1. Field of the Invention
Embodiments of the present invention generally describe achieving fault tolerance in an automated high productivity factory for batch array work-piece handling and processing sized for producing 1,000 or more work-pieces an hour and as high as 40,000 per hour or more. High volume factories require predictable production over some calendar period and require a method of fault tolerance.
2. Description of the Related Art
Solar energy from the sun may be converted to electricity by utilizing a solar power technology called photovoltaics (PV) that uses solar cells tiled into modules. Solar cells produce direct current electricity from the sun's rays, which can be used to power equipment, to recharge batteries, or be converted to AC for on-grid applications.
Increased productivity for manufacturing of PV cells and modules requires batch processing of multiple solar cell work-pieces simultaneously if supply is to meet customer demand. To produce the PV cells and modules, numerous processes may need to be performed upon a work-piece. The work-piece may thus need to be moved from one processing tool to another. A processing tool may comprise one or more chambers coupled together. For example, a processing tool that performs a vacuum based process may comprise one or more processing chambers and one or more load lock chambers coupled together. For a non-vacuum process such as metrology, the processing tool may comprise one or more metrology chambers.
FABs, sometimes referred to as factories, may be set up to arrange all of the necessary processing tools that are needed to process a work-piece into a single processing line. A work-piece exits the processing line after the necessary processes have occurred as a finished product. To increase the output of a FAB, numerous substantially identical processing lines may be present.
At any given point in time, it is not uncommon for a processing tool within a processing line to shut-down for repairs. When any processing tool is shut-down along a processing line, the processing line may shut-down and thus produce no product during the shut-down period. The shut-down thus affects the product output.
For example, suppose a processing tool has a mean time between failures (MTBF) of 320 hours, then a FAB having 50 tools will, on average, have 1 tool shut-down in any 8-hour period. Therefore, when a processing tool is shut-down for an 8-hour period, the processing line containing the shut-down tool is also shut-down for the 8-hour period and thus produces no product. Hence, the processing line has a fault. When a processing line produces no product, the total output of the FAB, as well as the average throughput over a specified period of time, will decrease. Fault tolerance is where a FAB would be able to compensate for a fault while maintaining the desired FAB throughput. It would be beneficial to overcome faults in FABs and still maintain a desired FAB throughput.
Therefore, there is a need in the art for achieving fault tolerance in a photovoltaic FAB.
The present invention generally comprises a method for achieving fault tolerance in a PV FAB which comprises equipment for an automated high volume work-piece manufacturing architecture consisting of work-piece handling and work-piece processing organized in a regular fashion from a group of lines comprising parallel channels. For descriptive purposes, factory architecture supports a river of work-pieces comprising streams (lines) which are further sub-divided into channels. Channels may operate in a continuous conveyor in some cases and in segmented piece-wise continuous batches in others. The batch array may be 1 or 2 dimensions, (i.e., 1×n or n×m work-pieces).
The work-pieces may be transported or presented to the equipment from a stacked orientation supply to a parallel orientation array of channels comprising a stream. Additionally, the work-pieces may be transferred between manufacturing architecture entities by an array to array batch transfer of channels. The work-pieces may be transferred within the manufacturing architecture in an array to array batch transfer operation as opposed to one work-piece at a time. The robotic operations on the streams of work-pieces may be between robotic devices, between robotic devices and processing equipment, and within processing equipment. In general, fault tolerance requires a factory operating policy which implements capability redundancy and reserves some capacity for managing production during equipment shutdown due to failure, repair or scheduled maintenance.
Whenever a processing tool within any processing line is shut-down, rather than shut-down the entire processing line containing the shut-down processing tool, work-piece flow channels may be routed around the shut-down processing tool by transferring the work-pieces to an adjacent processing line in the FAB if additional capacity is available.
