Patent Application
20100063750
The invention relates to automated inspection systems, and more particularly, to systems for inspecting of moving webs.
Inspection systems for the analysis of moving web materials have proven critical to modern manufacturing operations. Industries as varied as metal fabrication, paper, non-woven materials, and films rely on these inspection systems for both product certification and online process monitoring.
Products created on web process lines are subject to anomalies or defects from many sources. One particular concern is web line-induced anomalies, such as those created by continuously rotating equipment contacting the web in a regular, repeating pattern. Such equipment can generally be described as a “roll.” Typical rolls utilized within a web manufacturing line include but are not limited to casting wheels, pull rolls, nip rolls, microreplicating rolls, web cleaning components, and idlers.
For example, the surface of a roll may be damaged (e.g., scratched) or a may have a contaminant (e.g., dirt or other particle) that induces an anomaly or defect in the moving web carried by the roll. Moreover, the roll can cause so-called “repeating anomalies” in that a new anomaly may be imparted into the moving web with each rotation of the roll. On the resulting web product, these anomalies repeat at a distance equal to the roll's circumference in the same cross-direction or “cross-web” position. Web process lines may have hundreds of rolls, many of which may have similar diameters. Identifying the specific offending roll that induced a repeating anomaly or defect within the web can be difficult with conventional inspection systems.
For example, commercially available web inspection systems provide identification of repeating defects, including cross-web position and down-web repeat distance. However, these systems typically require a priori knowledge of existing roll diameters on a given process line in order to extract repeating defect information from the entire data stream. Moreover, in many cases there may be many idlers or other rolls within a given web process line with circumferences that are near a given repeat distance of a repeating anomaly, making defect-causing roll identification difficult. As one example, a length orienter on a film making line may have numerous rolls (e.g., twelve or more), all of nominally the same eight-inch diameter. It is often difficult to determine the unique defect-causing roll using traditional methods due in part to slight variations in diameter of each of these rolls. In addition, conventional systems are often unable to account for any spatial distortion (e.g., stretching) of the web between the defect-causing roll and the web inspection system. Further, undocumented roll changes to a web process line can also occur. For example, a six-inch diameter roll may be replaced by a five-inch diameter roll and may begin introducing repeat defects. Process operators using conventional web inspection systems might not check the changed roll as the source of anomalies or defects due to the change not being documented and the assumed diameter of the roll being incorrect.
In general, this application describes techniques for the automated inspection of moving webs. More specifically, the techniques described herein enable an automated inspection system to distinguish between anomalies that occur repeatedly from an identifiable source and random anomalies for which a source may not be determinable. Certain elements of a web manufacturing line may introduce repeated anomalies or defects into the web. For example, idler rollers, generally referred to herein as “rolls,” that support the web as it traverses the system may introduce repeated anomalies into the web at regular intervals. In accordance with the techniques described herein, the automated inspection system may identify these repeated anomalies within the web and determine the source of the anomalies. This may permit operators of the manufacturing line to locate the anomaly-causing element to repair or replace the offending element. Although explained for exemplary purposes in reference to anomalies, the techniques described herein may be readily applied to defects, and the term anomalies is used herein to mean potential or actual defects.
As described herein, the web inspection system identifies positions of anomalies or defects within the web and then correlates those positions with roll synchronization signals that were received during the manufacturing of the web. For example, each roll of interest for a web manufacturing process is equipped with a synchronization mark. During manufacturing of the web, the web inspection system receives a roll synchronization signal from each of the rolls indicating that the respective roll has completed a full rotation. The web inspection system records the position of each occurrence of these synchronization marks with respect to its downweb positional coordinate system. The web inspection system then correlates positional data of the roll synchronization signals with positional data for the anomalies or defects.
In one embodiment, the invention is directed to a method comprising receiving roll synchronization signals from a plurality of sensors of a web manufacturing system, wherein each of the sensors corresponds to a different roller of the web manufacturing system, and wherein each of the roll synchronization signals indicates that the corresponding roller has completed a full rotation during manufacturing of a web of material. The method further comprises receiving anomaly data from a web inspection system that identifies positions of anomalies on the web. The method further comprises identifying a set of two or more of the anomalies as repeated, identifying which of the rollers caused the repeated anomalies by correlating the positions of the repeated anomalies with the roll synchronization signals, and outputting an identification of the offending one of the rollers.
In another embodiment, the invention is directed to a system comprising a plurality of rollers in contact with a web of material, wherein two or more of the rollers each include a synchronization mark to indicate when the corresponding roller has completed a full rotation. The system includes a plurality of synchronization mark readers that read the synchronization marks of the plurality of rollers and output roll synchronization signals. Each of the roll synchronization signals indicates that the corresponding roller has completed a full rotation during manufacturing of the web. The system also includes an encoder on at least one of the rollers that outputs a position signal indicative of a down-web distance of the web, and an inspection system that inspects the web and outputs anomaly data identifying positions of anomalies on the web. A synchronization unit receives the position signal from the encoder and the plurality of roll synchronization signals from the synchronization mark readers, and converts the occurrence of each of the roll synchronization signals into down-web positions within a coordinate system associated with web process line. An analysis computer processes the anomaly data to identify a set of two or more of the anomalies as repeated anomalies. The analysis computer outputs an indication of which of the rollers caused the repeated anomalies by correlating the positions of the repeated anomalies with the down-web positions of the roll synchronization signals.
