Great research starts with great data.

Learn More
More >
Patent Analysis of

Optimal one-wafer scheduling of single-arm multi-cluster tools with tree-like topology

Updated Time 12 June 2019

Patent Registration Data

Publication Number

US10001773

Application Number

US14/918577

Application Date

21 October 2015

Publication Date

19 June 2018

Current Assignee

MACAU UNIVERSITY OF SCIENCE AND TECHNOLOGY

Original Assignee (Applicant)

MACAU UNIVERSITY OF SCIENCE AND TECHNOLOGY

International Classification

G05B19/41,G05B19/418

Cooperative Classification

G05B19/4187,G05B2219/45031,G05B2219/50291,Y02P90/205,Y02P90/26

Inventor

WU, NAIQI,ZHU, QINGHUA,QIAO, YAN,ZHOU, MENGCHU

Patent Images

This patent contains figures and images illustrating the invention and its embodiment.

US10001773 Optimal one-wafer scheduling single-arm 1 US10001773 Optimal one-wafer scheduling single-arm 2 US10001773 Optimal one-wafer scheduling single-arm 3
See all images <>

Abstract

The scheduling problem of a multi-cluster tool with a tree topology whose bottleneck tool is process-bound is investigated. A method for scheduling the multi-cluster tool to thereby generate an optimal one-wafer cyclic schedule for this multi-cluster tool is provided. A Petri net (PN) model is developed for the multi-cluster tool by explicitly modeling robot waiting times such that a schedule is determined by setting the robot waiting times. Based on the PN model, sufficient and necessary conditions under which a one-wafer cyclic schedule exists are derived and it is shown that an optimal one-wafer cyclic schedule can be always found. Then, efficient algorithms are given to find the optimal cycle time and its optimal schedule. Examples are used to demonstrate the scheduling method.

Read more

Claims

1. A method for controlling robots of a tree-like multi-cluster tool in wafer processing, the multi-cluster tool comprising K single-cluster tools denoted as Ci, i=1, 2, . . . , K, the single-cluster tool Ci having a robot Ri for wafer handling, the multi-cluster tool being decomposable into a plurality of cluster-tool-chains (CTCs) each having a serial topology, an individual CTC consisting of single-cluster tools Ci, Ci+1, . . . Ci′ being denoted as CTC[i, i′], processing steps performed by Ci being indexed from 0 to n[i], a first processing step of Ci having an index 0 and being a step of loading a wafer from a loadlock or a buffer module of Ci to Ri, the method comprising: determining values of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], wherein ωij is a robot waiting time for Ri to wait at a j-th step in Ci before unloading the wafer from a process module used for performing the j-th step in Ci; for any i∈{1, 2, . . . , K} and any j∈{0, 1, . . . , n[i]}, when Ri arrives at the process module for the j-th step for unloading the wafer therefrom in the j-th step performed in Ci, waiting, by Ri, for the robot waiting time of ωij; after the robot waiting time of ωij expires, unloading, by Ri, the wafer from the process module for the j-th step; after the wafer is unloaded from the process module for the j-th step, transporting, by Ri, the wafer to the process module for a next step that is either a (j+1)th step, or a zeroth step when j=n[i]; and after the wafer arrives at the process module for the next step, loading, by Ri, the wafer to the second process module; wherein the determining of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], comprises: determining candidate values of ωij for the single-cluster tools in an individual CTC by performing a CTC-Check algorithm under a constraint that Θ is equal to a lower bound of cycle time of the multi-cluster tool, the lower bound of cycle time of the multi-cluster tool being a fundamental period of a bottleneck tool among the K single-cluster tools; responsive to finding that the candidate ωij values for the individual CTC are non-negative, setting the candidate ωij values as the determined values of ωij for the individual CTC; andresponsive to finding that at least one of the candidate ωij values for the individual CTC is negative, performing: determining a minimum cycle time for the individual CTC; and determining the values of ωij for the individual CTC as the candidate values of ωij obtained by performing the CTC-Check algorithm under a revised constraint that Θ the determined minimum cycle time for the individual CTC; wherein the CTC-Check algorithm for CTC[f, l] is configured to compute the candidate values of ωij such that ωi(n[i]) is minimized for each value of i selected from N[f, l−1], and the CTC-Check algorithm comprises the steps of: setting ωij=min{Θ−(4λl+3μll(j+1)), Θ−2(n[l]+1)(λll)−Σm∈Ωj\{j}ωlm}, j∈Ωn[l]−1; setting ωl(n[l])=Θ−2(n[l]+1)(λll)−Σm=0n[l]−1ωlm; setting τl0=Θ−φl0; andsetting values of ωij's, i=l−1 to f, according to:

ωi(b[i]−1)=min{τ(i+1)0−(4λi+3μi), Θ−2(n[i]+1)(λii)};

ωij=min{Θ−(4λi+3μii(j+1)), Θ0−2(n[i]+1)(λii)−Σm∈Ωj\{j}ωim} for j∈Ωn[i]−1\{b[i]−1, n[i]};

ωi(n[i])=Θ−2(n[i]+1)(λii)−Σm=0n[i]−1ωim; and

τi0=Θ−φi0, where: b[l] is an index denoting a step of putting the wafer in Cl to a buffer module shared by Cl and Cl+1; θijij+4λi+3μii(j−1), i∈NK and j∈Nn[i]; θi0i0+4λi+3μii(n[i]), i∈NK; Θ is a cycle time of the multi-cluster tool; αij is a time taken in processing the wafer at the j-th step in Ci; τij is a wafer sojourn time at the j-th step in Ci; λi is a time taken by Ri to load or unload a wafer in Ci; μi is a time taken by Ri to move between process modules of Ci for transiting from one processing step to another; NL={1, 2, . . . , L} for a positive integer L; and ΩL=NL∪{0}.

2. The method of claim 1, wherein the robots Ri, i=1, 2, . . . , K, are single-arm.

3. A method for controlling robots of a tree-like multi-cluster tool in wafer processing, the multi-cluster tool comprising K single-cluster tools denoted as Ci, i=1, 2, . . . , K, the single-cluster tool Ci having a robot Ri for wafer handling, the multi-cluster tool being decomposable into a plurality of cluster-tool-chains (CTCs) each having a serial topology, an individual CTC consisting of single-cluster tools Ci, Ci+1, . . . Ci′ being denoted as CTC[i, i′], processing steps performed by Ci being indexed from 0 to n[i], a first processing step of Ci having an index 0 and being a step of loading a wafer from a loadlock or a buffer module of Ci to Ri, the method comprising: determining values of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], wherein ωij is a robot waiting time for Ri to wait at a j-th step in Ci before unloading the wafer from a process module used for performing the j-th step in Ci; for any i∈{1, 2, . . . , K} and any j∈{0, 1, . . . , n[i]}, when Ri arrives at the process module for the j-th step for unloading the wafer therefrom in the j-th step performed in Ci, waiting, by Ri, for the robot waiting time of ωij; after the robot waiting time of ωij expires, unloading, by Ri, the wafer from the process module for the j-th step; after the wafer is unloaded from the process module for the j-th step, transporting, by Ri, the wafer to the process module for a next step that is either a (j+1)th step, or a zeroth step when j=n[i]; and after the wafer arrives at the process module for the next step, loading, by Ri, the wafer to the second process module; wherein the determining of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], comprises: identifying an L-CTC set consisting of first individual CTCs each having a leaf tool therein, whereby one or more second individual CTCs not in the L-CTC set are also identified; for each of the first individual CTCs, determining a minimum cycle time and then determining the values of ωij by performing a CTC-Check algorithm under a constraint that Θ is equal to the determined minimum cycle time; for each second individual CTC, determining candidate values of ωij for each single-cluster tool therein by performing the CTC-Check algorithm under an initial constraint that Θ is equal to a maximum value of the minimum cycle times already determined for the first individual CTCs; responsive to finding that the candidate ωij values for the second individual CTC are non-negative, setting the candidate ωij values as the determined values of ωij for the second individual CTC; andresponsive to finding that at least one of the candidate ωij values for the second individual CTC is negative, performing: determining a minimum cycle time for the second individual CTC; and determining the values of ωij for the second individual CTC as the candidate values of ωij obtained by performing the CTC-Check algorithm under a revised constraint that Θ is equal to the determined minimum cycle time for the second individual CTC; wherein the CTC-Check algorithm for CTC[f, l] is configured to compute the candidate values of ωij such that ωi(n[i]) is minimized for each value of i selected from N[f, l−1], and the CTC-Check algorithm comprises the steps of: setting ωij=min{Θ−(4λl+3μll(j+1)), Θ−2(n[l]+1)(λll)−Σm∈Ωj\{j}ωlm}, j∈Ωn[l]−1; setting ωl(n[l])=Θ−2(n[l]+1)(λll)−Σm=0n[l]−1ωlm; setting τl0=Θ−φl0; andsetting values of ωij's, i=l−1 to f, according to:

ωi(b[i]−1)=min{τ(i+1)0−(4λi+3μi), Θ−2(n[i]+1)(λii)};

ωij=min{Θ−(4λi+3μii(j+1)), Θ0−2(n[i]+1)(λii)−Σm∈Ωj\{j}ωim} for j∈Ωn[i]−1\{b[i]−1, n[i]};