In one embodiment, a method of achieving fault tolerance for a photovoltaic FAB is disclosed. The method comprises disposing a batch of work-pieces into a FAB, the FAB comprising a plurality of substantially identical processing lines, each processing line having a plurality of processing tools coupled together, the FAB further comprising one or more buffer or stocker stations between each processing line at a location adjacent to each processing tool, batch processing the work-pieces within the individual processing lines without transferring work-pieces between the processing lines, the batch processing occurring at a predetermined optimum throughput that is below a maximum throughput, detecting a shut-down at one or more processing tools in a first processing line of the plurality of substantially identical processing lines, batch array transferring a plurality of work-pieces between processing lines at the one or more buffer or stocker stations at a location before the shut-down tool of the first processing line, batch array receiving the plurality of transferred work-pieces into a second processing line, and increasing throughput for the second processing line to a level above the optimum level in order to achieve fault tolerance and maintain a substantially constant throughput averaged over a predetermined period of time.
In another embodiment, a method of achieving fault tolerance for a photovoltaic FAB is disclosed. The method comprises determining a minimum economic unit for a processing line within the FAB to achieve an optimum throughout in the FAB, the processing line having a plurality of processing tools coupled together, coupling a plurality of processing lines together within the FAB, each processing line having the minimum economic unit, the plurality of processing lines coupled together at one or more of the processing tools, detecting a shut-down of one or more tools within one or more processing lines, and compensating for the shut-down of the one or more tools to achieve fault tolerance. The compensating comprises batch array transferring work-pieces between the plurality of processing lines at a first buffer or stocker station disposed prior to the shut-down of the one or more tools adjacent processing lines, batch array transferring the work-pieces between the plurality of processing lines at a second buffer or stocker station disposed after the shut-down of the one or more tools, and increasing throughput through the plurality of processing lines above the optimum throughput such that the average throughput for the FAB over a predetermined period of time remains substantially constant.
In yet another embodiment, a photovoltaic FAB having fault tolerance capabilities is disclosed. The FAB comprises a plurality of substantially identical processing lines disposed within the FAB, each processing line comprising a plurality of processing tools, one or more buffer or stocker stations disposed between adjacent processing lines at locations between each processing tool, and one or more batch array work-piece transfer devices, the devices capable of retrieving a plurality of work-pieces into a processing tool, and rotating and extending to the one or more buffer or stocker stations disposed between processing lines.
So that the manner in which the above recited features of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to embodiments, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only typical embodiments of this invention and are therefore not to be considered limiting of its scope, for the invention may admit to other equally effective embodiments.
To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements and features of one embodiment may be beneficially incorporated in other embodiments without further recitation.
The present invention generally comprises a method for achieving fault tolerance in a PV FAB.
The work-pieces may be transferred between processing tools along a processing line or stream by an array to array transfer whereby an array of work-pieces may be transferred from one processing tool to another processing tool as an array rather than individually transferring the work-pieces one at a time. The various processing tools may include one or more metrology tools. The processing tool may be arranged in a flow through manner whereby the processing tools are arranged in a linear fashion, a flow-by arrangement whereby the processing tools are arranged in a non-linear fashion, or a combination of flow through and flow-by arrangements.
Whenever a processing tool within any processing line is shut-down, rather than shut-down the entire processing line containing the shut-down processing tool, work-pieces may be routed around the shut-down processing tool by transferring the work-pieces to an adjacent processing line within the FAB at an interchange node. The plurality of work-pieces may be transferred to other processing lines or streams through buffer or stocker stations. A buffer station may permit transfer between adjacent processing lines or streams while a stocker station may permit transfer between non-adjacent processing lines or streams. The buffer stations may additionally used to store work-pieces while waiting to be disposed into the next processing tool. At a location after the shut-down processing tool, the work-pieces may be transferred back to the processing line containing the shut-down processing tool through buffer or stocker stations. After the processing within the processing line or stream is completed, the work-pieces may be transferred from an array arrangement back to a stack arrangement. During the time period that the processing tool is shut-down, the other processing lines within the FAB may increase their throughput in order to maintain a substantially constant optimum throughput for the FAB over a given period of time.