In another embodiment, the invention is directed to a computer-readable storage medium containing software instructions. The instructions cause a programmable processor of a computer to execute the software instructions and perform at least some of the functions set forth herein.
The techniques described herein may provide several advantages. For example, the techniques may achieve significant accuracy improvement over conventional systems. For example, the techniques can be used to easily differentiate roll sizes that differ less than 25 μm. This allows an offending roll to be identified from a group of rolls of similar diameters, thereby enabling simpler and more robust manufacturing process maintenance. Further, the techniques allow repeated anomalies or defects on a web to be detected even amidst a large number of random defects. In addition, the techniques allow the system to measure the exact crossweb and circumferential position of the defect producing area of a roll, and even differentiate between multiple repeating defects at the same crossweb position.
The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.
In general, web rolls 7, 10 may contain manufactured web material that may be any sheet-like material having a fixed dimension in one direction and either a predetermined or indeterminate length in the orthogonal direction. Examples of web materials include, but are not limited to, metals, paper, wovens, non-wovens, glass, polymeric films, flexible circuits or combinations thereof. Metals may include such materials as steel or aluminum. Wovens generally include various fabrics. Non-wovens include materials, such as paper, filter media, or insulating material. Films include, for example, clear and opaque polymeric films including laminates and coated films.
In order to produce a finished web roll 10 that is ready for conversion into individual sheets for incorporation into products 12, unfinished web rolls 7 may need to undergo processing from multiple process lines either within one web manufacturing plant, for instance, web manufacturing plant 6A, or within multiple manufacturing plants. For each process, a web roll is typically used as a source roll from which the web is fed into the manufacturing process. After each process, the web is typically collected again into a web roll 7 and moved to a different product line or shipped to a different manufacturing plant, where it is then unrolled, processed, and again collected into a roll. This process is repeated until ultimately a finished web roll 10 is produced.
An anomaly introduced into a web roll 7 by one plant, for example, web manufacturing plant 6A, may be detectable once plant 6A has finished its processes on web roll 7, but the anomaly may later become undetectable after another web manufacturing plant, such as web manufacturing plant 6B, has performed its manufacturing processes on web roll 7.
For many applications, the web materials for each of web rolls 7 may have numerous coatings applied at one or more production lines of one or more web manufacturing plants 6. The coating is generally applied to an exposed surface of either a base web material, in the case of the first manufacturing process, or a previously applied coating in the case of a subsequent manufacturing process. Examples of coatings include adhesives, hardcoats, low adhesion backside coatings, metalized coatings, neutral density coatings, electrically conductive or nonconductive coatings, or combinations thereof. A given coating may be applied to only a portion of the web material or may fully cover the exposed surface of the web material. Further, the web materials may be patterned or unpatterned.
During each manufacturing process for a given one of web rolls 7, one or more inspection systems acquire anomaly information for the web. For example, as illustrated in
In general, conversion control system 4 applies one or more defect detection algorithms that may be application-specific, i.e., specific to products 12, to select and generate a conversion plan for each web roll 10. A certain anomaly may result in a defect in one product, for instance product 12A, whereas the anomaly may not cause a defect in a different product, for instance, product 12B. Each conversion plan represents defined instructions for processing a corresponding finished web roll 10 for creating products 12, which may ultimately be sold to customers 14.
Certain elements of the process lines within web manufacturing plants 6 may introduce repeated anomalies or defects into the web. For example, “rolls” that engage the web as it traverses the process line may introduce repeated anomalies into the web at regular intervals. Example rolls utilized within a web process line include casting wheels, pull rolls, nip rolls, microreplicating rolls, web cleaning components, and idler rollers. In accordance with the techniques described herein, automated inspection systems, either located within manufacturing plants 6 or remote, identify these repeated anomalies within the web and determine the source roll that induced the repeated anomalies. This permits operators to locate the anomaly-causing element of the system and to repair or replace the offending element.
As described in further detail below, each of a web inspection system identifies positions of anomalies or defects within the web and then correlates those positions with roll synchronization signals that were received during the manufacturing of the web. For example, each roll of interest for a given web manufacturing process of manufacturing plants 6 may be equipped with a synchronization mark. During manufacturing of the web, the web inspection system receives a roll synchronization signal from each of the rolls indicating that the respective roll has completed a full rotation. The web inspection system records the occurrence of these synchronization marks. The web inspection system then converts the occurrence of each of the roll synchronization signals into the spatial domain of the inspection system for correlation with positional data for the anomalies or defects.
The techniques described herein may provide several advantages. For example, the techniques may achieve significant accuracy improvement over conventional systems. For example, the techniques can be used to easily differentiate roll sizes that differ less than 25 μm. This allows an offending roll to be identified from a group of rolls of similar diameters. Further, the techniques allow repeated anomalies or defects on a web to be detected even amidst a large number of random defects. In addition, the techniques allow the system to measure the exact crossweb and circumferential position of the defect producing area of a roll, and even differentiate between multiple repeating defects on the same roll or at the same crossweb position.