ωi(n[i])=Θ−2(n[i]+1)(λii)−Σm=0n[i]−1ωim; and

τi0=Θ−φi0, where: b[l] is an index denoting a step of putting the wafer in Cl to a buffer module shared by Cl and Cl+1; θijij+4λi+3μii(j−1), i∈NK and j∈Nn[i]; θi0i0+4λi+3μii(n[i]), i∈NK; Θ is a cycle time of the multi-cluster tool; αij is a time taken in processing the wafer at the j-th step in Ci; τij is a wafer sojourn time at the j-th step in Ci; λi is a time taken by Ri to load or unload a wafer in Ci; μi is a time taken by Ri to move between process modules of Ci for transiting from one processing step to another; NL={1, 2, . . . , L} for a positive integer L; and ΩL=NL∪{0}.

4. The method of claim 3, wherein the robots Ri, i=1, 2, . . . , K, are single-arm.

Read more

Claim Tree

  • 1
    1. A method for controlling robots of a tree-like multi-cluster tool in wafer processing, the multi-cluster tool comprising
    • K single-cluster tools denoted as Ci, i=1, 2, . . . , K, the single-cluster tool Ci having a robot Ri for wafer handling, the multi-cluster tool being decomposable into a plurality of cluster-tool-chains (CTCs) each having a serial topology, an individual CTC consisting of single-cluster tools Ci, Ci+1, . . . Ci′ being denoted as CTC[i, i′], processing steps performed by Ci being indexed from 0 to n[i], a first processing step of Ci having an index 0 and being a step of loading a wafer from a loadlock or a buffer module of Ci to Ri, the method comprising: determining values of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], wherein ωij is a robot waiting time for Ri to wait at a j-th step in Ci before unloading the wafer from a process module used for performing the j-th step in Ci
    • for any i∈{1, 2, . . . , K} and any j∈{0, 1, . . . , n[i]}, when Ri arrives at the process module for the j-th step for unloading the wafer therefrom in the j-th step performed in Ci, waiting, by Ri, for the robot waiting time of ωij
    • after the robot waiting time of ωij expires, unloading, by Ri, the wafer from the process module for the j-th step
    • after the wafer is unloaded from the process module for the j-th step, transporting, by Ri, the wafer to the process module for a next step that is either a (j+1)th step, or a zeroth step when j=n[i]
    • and after the wafer arrives at the process module for the next step, loading, by Ri, the wafer to the second process module
    • wherein the determining of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], comprises: determining candidate values of ωij for the single-cluster tools in an individual CTC by performing a CTC-Check algorithm under a constraint that Θ is equal to a lower bound of cycle time of the multi-cluster tool, the lower bound of cycle time of the multi-cluster tool being a fundamental period of a bottleneck tool among the K single-cluster tools
    • responsive to finding that the candidate ωij values for the individual CTC are non-negative, setting the candidate ωij values as the determined values of ωij for the individual CTC
    • andresponsive to finding that at least one of the candidate ωij values for the individual CTC is negative, performing: determining a minimum cycle time for the individual CTC
    • and determining the values of ωij for the individual CTC as the candidate values of ωij obtained by performing the CTC-Check algorithm under a revised constraint that Θ the determined minimum cycle time for the individual CTC
    • wherein the CTC-Check algorithm for CTC[f, l] is configured to compute the candidate values of ωij such that ωi(n[i]) is minimized for each value of i selected from N[f, l−1], and the CTC-Check algorithm comprises the steps of: setting ωij=min{Θ−(4λl+3μll(j+1)), Θ−2(n[l]+1)(λll)−Σm∈Ωj\{j}ωlm}, j∈Ωn[l]−1
    • setting ωl(n[l])=Θ−2(n[l]+1)(λll)−Σm=0n[l]−1ωlm
    • setting τl0=Θ−φl0
    • andsetting values of ωij's, i=l−1 to f, according to: ωi(b[i]−1)=min{τ(i+1)0−(4λi+3μi), Θ−2(n[i]+1)(λii)}
    • ωij=min{Θ−(4λi+3μii(j+1)), Θ0−2(n[i]+1)(λii)−Σm∈Ωj\{j}ωim} for j∈Ωn[i]−1\{b[i]−1, n[i]}
    • ωi(n[i])=Θ−2(n[i]+1)(λii)−Σm=0n[i]−1ωim
    • and τi0=Θ−φi0, where: b[l] is an index denoting a step of putting the wafer in Cl to a buffer module shared by Cl and Cl+1
    • θijij+4λi+3μii(j−1), i∈NK and j∈Nn[i]
    • θi0i0+4λi+3μii(n[i]), i∈NK
    • Θ is a cycle time of the multi-cluster tool
    • αij is a time taken in processing the wafer at the j-th step in Ci
    • τij is a wafer sojourn time at the j-th step in Ci
    • λi is a time taken by Ri to load or unload a wafer in Ci
    • μi is a time taken by Ri to move between process modules of Ci for transiting from one processing step to another
    • NL={1, 2, . . . , L} for a positive integer L
    • and ΩL=NL∪{0}.
    • 2. The method of claim 1, wherein
      • the robots Ri, i=1, 2, . . . , K, are single-arm.
  • 3
    3. A method for controlling robots of a tree-like multi-cluster tool in wafer processing, the multi-cluster tool comprising
