniently be achieved by recording the position, width, amplitude and confidence of each identifiable peak in the response, although other methods may also be used. The parameterization of each peak in the magnetic response can be achieved by any of a number of known numerical analysis techniques.
To try to improve distinguishing between random tags in the presence of measurement variations, two options are available: "single peak suppression" and "multi peak suppression". These attempt to counteract the tendency of low peaks to sometimes be measured below the threshold and sometimes above it. "Single peak suppression" will try and match the current tag to tags in the database assuming there is one peak too many or too few in the current measurement. All permutations of missing peak will be tried. "Multi peak suppression" will try and match tags when more than one peak is missing or extra. Only the smallest peaks will be looked at in this process. The level to which peaks will be looked at is an adjustable parameter.
The searching may be accelerated by performing by a series of incrementally more stringent pre-searches, beginning with relatively crude test conditions such as the
number of peaks above various thresholds or the angular positions or gaps there between, and culminating in one or more correlation calculations between the compressed magnetic response from the currently interrogated marker and the most likely candidates from the database pre-searches. The database may additionally be organized to store entries according to key features of the pre-searches, to reduce the searching time. The correlation calculation can be designed to provide greater emphasis on the angular positions of large peaks, with less weight given to smaller peaks. Similarly, the angular position information can be given greater weight than the peak amplitude or width information. The confidence factor is given approximately by the ratio of the best match correlation to the sum of correlations from all the tags in the database (ignoring negative values). Hence, if the best match correlation is much bigger than any of the others, the confidence of having successfully found the correct tag is high. If the best correlation is only slightly higher than correlations from other tags in the database, the confidence will be low. Even greater confidence and error tolerance may be achieved if a final correlation calculation is performed on the uncompressed data array of the magnetic response of the current marker and the most likely candidate matches from the database. Such correlation calculations are widely known.
Specific system applications
Mobile communication solution
Printed data representing random magnetic data to be verified on board detector
Example application 1.
Self-adhesive labels are manufactured to contain a random distribution of amorphous metal fibers manufactured by MXT Inc., where the alloy composition and fiber production parameters have been controlled to give a uniquely narrow and square hysteresis shape. The tags are 35mm square and contain an average of 3mg of MXT fiber with an average fiber length of 25mm. The tags may additionally contain visual or optically verifiable authentication features such as holograms or diffractively verifiable images. When tags are interrogated in the reading apparatus, the magnetic response contains several peaks identifiable as unique to this fiber, distributed at different angles within the
tag. When the manufacturer integrates these tags into the packaging of brand name retail goods, the authenticity of manufacture can be checked, without even opening the packaging, using one of the detectors disclosed herein.
Example application 2.
Markers of dimensions 35mm x 35mm are manufactured to contain a random distribution of straight amorphous metal fibers manufactured by MXT Inc. with an average loading of 5mg per marker and an average length of 25mm, plus a 35mm length of ribbon of amorphous metal manufactured by Vacuumschmelze Corporation of cross section 25micrometers by 700micrometers. These markers are formed with pressure sensitive adhesive on one side, and they are spaced apart at 70mm intervals on a release liner of width 50mm. Rolls of the markers are compatible with automated label applicators. After fabrication, each roll is passed through an inspecting station where a single sided reader of the sort described herein interrogates it. The reader verifies the presence of the amorphous ribbon and measures the angular magnetic response of the ribbon plus fibers. Tags which the reader deems could give unreliable or ambiguous future readings are removed from the roll and discarded. The entire magnetic response is recorded in a database as a one-dimensional array, along with a compressed version thereof. The compressed data is deduced from a mathematical model which is applied to the magnetic response to find the best fitting linear combination of individual fiber peaks to describe the data. The peak positions, amplitudes and confidence factors form the compressed data. The data may be pre-sorted in the database by a method which will speed subsequent searches. Many such methods are well known in the art.
The tags are applied to manufactured goods or their packaging, and a database of product information linked with tag data is maintained. At any point thereafter in the product life, the goods may be interrogated by one of the readers described herein for purposes of tracking, tracing, or verifying the authenticity of the product. This may be achieved by comparing the magnetic response and/or the compressed form thereof with the computer database of pre-measured markers, possibly facilitated by wireless communication of the data with a remote computer, using the data search techniques described herein. The best match of tag magnetic signature is used to identify the product, and the confidence of the match is a measure of authenticity.
Example application 3
A simplification to Example application 2 may avoid the use of the database by measuring the magnetic response of the manufactured tag and translating the measured magnetic response into a compressed data set and printing an optical bar code onto the tag which represents an encoded form of that compressed magnetic data. Field interrogation requires reading the magnetic response and the optical bar code and verifying that the 2 are correlated through the secure encryption algorithm used to form the optical bar code.
It will be appreciated that modifications can be made. o embodiments of invention described in the foregoing without departing from the scope of the invention.