Further, in some cases anomalies often appear the same to conventional inspection systems regardless of whether the anomaly occurs on the top side of a web or on the bottom side of a web. However, it is often desirous to know on which side of the web defects occur because, for example, anomalies on the one side of the web, say the bottom, may be healed by coatings on subsequent processes, but anomalies on the top side will still be visible after the final manufacturing operation. Thus, by determining the causal roll for a particular repeating anomaly, the inspection system can determine which side of the web an anomaly is on by storing data specifying the side (i.e., top or bottom) on which each roller is located and correlating each repeated anomaly to an individual roller in an automated manner. Data can be output indicating the side of the roller causing the anomaly by displaying and indication to a user, storing the data in a database or communicating the data to other electronic systems or devices.
The inspection system described herein may be further configured to automatically disregard repeated anomalies on the bottom side of the web without alerting the operator, while immediately alerting for defects on the top side. Alternatively, such anomalies on the bottom of the web may be designated as at a lower alert or warning level. Thus another potential advantage of the techniques described herein may be efficiently detecting and reporting of anomalies of varying degrees of importance.
Image acquisition devices 26 may be conventional imaging devices that are capable of reading a sequential portion of the moving web 20 and providing output in the form of a digital data stream. As shown in
The image may be optionally acquired through the utilization of optic assemblies that assist in the procurement of the image. The assemblies may be either part of a camera, or may be separate from the camera. Optic assemblies utilize reflected light, transmitted light, or transflected light during the imaging process. Reflected light, for example, is often suitable for the detection of defects caused by web surface deformations, such as surface scratches.
In some embodiments, fiducial mark controller 30 controls fiducial mark reader 29 to collect roll and position information from web 20. For example, fiducial mark controller 30 may include one or more photo-optic sensors for reading bar codes or other indicia from web 20. In addition, fiducial mark controller 30 may receive position signals from one or more high-precision encoders engaged with web 20 and/or rollers 22, 24. Based on the position signals, fiducial mark controller 30 determines position information for each detected fiducial mark. Fiducial mark controller 30 communicates the roll and position information to analysis computer 28. Techniques for applying and using fiducial marks to identify specific locations on a web are described in co-pending patent application “Apparatus and Method for the Automated Marking on Webs of Material” to Floeder et al., assigned to the assignee of the present application Ser. No. 10/826,995, filed Apr. 19, 2004, the entire contents of which are hereby incorporated by reference. Although discussed with respect to fiducial marks and a fiducial mark controller 30 and reader 29, fiducial marks may not be necessary in all embodiments to effect the techniques described herein. In other embodiments, other means may be used to determine locations of anomalies and other information on a web without departing from the techniques described herein.
Analysis computer 28 processes image streams from acquisition computers 27. Analysis computer 28 processes the digital information with one or more initial algorithms to generate local anomaly information that identifies any regions of web 20 containing anomalies that may ultimately qualify as defects. For each identified anomaly, analysis computer 28 extracts from the image data an anomaly image that contains pixel data encompassing the anomaly and possibly a surrounding portion of web 20. Analysis computer 28 may classify an anomaly into different defect classes if necessary. For instance, there may be unique defect classes to distinguish between spots, scratches, and oil drips. Other classes may distinguish between further types of defects. In accordance with the techniques described herein, analysis computer 28 may further determine in which of products 12 an anomaly may cause a defect.
Based on the position data produced by fiducial mark controller 30, analysis computer 28 determines the spatial position of each anomaly within the coordinate system of the process line. That is, based on the position data from fiducial mark controller 30, analysis computer 28 determines the x, y, and possibly z position for each anomaly within the coordinate system used by the current process line. For example, a coordinate system may be defined such that the x dimension represents a distance across web 20, a y dimension represents a distance along a length of the web, an the z dimension represents a height of the web, which may be based on the number of coatings, materials or other layers previously applied to the web. Moreover, an origin for the x, y, z coordinate system may be defined at a physical location within the process line, and is typically associated with an initial feed placement of the web 20.
In any case, analysis computer 28 records in database 32 the spatial location of each anomaly with respect to the coordinate system of the process line, this information being referred to herein as local anomaly information. That is, analysis computer 28 stores the local anomaly information for web 20, including roll information for the web 20 and position information for each anomaly, within database 32. Analysis computer 28 may also record, for each anomaly, those products of products 12 for which the anomaly may cause a defect. Database 32 may be implemented in any of a number of different forms including a data storage file or one or more database management systems (DBMS) executing on one or more database servers. The database management systems may be, for example, a relational (RDBMS), hierarchical (HDBMS), multidimensional (MDBMS), object oriented (ODBMS or OODBMS) or object relational (ORDBMS) database management system. As one example, database 32 is implemented as a relational database provided by SQL Server™ from Microsoft Corporation.
Once the process has ended, analysis computer 28 transmits the data collected in database 32 to conversion control system 4 via network 9. Specifically, analysis computer 28 communicates the roll information as well as the local anomaly information and respective sub-images to conversion control system 4 for subsequent, offline, detailed analysis. For example, the information may be communicated by way of database synchronization between database 32 and conversion control system 4. In some embodiments, conversion control system 4 may determine those products of products 12 for which each anomaly may cause a defect, rather than analysis computer 28. Once data for the finished web roll 10 has been collected in database 32, the data may be used to mark anomalies on the web roll, either directly on the surface of the web with a removable or washable mark, or on a cover sheet that may be applied to the web before or during marking of anomalies on the web.