    • K single-cluster tools denoted as Ci, i=1, 2, . . . , K, the single-cluster tool Ci having a robot Ri for wafer handling, the multi-cluster tool being decomposable into a plurality of cluster-tool-chains (CTCs) each having a serial topology, an individual CTC consisting of single-cluster tools Ci, Ci+1, . . . Ci′ being denoted as CTC[i, i′], processing steps performed by Ci being indexed from 0 to n[i], a first processing step of Ci having an index 0 and being a step of loading a wafer from a loadlock or a buffer module of Ci to Ri, the method comprising: determining values of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], wherein ωij is a robot waiting time for Ri to wait at a j-th step in Ci before unloading the wafer from a process module used for performing the j-th step in Ci
    • for any i∈{1, 2, . . . , K} and any j∈{0, 1, . . . , n[i]}, when Ri arrives at the process module for the j-th step for unloading the wafer therefrom in the j-th step performed in Ci, waiting, by Ri, for the robot waiting time of ωij
    • after the robot waiting time of ωij expires, unloading, by Ri, the wafer from the process module for the j-th step
    • after the wafer is unloaded from the process module for the j-th step, transporting, by Ri, the wafer to the process module for a next step that is either a (j+1)th step, or a zeroth step when j=n[i]
    • and after the wafer arrives at the process module for the next step, loading, by Ri, the wafer to the second process module
    • wherein the determining of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], comprises: identifying an L-CTC set consisting of first individual CTCs each having a leaf tool therein, whereby one or more second individual CTCs not in the L-CTC set are also identified
    • for each of the first individual CTCs, determining a minimum cycle time and then determining the values of ωij by performing a CTC-Check algorithm under a constraint that Θ is equal to the determined minimum cycle time
    • for each second individual CTC, determining candidate values of ωij for each single-cluster tool therein by performing the CTC-Check algorithm under an initial constraint that Θ is equal to a maximum value of the minimum cycle times already determined for the first individual CTCs
    • responsive to finding that the candidate ωij values for the second individual CTC are non-negative, setting the candidate ωij values as the determined values of ωij for the second individual CTC
    • andresponsive to finding that at least one of the candidate ωij values for the second individual CTC is negative, performing: determining a minimum cycle time for the second individual CTC
    • and determining the values of ωij for the second individual CTC as the candidate values of ωij obtained by performing the CTC-Check algorithm under a revised constraint that Θ is equal to the determined minimum cycle time for the second individual CTC
    • wherein the CTC-Check algorithm for CTC[f, l] is configured to compute the candidate values of ωij such that ωi(n[i]) is minimized for each value of i selected from N[f, l−1], and the CTC-Check algorithm comprises the steps of: setting ωij=min{Θ−(4λl+3μll(j+1)), Θ−2(n[l]+1)(λll)−Σm∈Ωj\{j}ωlm}, j∈Ωn[l]−1
    • setting ωl(n[l])=Θ−2(n[l]+1)(λll)−Σm=0n[l]−1ωlm
    • setting τl0=Θ−φl0
    • andsetting values of ωij's, i=l−1 to f, according to: ωi(b[i]−1)=min{τ(i+1)0−(4λi+3μi), Θ−2(n[i]+1)(λii)}
    • ωij=min{Θ−(4λi+3μii(j+1)), Θ0−2(n[i]+1)(λii)−Σm∈Ωj\{j}ωim} for j∈Ωn[i]−1\{b[i]−1, n[i]}
    • ωi(n[i])=Θ−2(n[i]+1)(λii)−Σm=0n[i]−1ωim
    • and τi0=Θ−φi0, where: b[l] is an index denoting a step of putting the wafer in Cl to a buffer module shared by Cl and Cl+1
    • θijij+4λi+3μii(j−1), i∈NK and j∈Nn[i]
    • θi0i0+4λi+3μii(n[i]), i∈NK
    • Θ is a cycle time of the multi-cluster tool
    • αij is a time taken in processing the wafer at the j-th step in Ci
    • τij is a wafer sojourn time at the j-th step in Ci
    • λi is a time taken by Ri to load or unload a wafer in Ci
    • μi is a time taken by Ri to move between process modules of Ci for transiting from one processing step to another
    • NL={1, 2, . . . , L} for a positive integer L
    • and ΩL=NL∪{0}.
    • 4. The method of claim 3, wherein
      • the robots Ri, i=1, 2, . . . , K, are single-arm.
See all independent claims <>

Description

LIST OF ABBREVIATIONS

BM buffer module

CTC cluster-tool-chain

FP fundamental period

LL loadlock

PM process module

PN Petri net

R robot

ROPN resource-oriented PN

BACKGROUND

Field of the Invention

The present invention generally relates to scheduling a tree-like multi-cluster tool. In particular, the present invention relates to a method for scheduling this multi-cluster tool to thereby generate an optimal one-wafer cyclic schedule.