Each processing line 102, 104, 106, 108 may comprise one or more processing tools 118 coupled together. The processing tools 118 may comprise one or more chambers 116. The processing tools 118 may be coupled together in the processing lines 102, 104, 106, 108 such that the optimum throughput of each processing line 102, 104, 106, 108 may be achieved at the minimum economic unit. The minimum economic unit means the greatest efficiency possible taking everything into account such as work-piece sequencing, processing time, transfer times, facility space, etc. The chambers 116 may comprise load lock chambers, preheat chambers, and processing chambers such as chemical vapor deposition (CVD) chambers, physical vapor deposition (PVD) chambers, cleaning chambers, etching chambers, inspection chambers, implantation chambers, etc.
Work-pieces may enter the processing tool 118 at an entrance position 110 and exit the processing tool 118 at an exit location 112. If necessary, the work-pieces may be temporarily stored at a buffer or stocker station 114. The buffer or stocker station 114 may be used to store the work-pieces between processing tools 118 whenever one processing tool 118 has a different processing time compared to the adjacent processing tool 118. It is to be understood that whenever one processing tool 118 has a longer processing time than an adjacent processing tool 118, the processing tool 118 may comprise a plurality of chambers 116 so that the faster processing tool 118 may have a minimum amount of down-time waiting for work-pieces.
For example, if one processing tool 118 takes two times as long as the adjacent processing tool 118, then by increasing the number of slow processing tools 118, the amount of work-pieces processed by the combined slow processing tools 118 is doubled and each slow processing tool 118 may process a batch of work-pieces. However, the faster processing tool 118 may then sequentially process a batch of work-pieces from each slow processing tool 118 such that when the faster processing tool 118 has completed both batches of work-pieces, two more batches of work-pieces are ready to be transferred from the slower processing tools 118. If only one slow processing tool 118 were present, then the faster processing tool would have a period of non-operation. Thus, by increasing the number of slow processing tools 118, the faster processing tool 118 may continuously run.
The processing tools 118 are not limited to the arrangement shown in
When the work-pieces are transferred between processing tools 118 along a processing line 102, 104, 106, 108, the work-pieces may be extracted from the processing tool 118 at an exit location 112 by an end effector that may extend into the processing tool as shown by arrows C. An end effector may comprise a device at the end of a robotic arm, designed to interact with an environment. The exact nature of the end effector depends on the application of the robot. The end effector is, in essence, the “hand” connected to a robot's arm which may retain the photovoltaic work-pieces. The work-pieces may then be translated over to an adjacent processing tool 118 by moving the end effector 120 along a track 124 as shown by arrow B. The end effector 120 may then extend into a processing tool as shown by arrows C to insert the work-pieces into the adjacent processing tool 118.
One or more end effectors 120 may be disposed on a common track 124. Each end effector 120 may access one or more processing tool 118. In one embodiment, the end effectors 120 may translate as shown by arrows A, B along the track 124 to an adjacent processing tool 118 as shown by the end effectors 120 shown in shadow. More than one end effector 120 may access a processing tool 118 simultaneously as shown in
During normal operations, the FAB 100 may operate with the processing lines 102, 104, 106, 108 at a predetermined optimum throughput level for a predetermined period of time. The predetermined period of time may be an hour, a day, a week, a month, a year, etc. The predetermined optimum throughput level for the predetermined period of time may be equal to a value below the maximum throughput capacity for the FAB 100. Should one of the processing tools 118 within any processing line 102, 104, 106, 108 within the FAB 100 shut-down, the work-pieces may be transferred from one processing line 102, 104, 106, 108 to another processing line 102, 104, 106, 108.
Once work-pieces have been extracted from a processing tool 118 at an exit location 112, rather than translating the work-pieces over to an entrance position 110 of the adjacent processing tool 118 that is shut-down, the work-pieces may translate over to a buffer or stocker station 114 where the work-pieces may be translated over to an adjacent processing line 102, 104, 106, 108.