Manufacturing system 40 produces web 44, typically by pulling a substrate from lead roller 41 through manufacturing components 48A-48M (manufacturing components 48) to produce web 44 that is collected onto web roller 42. Accordingly, web 44 may traverse web manufacturing components 48, which may manufacture web 44 in various ways. For example, one of manufacturing components 48, e.g. manufacturing component 48A, may apply a coating to web 44.
Idler rollers 46A-46N (idler rollers 46) provide support for web 44 as web 44 traverses web manufacturing system 40. That is, web 44 may rest upon idler rollers 46 while undergoing manufacturing from manufacturing components 48. Although idler rollers 46 may be necessary to properly position web 44, idler rollers 46 may impart anomalies or defects into web 44. For example, one or more of idler rollers 46 may scratch the bottom side of web 44. Although discussed with respect to idler rollers 46, other types of roll, such as casting wheels, pull rolls, nip rolls, microreplicating rolls, or web cleaning components, may be present in web manufacturing system 40, in addition to or in lieu of idler rollers 46. Thus the techniques described herein are not limited to use with idler rollers, but can be applied to any roll of interest within the web process line. The use of idler rollers is merely exemplary for the purpose of demonstration.
The techniques explained herein identify positions of anomalies or defects within the web and correlate those positions with roll synchronization signals. For example, each roll of interest for a web manufacturing process 40 may be equipped with a respective synchronization mark 47A-47N. Further, synchronization mark readers 50A-50N (synchronization mark readers 50) are associated with each one of the rolls of interest (each one of idler rollers 46 in this example) for sensing the respective synchronization mark. Each of synchronization mark readers 50 may detect when the corresponding one of idler rollers 46 has made a full rotation and then emit a roll synchronization signal in the form of a trigger pulse, which remote synchronization unit 54 detects. That is, each of synchronization mark readers 50 may output a short pulse upon a complete rotation of the respective one of rollers 46, and the leading edge of each short pulse may indicate the complete rotation has been detected. In one embodiment, each of synchronization mark readers 50 may be a photo-optic sensor. For example, readers 50 may be from the D10 Family of sensors from Banner Engineering Corp. In general, readers 50 detect corresponding synchronization marks 47 as the marks rotate past the reader. In the example embodiment, synchronization marks 47 may be a target such as a retro-reflecting material or a machined section of the roll. Upon detecting reference point synchronization marks 47 on a corresponding one of rollers 46, the one of readers 50 outputs the synchronization mark signal. Therefore, each of readers 50 outputs a discrete signal for each rotation of the corresponding one of rollers 46.
To aid converting the roll synchronization signals into a spatial domain of a coordinate system associated with web process line 40, a rotational encoder is affixed to one or more rolls along the process line. In this example, rotational encoder 52 is affixed to web roller 41. In other embodiments, an encoder may be used with one or more of rollers 46 in lieu of, or in addition to, encoder 52. Encoder 52, in one embodiment, may be a sine encoder based position sensor. Other embodiments may utilize other types of position sensors or encoders. In general, encoder 52 outputs an electrical pulse train that is directly synchronized to the physical movement of web roller 41. For example, encoder 52 may emit a series of pulses for each rotation of web roller 41. In one embodiment, for example, encoder 52 may emit four million pulses per rotation, thus providing a high degree of positional accuracy.
Remote synchronization unit 54 receives the positional pulses from encoder 52 and the roll synchronization signals from synchronization mark readers 50 and generates a logical map that identifies various sections of web 44 that align with each of idler rollers 46. For example, for each of the rollers, remote synchronization unit 54 divides the spatial domain of the web into a series of sections, each of the sections being as long as the circumference of respective roller. Each web section corresponding to idler roller 46A, for example, is 18.85 inches, i.e. 6.00 inches*π. Each web section corresponding to idler roller 46B is 18.91 inches, and a web section that corresponds to idler roller 46C is 18.79 inches. In this way, remote synchronization unit 54 uses the positional data from encoder 52 as well as roll synchronization signals from synchronization mark readers 50 to convert the roll synchronization signals into the spatial domain of the coordinate system for process 40 for determining web sections within the spatial domain for each or the rollers of interest. As a result, remote synchronization unit 54 need not require a priori data regarding the exact diameter of each of rollers 46 in order to determine the web sections and ultimate detect repeated defects.
In some cases, some or all of the rolls of interest may have approximately the same diameter. For example, a subset or all of idler rollers 46 may have approximately the same diameter of six inches. However, this subset of idler rollers 46 typically does not have exactly the same diameter due to manufacturing variability. For example, the diameter of idler roller 46A may be 6.01 inches, the diameter of idler roller 46B may be 6.02 inches, and the diameter of idler roller 46C may be 5.98 inches. The techniques described leverages averaging captured by calculating variations in the relative offset between a repeated defect and the corresponding roll synchronization signal for a given roller. This provides precise accuracy that allows for repeat defect detection even in manufacturing lines having substantially similar sized rollers but for manufacturing variability in the rollers themselves.