List of References

There follows a list of references that are occasionally cited in the specification. Each of the disclosures of these references is incorporated by reference herein in its entirety.

    • W. K. Chan, J. G. Yi, and S. W. Ding, “Optimal Scheduling of Multicluster Tools with Constant Robot Moving Times, Part I: Two-Cluster Analysis,”IEEE Transactions on Automation Science and Engineering, vol. 8, no. 1, pp. 5-16, 2011a.
    • W. K. Chan, J. G. Yi, S. W. Ding, and D. Z. Song, “Optimal Scheduling of Multicluster Tools with Constant Robot Moving Times, Part II: Tree-Like Topology Configurations,”IEEE Transactions on Automation Science and Engineering, vol. 8, no. 1, pp. 17-28, 2011b.
    • T. H. Cormen, C. E. Leiserson, R. L. Rivest, and C. Stein, Introduction to algorithms: MIT press, 2001.
    • M. Dawande, C. Sriskandarajah, and S. P. Sethi, “On Throughput Maximization in Constant Travel-time Robotic Cells,”Manufacture Service Operation Manage, vol. 4, no. 4, pp. 296-312, 2002.
    • S. W. Ding, J. G. Yi, and M. T. Zhang, “Multicluster Tools Scheduling: an Integrated Event Graph and Network Model Approach,”IEEE Transactions on Semiconductor Manufacturing, vol. 19, no. 3, pp. 339-351, 2006.
    • I. Drobouchevitch, S. P. Sethi, and C. Sriskandarajah, “Scheduling Dual Gripper Robotic Cells: One-unit Cycles,”European Journal of Operational Research, vol. 171, no. 2, pp. 598-631, 2006.
    • H. Geismar, C. Sriskandarajah, and N. Ramanan, “Increasing Throughput for Robotic Cells With Parallel Machines and Multiple Robots,”IEEE Transactions on Automation Science and Engineering, vol. 1, no. 1, pp. 84-89, 2004.
    • J. -H. Kim, T. -E. Lee, H. -Y. Lee, and D. -B. Park, “Scheduling analysis of timed-constrained dual-armed cluster tools,”IEEE Transactions on Semiconductor Manufacturing, vol. 16, no. 3, pp. 521-534, 2003.
    • J. -H. Lee, T. -E. Lee, “Concurrent processing of multiple wafer types in a single-armed cluster tool,”IEEE Conference on Automation Science and Engineering, pp. 102-107, 2011.
    • T. -E. Lee, H. -Y. Lee, and Y. -H. Shin, “Workload balancing and scheduling of a single-armed cluster tool,” in Proceedings of the 5th APIEMS Conference, Gold Coast, Australia, pp. 1-15, 2004.
    • T. -E. Lee and S. -H. Park, “An extended event graph with negative places and tokens for timed window constraints,”IEEE Transactions on Automation Science and Engineering, vol. 2, no. 4, pp. 319-332, 2005.
    • M. -J. Lopez and S. -C. Wood, “Systems of multiple cluster tools—configuration, reliability, and performance,”IEEE Transactions on Semiconductor Manufacturing, vol. 16, no. 2, pp. 170-178, 2003.
    • T. Murata, “Petri nets: Properties, analysis and applications,”Proceedings of IEEE, vol. 77, no. 4, pp. 541-580, April 1989.
    • M. X. Liu and B. H. Zhou, “Modelling and scheduling analysis of multi-cluster tools with residency constraints based on time constraint sets”. International Journal of Production Research, vol. 51, no. 16, pp. 4835-4852, 2013.
    • T. L. Perkinson, P. K. MacLarty, R. S. Gyurcsik, and R. K. Cavin III, “Single-wafer cluster tool performance: An analysis of throughput,”IEEE Transactions on Semiconductor Manufacturing, vol. 7, no. 3, pp. 369-373, 1994.
    • T. L. Perkinson, R. S. Gyurcsik, and P. K. MacLarty, “Single-wafer cluster tool performance: An analysis of the effects of redundant chambers and revisitations sequences on throughput,”IEEE Transactions on Semiconductor Manufacturing, vol. 9, no. 3, pp. 384-400, 1996.