Because the work-pieces are transferred to an adjacent processing line 102, 104, 106, 108, the throughput for the FAB 100 may decrease and thus, the throughput for the FAB 100 over a predetermined period of time may not meet the optimum throughput level. The predetermined period of time may be an hour, a day, a week, a month, or a year. To compensate for the decreased throughput, the throughput for the processing lines 102, 104, 106, 108 that do not have a shut-down processing tool 118 may increase their work-piece throughout to a level above their optimum operating capacity up to their maximum capacity. By increasing the throughput through the processing lines 102, 104, 106, 108 that do not have a shut-down processing tool 118, the processing lines 102, 104, 106, 108 may compensate for the shut-down processing tool 118. Once the shut-down processing tool 118 is back on line, then the work-pieces may be processed within the processing tool 118 and work-pieces may not be transferred between processing lines 102, 104, 106, 108. The throughput for each processing line 102, 104, 106, 108 may thus return to optimum throughput levels for the processing lines 102, 104, 106, 108 within the FAB 100 and thus, the average throughput for the predetermined period of time for the FAB 100 may remain substantially constant and a fault tolerance, or down-time compensation, is achieved. As throughput demands for the FAB 100 increase, additional processing lines 102, 104, 106, 108 may be added to achieve the optimum throughput and fault tolerance.
The work-pieces may be extracted from the load lock chambers 214 by an end effector 224 that may extend into the unload lock chamber 214 as shown by arrows D. The end effector 224 may then rotate and insert the work-pieces into a load lock chamber 214 on an adjacent processing tool 218 as shown by arrows D. Should the processing tool 218 be shut-down, the end effector 224 may extend as shown by arrows E along a track 212 into a buffer or stocker station 222 that transfers the work-pieces between processing lines 202, 204, 206, 208. A stocker station 210 may transfer the work-pieces to processing lines 202, 204, 206, 208 that are not adjacent by an up and over transfer as discussed above. The work-pieces may thus be routed around the shut-down processing tool 218. In a manner similar to that discussed above with respect to
It is to be understood that while
By transferring work-pieces between processing lines within a FAB during a shut-down of a processing tool, the average throughput over a predetermined period of time for a FAB may be maintained. The transferring of work-pieces permits a FAB to achieve a fault tolerance and thus, maintain a substantially constant throughput over the predetermined period of time.
While the foregoing is directed to embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.
This application is a continuation of U.S. patent application Ser. No. 11/747,583, filed May 11, 2007 now U.S. Pat. No. 7,496,423, which is hereby incorporated by reference.
Number | Name | Date | Kind |
---|---|---|---|
4237598 | Williamson | Dec 1980 | A |
4668365 | Foster et al. | May 1987 | A |
4875327 | Wilde | Oct 1989 | A |
5046909 | Murdoch | Sep 1991 | A |
5174707 | Suekane et al. | Dec 1992 | A |
5183370 | Cruz | Feb 1993 | A |
5223112 | Tepman | Jun 1993 | A |
5462080 | Plavidal et al. | Oct 1995 | A |
5525024 | Freerks et al. | Jun 1996 | A |
5789878 | Kroeker et al. | Aug 1998 | A |
5810549 | Wytman | Sep 1998 | A |