In order to associate an anomaly with one of idler rollers 46, an inspection system may first collect data regarding web 44. Using the pulses from encoder 52 and the roll synchronization signals from synchronization mark readers 50 that has been collected and correlated by remote synchronization unit 54, the inspection system analyzes the anomaly data for the identified web sections for each of the rollers. The inspection system may average the results of the data over many instances of these web sections. For example, in one embodiment, the inspection system may collect 100 instances of web segment data for a given roller. The inspection system then analyzes the data to attempt to distinguish between repeated anomalies and random anomalies. The inspection system may determine that an anomaly is a repeated anomaly, caused by one of idler rollers 46 for example, if an anomaly occurs in a majority of the instances of analyzed web sections for a given roller at or relatively near the same position in those instances in which the anomaly occurs. For example, if idler roller 46A causes an anomaly in web 44, the anomaly will probably be repeated, and the instances of the repeated anomaly should occur approximately 18.85 inches apart, given a diameter of roller 46A of 6.00 inches.
In some arrangements, at least some of the anomalies imparted to web 44 by idler rollers 46 may be healed, i.e. corrected, by the time web 44 is ready to be converted into sheets. In other words, although idler rollers 46 may impart an anomaly into web 44, the anomaly may not cause a defect because the anomaly may be corrected through other manufacturing processes before web 44 is ready to be converted. For example, anomalies imparted to web 44 by idler rollers 46 will be on the bottom side of web 44. Anomalies occurring on the top of web 44 may not be healed or corrected in web 44. That is, anomalies occurring on the top surface of web 44 may cause defects in products 12 if a web segment or individual sheet containing such anomalies is converted into one of products 12. In accordance with the techniques described herein, an inspection system may be able to determine whether an anomaly occurred on the top side or bottom side of web 44. Moreover, the inspection system may be able to trace the source of anomalies occurring on the top side to a particular one of idler rollers 46, for example, idler roller 46A. Accordingly, an operator of manufacturing system 40 may locate the portion of idler roller 46A that caused the anomalies and repair idler roller 46A.
In general, example remote synchronization unit 54 senses the occurrence of each roll synchronization signal (illustrated in
In the example embodiment, remote synchronization unit 54 includes counters 56A-56N (“counters 56”) and registers 58A-58N (“registers 58”). Each of synchronization mark readers 50 is associated with one of counters 56, which is in turn associated with one of registers 58. The pulse signal from encoder 52 is used as a global increment driving counters 56. That is, as encoder 52 detects web movement, encoder 52 sends a series of pulses that are used to simultaneously increment each of counters 56. In the exemplary embodiment of
The roll synchronization signals from each of the rollers are used as triggers for recording the value within the rollers' respective counters. Specifically, during a full rotation of any of rollers 46, the corresponding synchronization mark 47 of that roller will pass the associated synchronization mark reader 50. For example, for each rotation of roller 46A, synchronization mark reader 50A will detect synchronization mark 47A. Upon detecting mark 47A, synchronization mark reader 50A outputs a roll synchronization signal to remote synchronization unit 54 in the form of a short pulse. In response to this pulse, remote synchronization unit 54 latches the current value of the corresponding counter, in this case, counter 56A, into the corresponding data register, register 58A.
Controller 60 polls each of registers 58 at a high rate or is interrupt driven to retrieve the most recent counter data. Accordingly, the polling cycle of controller 60 is faster than the rotations of all of rollers 46. If, upon polling one of registers 58, e.g. register 58A, the counter data is the same as the previous poll, controller 60 may ignore the current counter data. However, if the counter data has changed, controller 60 may retrieve the counter data and transmit the counter data, along with the roller number, to analysis computer 59 (
In one embodiment, analysis computer 59 may be a server-class computer. In other embodiments, analysis computer 59 may be a distributed computing system or other computing system capable of handling the high amounts of data required for processing the inspection and position information.
As described above, controller 60 of remote synchronization unit 54 transmits roll position data upon detecting a rotation of one of the rollers 46, and the roll position data may be in the form of a roller identification (i.e., a trigger number) and the current encoder position recorded representing the downweb position (DW position) for a given complete rotation of that roller. In some embodiments, encoder 52 may transmit positional pulses both to remote synchronization unit 54 and to the inspection systems to allow correlation within the spatial domain of the web segments of the rolls and detected anomalies. In other embodiments, two distinct encoders may be used to provide positional reference information that is reconciled by analysis computer 59. In still other embodiments, a different means of tracking distance down the web, such as fiducial marks, may be employed by the inspection system. Techniques for using fiducial marks with a web are discussed in co-pending patent application “Fiducial Marking for Multi-Unit Process Spatial Synchronization,” Ser. No. 11/828,376, to Floeder et al., assigned to the assignee of the present application, filed Jul. 26, 2007, the entire contents of which are hereby incorporated by reference.