    • Y. Qiao, N. Q. Wu, and M. C. Zhou, “Petri Net Modeling and Wafer Sojourn Time Analysis of Single-Arm Cluster Tools With Residency Time Constraints and Activity Time Variation,”IEEE Transactions on Semiconductor Manufacturing, vol. 25, no. 3, pp. 432-446, 2012a.
    • Y. Qiao, N. Q. Wu, and M. C. Zhou, “Real-Time Scheduling of Single-Arm Cluster Tools Subject to Residency Time Constraints and Bounded Activity Time Variation,”IEEE Transactions on Automation Science and Engineering, vol. 9, no. 3, pp. 564-577, 2012b.
    • Y. Qiao, N. Q. Wu, and M. C. Zhou, “A Petri netbased Novel Scheduling Approach and Its Cycle Time Analysis for Dual-Arm Cluster Tools with Wafer Revisiting,”IEEE Transactions on Semiconductor Manufacturing, vol. 26, no. 1, pp. 100-110, 2013.
    • S. Rostami, B. Hamidzadeh, and D. Camporese, “An optimal periodic scheduler for dual-arm robots in cluster tools with residency constraints,”IEEE Transactions on Robotics and Automation, vol. 17, no. 5, pp. 609-618, 2001.
    • S. Venkatesh, R. Davenport, P. Foxhoven, and J. Nulman, “A steady state throughput analysis of cluster tools: Dual-blade versus single-blade robots,”IEEE Transactions on Semiconductor Manufacturing, vol. 10, no. 4, pp. 418-424, 1997.
    • N. Q. Wu, “Necessary and Sufficient Conditions for Deadlock-free Operation in Flexible Manufacturing Systems Using a Colored Petri Net Model,”IEEE Transaction on Systems, Man, and Cybernetics, Part C, vol. 29, no. 2, pp. 192-204, 1999.
    • N. Q. Wu, C. B. Chu, F. Chu, and M. C. Zhou, “A Petri net method for schedulability and scheduling problems in single-arm cluster tools with wafer residency time constraints,”IEEE Transactions on Semiconductor Manufacturing, vol. 21, no. 2, pp. 224-237, 2008.
    • N. Q. Wu, F. Chu, C. B. Chu, and M. C. Zhou, “Petri net modeling and cycle time analysis of dual-arm cluster tools with wafer revisiting,”IEEE Transactions on Systems, Man, &Cybernetics: Systems, vol. 43, no. 1, pp. 196-207, 2013a.
    • N. Q. Wu, F. Chu, C. Chu, and M. Zhou, “Petri Net-Based Scheduling of Single-Arm Cluster Tools With Reentrant Atomic Layer Deposition Processes,”IEEE Transactions on Automation Science and Engineering, vol. 8, no. 1, pp. 42-55, 2011.
    • N. Q. Wu, F. Chu, C. B. Chu, and M. C. Zhou, “Petri net modeling and cycle time analysis of dual-arm cluster tools with wafer revisiting,”IEEE Transactions on Systems, Man, &Cybernetics: Systems, vol. 43, no. 1, pp. 196-207, 2013b.
    • N. Q. Wu and M. C. Zhou, “Avoiding deadlock and reducing starvation and blocking in automated manufacturing systems based on a Petri net model,”IEEE Transactions on Robotics and Automation, vol. 17, no. 5, 658-669, 2001.
    • N. Q. Wu and M. C. Zhou, “Modeling and deadlock control of automated guided vehicle systems,”IEEE/ASME Transactions on Mechatronics, vol. 9, no. 1, pp. 50-57, 2004.
    • N. Q. Wu and M. C. Zhou, System modeling and control with resource-oriented Petri nets, CRC Press, Taylor & Francis Group, New York, October 2009.
    • N. Q. Wu and M. Zhou, “Analysis of wafer sojourn time in dual-arm cluster tools with residency time constraint and activity time variation,”IEEE Transactions on Semiconductor Manufacturing, vol. 23, no. 1, pp. 53-64, 2010a.