5882165 | Maydan et al. | Mar 1999 | A |
5957648 | Bachrach | Sep 1999 | A |
5993141 | Wytman | Nov 1999 | A |
6016611 | White et al. | Jan 2000 | A |
6053687 | Kirkpatrick et al. | Apr 2000 | A |
6079693 | Ettinger et al. | Jun 2000 | A |
6086362 | White et al. | Jul 2000 | A |
6103055 | Maher et al. | Aug 2000 | A |
6132165 | Carducci | Oct 2000 | A |
6134482 | Iwasaki | Oct 2000 | A |
6149365 | White et al. | Nov 2000 | A |
6176668 | Kurita et al. | Jan 2001 | B1 |
6182603 | Shang et al. | Feb 2001 | B1 |
6193507 | White et al. | Feb 2001 | B1 |
6206176 | Blonigan et al. | Mar 2001 | B1 |
6213704 | White et al. | Apr 2001 | B1 |
6215897 | Beer et al. | Apr 2001 | B1 |
6235634 | White et al. | May 2001 | B1 |
6267851 | Hosokawa | Jul 2001 | B1 |
6280134 | Nering | Aug 2001 | B1 |
6286230 | White et al. | Sep 2001 | B1 |
6287386 | Perlov et al. | Sep 2001 | B1 |
6298685 | Tepman | Oct 2001 | B1 |
6308932 | Ettinger et al. | Oct 2001 | B1 |
6338161 | Ali Muhammad | Jan 2002 | B1 |
6359250 | Blonigan et al. | Mar 2002 | B1 |
6371712 | White et al. | Apr 2002 | B1 |
6406359 | Birang et al. | Jun 2002 | B1 |
6435868 | White et al. | Aug 2002 | B2 |
6460369 | Hosokawa | Oct 2002 | B2 |
6471459 | Blonigan et al. | Oct 2002 | B2 |
6477980 | White et al. | Nov 2002 | B1 |
6506693 | Heyder et al. | Jan 2003 | B2 |
6517048 | Ettinger et al. | Feb 2003 | B2 |
6517303 | White et al. | Feb 2003 | B1 |
6520839 | Gonzalez-Martin et al. | Feb 2003 | B1 |
6540466 | Bachrach | Apr 2003 | B2 |
6552297 | Blonigan et al. | Apr 2003 | B2 |
6572321 | Nulman | Jun 2003 | B1 |
6647993 | Shang et al. | Nov 2003 | B2 |
6679671 | Blonigan et al. | Jan 2004 | B2 |
6679755 | Sommer et al. | Jan 2004 | B1 |
6698991 | Bachrach et al. | Mar 2004 | B1 |
6719516 | Kroeker | Apr 2004 | B2 |
6746198 | White et al. | Jun 2004 | B2 |
6841728 | Jones et al. | Jan 2005 | B2 |
6847730 | Beer et al. | Jan 2005 | B1 |
6955197 | Elliott et al. | Oct 2005 | B2 |
7076326 | Wu et al. | Jul 2006 | B2 |
20010015074 | Hosokawa | Aug 2001 | A1 |
20010024609 | White et al. | Sep 2001 | A1 |
20010041122 | Kroeker | Nov 2001 | A1 |
20020031420 | Kroeker | Mar 2002 | A1 |
20020050581 | Ettinger et al. | May 2002 | A1 |
20020090282 | Bachrach | Jul 2002 | A1 |
20020096114 | Carducci et al. | Jul 2002 | A1 |
20020187024 | Nulman | Dec 2002 | A1 |
20030021658 | Blonigan et al. | Jan 2003 | A1 |
20030113190 | Bachrach | Jun 2003 | A1 |
20030141820 | White et al. | Jul 2003 | A1 |
20030202868 | Bachrach | Oct 2003 | A1 |
20040062633 | Rice et al. | Apr 2004 | A1 |
20040065255 | Yang et al. | Apr 2004 | A1 |
20040076496 | Elliott et al. | Apr 2004 | A1 |
20040081545 | Elliott et al. | Apr 2004 | A1 |
20040081546 | Elliott et al. | Apr 2004 | A1 |
20040191030 | Rice et al. | Sep 2004 | A1 |
20050005808 | Wakabayashi et al. | Jan 2005 | A1 |
20050081785 | Lubomirsky et al. | Apr 2005 | A1 |
20050121293 | Rice et al. | Jun 2005 | A1 |
20050145464 | Rice et al. | Jul 2005 | A1 |
20050186063 | Rice et al. | Aug 2005 | A1 |
20050209721 | Teferra et al. | Sep 2005 | A1 |
20050224315 | Rice et al. | Oct 2005 | A1 |
20050260345 | Lubomirsky et al. | Nov 2005 | A1 |
20070073430 | Govind et al. | Mar 2007 | A1 |
20070148336 | Bachrach et al. | Jun 2007 | A1 |
20070158654 | Kholodenko et al. | Jul 2007 | A1 |
20070219660 | Kaneko et al. | Sep 2007 | A1 |
Number | Date | Country | |
---|---|---|---|
20090012643 A1 | Jan 2009 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 11747583 | May 2007 | US |
Child | 12212522 | US |