In any case, analysis computer 59 correlates the roll position data from remote synchronization unit 54 with positional data of anomalies on the web as determined by the inspection system. Video or other image data may be passed from the inspection sensors to acquisition computers 62A-62M (“acquisition computers 62”). These computers represent software and/or hardware capable of acquiring and processing inspection data for detection of various types of anomalies on the web, e.g. scratches, spots, drips, spills, or other types of anomalies. For example, acquisition computers 62 may be software modules executing on analysis computer 59 or analysis computer 29 of
Analysis computer 59 processes the anomaly data and the roll position data to determine whether any anomalies repeatedly occur at substantially the same cross-web position and substantially the same downweb offset within multiple web segments for the same roller. For example, if one of rollers 46 causes a repeated anomaly, the repeated anomaly occurs at substantially the same cross-web position and will repeat with a spacing of the circumference of the corresponding roller, i.e. the circumference of the roller causing the repeated anomaly. In this manner, analysis computer 59 may determine that repeated anomalies are occurring. Moreover, correlating the downweb positions of the repeated anomalies with the downweb positions of the web segments for the different rollers, analysis computer 59 is able to identify which of rollers 46 is causing each of the repeated anomalies.
Encoder 52 and synchronization mark readers 50 create a series of pulses that graphically depict the position of each of rollers 46 over time along the downweb length of web 67. Encoder pulses from encoder 52 and synchronization pulses from synchronization mark readers 50 are represented signals 76 and graphs 78A-78N (“graphs 78”), respectively. Based on the data, roll position data, analysis computer 59 determines the number of encoder pulses from encoder 52 that occur between synchronization pulses from one of synchronization mark readers 50. In the example of
Analysis computer 59 determines that anomalies 64A-64D (“anomalies 64”) occur at a similar cross-web position. Analysis computer further determines that one of anomalies 64 occurs one encoder pulse after each synchronization pulse from roller 46C. That is, in this example the downweb positions of the anomalies are constant offsets from the start of new web segments for roller 46C. Therefore, analysis computer 59 determines that anomalies 64 are repeated anomalies and that roller 46C is the cause. An operator may then inspect roller 46C at the cross-web position of repeated anomalies 64 and either repair or replace roller 46C.
Similarly, a set of anomalies 66A-66D (“anomalies 66”) all occur at the same cross-web position. However, there are missing anomalies 68A and 68B that were expected to occur. It is possible that the offending roller did not cause an anomaly, or that the inspection system failed to detect an anomaly at one or both of positions 68A and 68B. In either case, however, analysis computer 59 may still determine the presence of a repeated anomaly. This is because, even with missing anomalies 68A and 68B, analysis computer 59 determines the presence of a repeated anomaly when a majority of the new web segments for a roller contain anomalies in the same cross-web position and the substantially the same distance from the synchronization pulses, i.e., the start of a new web segment for that roller. In this case, each of repeated anomalies 66 occur 7 encoder pulses after a majority of synchronization pulses of signal 78A. Therefore, analysis computer 59 may determine that roller 46A is causing a repeated anomaly.
The techniques described herein may even be used to detect repeated anomalies 70A-70G (“repeated anomalies 70”) and repeated anomalies 72A-72G (“repeated anomalies 72”) and to distinguish repeated anomalies 70 from repeated anomalies 72. Repeated anomalies 70 and repeated anomalies 72 each occur at the same cross-web position. Repeated anomalies 70 each occur 1 encoder pulse after a synchronization pulse of graph 78B and 4 encoder pulses after a synchronization pulse of graph 78D. Repeated anomalies 72 each occur 7 encoder pulses after a synchronization pulse of graph 78B and 1 encoder pulse after a synchronization pulse of graph 78D. Although it would appear that either of rollers 46B or 46D could be causing either of these repeated defects, analysis computer 59 may still determine which of repeated defects 70 and 72 are caused by rollers 46B and 46D, because the diameters of rollers 46B and 46D likely differ by some detectable amount. For ease of visualization and readability, a small number of encoder pulses is shown in the example of
Because of this segmentation, it is possible to determine whether, for example, idler roller 46A is causing any of the anomalies on web 44. Web segment 82A includes anomalies 84A, 86A, 88A, 90, and 92. Web segment 82B includes anomalies 84B, 86B, 88B, and 94. Web segment 82C includes anomalies 84C, 86C, 88C, 96, and 98. Web segment 82D includes anomalies 84D, 86D, 88D, and 80. To determine whether any of these anomalies is a repeated anomaly caused by idler roller 46A, the analysis computer determines the distance between each anomaly and each synchronization pulse, i.e. the beginning of each web segment as represented by each dashed line. Although only four web segments 82 are shown in
The analysis computer repeats this analysis for each of the rollers. That is, the analysis computer segments the web in a similar manner for each synchronization pulse, allowing the computer to identify the specific source of the repeated anomalies.
In composite map 110, anomalies 84, 86, and 88 each appear to be repeated anomalies. However, composite map 110 shows that repeated anomalies 84 shift in the down-web direction during different web segments. That is, anomalies 84 and 88 may be repeated anomalies, but they are not repeated at the interval of the circumference of idler roller 46A. Analysis computer 59 may determine this by determining that the distance from the synchronization pulse for this specific roller to each of anomalies 84 and 88 exceeds a threshold difference for each instance of anomalies 84 and 88.