    • N. Q. Wu and M. C. Zhou, “A closed-form solution for schedulability and optimal scheduling of dual-arm cluster tools based on steady schedule analysis,”IEEE Transactions on Automation Science and Engineering, vol. 7, no. 2, pp. 303-315, 2010b.
    • N. Q. Wu and M. C. Zhou, “Modeling, analysis and control of dual-arm cluster tools with residency time constraint and activity time variation based on Petri nets,”IEEE Transactions on Automation Science and Engineering, vol. 9, no. 2, 446-454, 2012a.
    • N. Q. Wu and M. C. Zhou, “Schedulability analysis and optimal scheduling of dual-arm cluster tools with residency time constraint and activity time variation,”IEEE Transactions on Automation Science and Engineering, vol. 9, no. 1, pp. 203-209, 2012b.
    • N. Q. Wu, M. C. Zhou, F. Chu, and C. B. Chu, “A Petri-net-based scheduling strategy for dual-arm cluster tools with wafer revisiting,”IEEE Transactions on Systems, Man, &Cybernetics: Systems, vol. 43, no. 5, pp. 1182-1194, September 2013b.
    • N. Q. Wu, M. C. Zhou, and Z. W. Li, Resource-oriented Petri net for deadlock avoidance in flexible assembly systems, IEEE Transactions on System, Man, &Cybernetics, Part A, vol. 38, no. 1, 56-69, 2008b.
    • F. J. Yang, N. Q. Wu, Y. Qiao, and M. C. Zhou, Optimal one-wafer cyclic scheduling of single-arm multi-cluster tools with two-space buffering modules, IEEE Transactions on Systems, Man, &Cybernetics: Systems, vol. 44, no. 12, 1584-1597, 2014.
    • J. Yi, S. Ding, and D. Song, “Steady-state throughput and scheduling analysis of multi-cluster tools for semiconductor manufacturing: A decomposition approach,” in Proceedings of 2005 IEEE International Conference on Robotics and Automation, pp. 292-298, 2005.
    • J. G. Yi, S. W. Ding, D. Z. Song, and M. T. Zhang, “Steady-State Throughput and Scheduling Analysis of Multi-Cluster Tools for Semiconductor Manufacturing: A Decomposition Approach,”IEEE Transactions on Automation Science and Engineering, vol. 5, no. 2, pp. 321-336, 2008.
    • M. C. Zhou and M. D. Jeng, “Modeling, Analysis, Simulation, Scheduling, and Control of Semiconductor Manufacturing Systems: A Petri Net Approach,”IEEE Transactions on Semiconductor Manufacturing, vol. 11, no. 3, pp. 333-357, 1998.
    • W. M. Zuberek, “Timed Petri nets in modeling and analysis of cluster tools,”IEEE Transactions on Robotics Automation, vol. 17, no. 5, pp. 562-575, October 2001.
    • Q. H. Zhu, N. Q. Wu, Y. Qiao, and M. C. Zhou, “Scheduling of Single-Arm Multi-Cluster Tools to Achieve the Minimum Cycle Time,” in Proceedings of 2013 IEEE International Conference on Robotics and Automation, Karlsruhe, Germany, pp. 3555-3560, May 2013a.
    • Q. H. Zhu, N. Q. Wu, Y. Qiao, and M. C. Zhou, “Petri Net-Based Optimal One-Wafer Scheduling of Single-Arm Multi-Cluster Tools in Semiconductor Manufacturing,”IEEE Transactions on Semiconductor Manufacturing, vol. 26, no. 4, pp. 578-591, 2013b.
    • Q. H. Zhu, N. Q. Wu, Y. Qiao, and M. C. Zhou, “Modeling and Schedulability Analysis of Single-Arm Multi-Cluster Tools with Residency Time Constraints via Petri Nets” in Proc. IEEE Int. Conf. Auto. Sci. Eng., Taipei, Taiwan, pp. 81-86, 2014.
    • Q. H. Zhu, N. Q. Wu, Y. Qiao, and M. C. Zhou, “Scheduling of Single-Arm Multi-cluster Tools With Wafer Residency Time Constraints in Semiconductor Manufacturing”IEEE Transactions on Semiconductor Manufacturing, vol. 28, no. 1, 2015.