In contrast, anomalies 86 are repeated anomalies and caused by the roller for which the data has been segmented because, as shown by composite map 110, they are spaced at substantially the same interval of the circumference of idler roller 46A. That is, for each instance of anomalies 86, the distance between the synchronization pulse and the instance of anomalies 86 is within a tolerance distance. The tolerance distance may, for example, be ±2 pulses depending on the positional resolution of the encoder. Therefore, the inspection system may determine that anomalies 86 are repeated anomalies caused by idler roller 46A. For example, anomalies 86 may be scratches on the bottom side of web 80 caused by a rough spot on idler roller 46A. Using this determination, an operator may attempt to repair idler roller 46A at this position to prevent idler roller 46A from causing more anomalies.
In some embodiments, the inspection system may be reprogrammed to disregard anomalies occurring at a similar position in later web segments, as these anomalies will very likely not actually cause a defect once web 80 is finally converted into products. That is, nearly all of anomalies caused by idlers or other rollers known to be located on a bottom side of the web may be cured at some point during the manufacturing of web 80.
The random anomalies 90, 92, 94, 96, 98, and 80 of web 80, however, probably occurred on the top side of web 80, and anomalies 90, 92, 94, 96, 98, and 80 will probably not be cured during the remainder of the manufacturing of web 80. Therefore, the inspection system may mark the positions of these anomalies in a database, such as database 32 of
Analysis computer 59 also receives roller data (122). The roller data may include identifications of each roller, as well as data characterizing the occurrences of complete rotations of each roller. For example, the roller data may identify roller 46A using a unique identifier or label, which may be assigned by a user, and include trigger numbers (e.g., a sequence number) and a down web position for each instance when synchronization mark reader 50A read synchronization mark 47A.
Next, analysis computer 59 correlates the received anomaly data with the received roller data (124). Initially, analysis computer 59 processes the roller data to logically partition the web into a series of segments, and may repartition the web in a similar manner for each roller of interest. That is, for each roller of interest, the length of each segment in the series is defined by the distance between two sequential trigger signals from its corresponding one of synchronization mark readers 50. As a result, the length of each of the segments for that partitioning is substantially equal to the circumference of the corresponding one of rollers 46. For example, analysis computer 59 may logically partition the web into a set of segments for roller 46A. The down web distance between signals from synchronization mark reader 50A with respect to a coordinate system of the process line will, accordingly, be equal to the circumference of roller 46A. As described in greater detail below, the anomaly data for each of these logical segments for a given roll of interest may be analyzed to determine the presence of anomalies in substantially common positions within the segments, that is, anomalies that occur at a common cross-web location and a common down-web distance from the beginning of all or a threshold number of the logical segment. This threshold, in one embodiment, may be a majority of the segments. In some embodiments, the width of each segment may be the width of the web. In other embodiments, such as that described with respect to
Based on the logical partitioning of the web for each of the rollers of interest, analysis computer 59 identifies positions of anomalies on each of the segments. In this manner, analysis computer 59 determines the positions of each anomaly relative to each rotation of each roller. Analysis computer 59 then analyzes the anomaly data (126) to determine the presence of repeating anomalies (128). Analysis computer 59 determines, for each roller, whether an anomaly is occurring in substantially the same position relative to the rotation of the roller. That is, analysis computer 59 determines whether any of the anomalies is in substantially the same position on the logical segments for any of the rollers. For example, analysis computer 59 may determine that an anomaly occurs 16 inches cross-web and five inches down-web for all or for a threshold number of the segments for a given partitioning.
By determining the presence of a repeating anomaly, analysis computer 59 may then identify the anomaly-causing roller of rollers 46 (130). For example, analysis computer 59 may determine that roller 46A is causing a repeated anomaly because the anomaly at issue occurs substantially the same cross web and down web location after each rotation of roller 46A. In response, analysis computer 59 may output the identity of the anomaly-causing roller (132). For example, analysis computer 59 may output the identity to a computer screen. Other means of identifying the anomaly-causing roller may also be used, such as, for example, causing a light on or near the offending roller to illuminate. As another example, analysis computer could illuminate a light-emitting diode (LED) that is associated with the offending roller, wherein each roller is associated with an LED and the LEDs may be positioned on a board to provide a central viewing location to an operator. Additionally, analysis computer 59 may further output the position of the anomaly to assist an operator in repair of the offending roller. The operator may determine the position of the synchronization mark on the roller and, using the position of the anomaly, inspect the roller at the position of the repeated anomaly to determine whether the repeated anomaly-causing element is repairable.
In the example embodiment of
Image acquisition devices, such as image acquisition devices 26 of
Because a repeating anomaly may occur in an overlap of lanes, such as in the overlapping region between lanes 154A and 154B, such an anomaly may be detected twice. The inspection system may use various factors to reconcile such a duplicate detection. For example, the inspection system may compare the cross-web position of the repeated anomalies, as well as the down-web position of each instance of the anomalies and the repeat interval between instances. When two repeated anomalies are discovered with the same cross-web position and instances of the anomalies occurring at the same down-web positions at the same interval, the system may discard one of the repeating anomalies so as not to trigger two alerts for the same repeating anomaly.
Initially, analysis computer 59, in an example embodiment, determines a starting point A, which may be a first detected anomaly (160). As discussed above, a repeating anomaly is an anomaly that is caused by an element of a web production or manufacturing system, such as an idler roller. Therefore, there is a certain distance, herein referred to as “Rmin” which is the minimum possible repeating distance for a repeated anomaly. For example, in the case of repeated anomalies caused by one or more of a plurality of idler rollers used within a web process, Rmin is the circumference of the smallest idler roller of interest. Accordingly, analysis computer 59 may search for a point B in lane 154A such that the cross-web position of points A and B are the same and the down-web distance between points A and B is at least Rmin (162).
Analysis computer 59 may then determine whether a point C exists in lane 154A such that the cross-web position of point C is the same as that of A and B and such that the down-web distance between points B and C is a certain multiple of the distance between points A and B (164). A repeated anomaly may not repeat in every expected instance. Several instances of the repeated anomaly may be skipped, as discussed with respect to
If no point C can be found at a distance from point B that is, for example, 1, ½, ⅓, 2, or 3 times the distance between points A and B (“NO” branch of 166), analysis computer 59 may obtain a new starting point anomaly A (168) and attempt to determine whether the new starting point is part of a repeating anomaly. If analysis computer 59 does find such a point C, however, (“YES” branch of 166), analysis computer 59 may then search for a point D where the distance between points C and D is a multiple of the distance between points A and B (170). In one embodiment, the same set of potential multiples may be used as in step 164, e.g. 1, ½, ⅓, 2, and 3. Point D may be used to confirm that points A, B, and C are indeed part of a sequence of repeating anomalies.
If no point D is found (“NO” branch of 172), analysis computer 59 may again restart the process of selecting a new starting anomaly point A (168). No point D may be found if, for example, anomalies at points A, B, and C were not part of a repeating anomaly and the distances between points A and B and between points B and C were merely coincidental. However, if analysis computer 59 does find a point D (“YES” branch of 172), it is quite likely that points A, B, C, and D make up a column of repeated anomalies. Therefore, analysis computer 59 may determine a repeat distance as the minimum of the distances between points A and B, points B and C, and points C and D (174). Analysis computer 59 may then expect to discover anomalies repeated at the determined repeat distance from point D at the cross-web position of points A, B, C, and D. Analysis computer 59 may analyze each of lanes 154 to discover repeated anomalies in a similar manner.
After having determined a repeated anomaly, analysis computer 59 may determine the source roller of the repeated anomaly, per the method of
The error tolerance applied by analysis computer 59 may be a function of the expected number of complete rotations separating the anomalies. For example, for two nearly identical rollers having diameters of 20.000 cm and 20.0001 cm, the down web distance separating two repeat intervals for the rollers will be approximately 62.800 cm and 62.803 cm, which may be too small to measure. However, after 50 expected complete rotations for the rollers, the down web positions of the end of the web segment will be 3140 cm and 3140.15 cm, yielding a positional difference of 0.15 cm, which is a measurable error tolerance applied by analysis computer 59.
As an example, the first position, i.e. the position of point A, for a repeated anomaly series may have been recorded at 0.4924 m and the nth instance of a repeated anomaly may have occurred at a down-web distance of 79.5302 m. The offset would then be 79.1008 m (79.5302 m−0.4924 m). The first synchronization mark of roller 46A (
Although discussed with respect to the use of lanes 154, the method described above is not limited to the use of lanes 154. For example, the method may be applied to a full web 152 that has not been divided into lanes 154. In another example embodiment, multiple analysis computers may be used, one analysis computer for each lane, rather than a single analysis computer 59. For example, acquisition computers 27 of
In the exemplary embodiment of
Entries in priority column 188 indicate to a user how important or significant the detected repeated anomaly is. In the example of
Entries in action description column 190 indicate to a user the suggested or required action that the user should take. For example, the first entry in description column 190 is “pull roll #3”. A user viewing this display should replace the roller identified with the number “3” with a new roller. Moreover, given a priority of “high” in priority column 188, a user should replace roll “3” as soon as possible.
Map column 192 allows a user to select a roller and view the composite map on map screen 184. For example, a user may use a mouse connected to analysis computer 59 to direct a pointer to one of the cells in column 192 and press a button to select the corresponding roller. In the example of
Slip area 183B displays information regarding whether rollers slip as the web traverses the manufacturing system. This slip may be caused, for example, when the web does not make constant contact with the roller. This may cause anomalies or defects to occur in the web when the web does come in contact with the roller. In any case, slip area 183B displays a roll identifier column 194 and a priority column 196. Roll identifier column 194 displays information that identifies the relevant roller. Priority column 196 indicates the priority, such as the severity, of the roll slippage. Again, in other embodiments, other types of priorities could be used, such as color-coded priorities or numeric priorities.
In one embodiment, analysis computer 59 may automatically sort the data displayed in data output area 182 based on priority, from highest priority to lowest, based on the values in priority columns 188 and 196. In one embodiment, analysis computer 59 may automatically populate the user interface, that is, without the need for a user to “refresh” the data. In one embodiment, data output area 182 may display between 0 and 20 entries. In one embodiment, data output area 182 may include a scroll bar, tabs, or other means by which to display a large number of entries.
Various embodiments of the invention have been described. These and other embodiments are within the scope of the following claims.