There follows a list of patent(s) and patent application(s) that are occasionally cited in the specification.

    • D. Jevtic, “Method and Apparatus for Managing Scheduling a Multiple Cluster Tool,” European Patent Application Publication No. 1,132,792 A2, Sep. 12, 2001.
    • D. Jevtic and S. Venkatesh, “Method and Apparatus for Scheduling Wafer Processing within a Multiple Chamber Semiconductor Wafer Processing Tool Having a Multiple Blade Robot,” U.S. Pat. No. 6,224,638 B1, May 1, 2001.

Description of Related Art

With single wafer processing technology, cluster tools are increasingly used in the semiconductor industry. Generally, a single-cluster tool is composed of four to six process modules (PM), one wafer-delivering robot (R), and two loadlocks (LL) for cassette loading/unloading. The robot can be single or dual-arm one resulting in single or dual-arm cluster tools. Several single-cluster tools are connected together by shared buffer modules (BMs) to form a more integrated manufacturing system called a multi-cluster tool. It has higher performance and becomes more and more popular in the semiconductor industry. However, it is very challenging to effectively operate such a system.

Extensive studies have been done on modeling, performance evaluation, and scheduling for single-cluster tools [Kim et al., 2003; Lee et al., 2004; Lee and Park, 2005; Lopez and Wood, 2003; Wu et al., 2008a; Wu and Zhou, 2010b; Qiao et al., 2012a and 2012b; and Wu et al., 2013a and 2013b]. It is found that a cluster tool may operate in process-bound or transport-bound region. For a single-cluster tool, if one of its process steps is the bottleneck of the tool in the sense of workload, it is process-bound. In this case, the cycle time of the system is determined by the wafer processing time. Otherwise, if the robot is the bottleneck, it is transport-bound and the cycle time is determined by the robot task time. In practice, the robot task time is much shorter than the wafer processing time [Kim et al., 2003; and Lopez and Wood, 2003] such that, for a single-arm cluster tool, a backward strategy is shown to be optimal [Lee et al., 2004; Dawande 2002; and Lopez and Wood, 2003]. Some wafer fabrication processes require that a processed wafer should leave a PM within a given time interval, which is called wafer residency time constraints [Kim et al., 2003; Lee and Park, 2005]. Without immediate buffer between the PMs in a tool, this greatly complicates the problem of scheduling single-cluster tools. This problem is studied by using Petri nets and mathematical programming models for dual-arm cluster tools to find an optimal periodic schedule in [Kim et al., 2003; Lee and Park, 2005]. It is further studied for both single and dual-arm cluster tools in [Wu et al., 2008; Wu and Zhou, 2010b; and Qiao et al., 2012a and 2012b] by developing generic Petri net (PN) models. With these models, robot waiting is explicitly modeled as an event to parameterize a schedule by robot waiting time. Then, to find a schedule is to determine the robot waiting time. With these models, schedulability conditions are presented and closed-form scheduling algorithms are given to find an optimal periodic schedule if schedulable.

In recent years, attention has been paid to the problem of scheduling multi-cluster tools. A multi-cluster tool composed of K single-cluster tools is called a K-cluster tool. A heuristic method is proposed in [Jevtic, 2001] for it by dynamically assigning priorities to PMs. However, the performance of such a schedule is difficult to evaluate. Geismar et al. examine a serial 3-cluster tool composed of single-arm tools and parallel processing modules. By simulation, they find that, for 87% of instances, the backward strategy achieves the lower-bound of cycle time. In [Ding et al., 2006], an event graph model is used to describe the dynamic behavior of the system and a simulation-based search method is proposed to find a periodic schedule.

To reduce the computational complexity, without considering the robot moving time, a decomposition method is proposed in [Yi et al., 2005 and 2008]. By this method, the fundamental period (FP) for each tool is calculated as done for scheduling single-cluster tools. Then, by analyzing time delays resulting from accessing the shared buffers, the global fundamental period, or the cycle time for the system is determined. In this way, a schedule is found.

With robot moving time considered, a polynomial algorithm is presented to find a multi-wafer schedule for a serial multi-cluster tool in [Chan et al., 2011a]. A K-cluster tool is said to be process-dominant if its bottleneck tool is process-bound. It is known that there is always an optimal one-wafer schedule for a process-dominant serial multi-cluster tool [Zhu et al., 2013a, 2012; and Zhu et al., 2013b]. In studying the effect of buffer spaces in BMs on productivity, Yang et al. [2014] show that, for a process-dominant serial multi-cluster tool with two-space BMs, there is a one-wafer cyclic schedule such that the lower bound of cycle time is reached. For a single-arm multi-cluster tool with two-space BMs and wafer residency time constraints, Liu and Zhou [2013] propose a non-linear programming model and a heuristic algorithm to solve it. For a single-arm multi-cluster tool with single-space BMs and wafer residency time constraints, Zhu et al. [2014 and 2015] present sufficient and necessary schedulability conditions and an efficient algorithm to find a one-wafer optimal cyclic schedule if schedulable.

Chan et al. [2011b] study the problem of scheduling a tree-like multi-cluster tool, which is the only report on tree-like multi-cluster tool scheduling to our best knowledge. The obtained schedule is a multi-wafer cyclic one. It derives conditions under which a tree-like multi-cluster tool can be scheduled by a decomposition method. Further, if decomposable, conditions under which a backward scheduling strategy is optimal are presented.

Since a one-wafer cyclic schedule is easy to implement and understand by a practitioner, it gives rise to a question if there exists a one-wafer cyclic schedule for a tree-like multi-cluster tool and how such a schedule can be found if it exists. There is a need in the art to derive the conditions for which this schedule exists and to develop a method for scheduling a tree-like multi-cluster tool with a one-wafer cyclic schedule.

SUMMARY OF THE INVENTION

An aspect of the present invention is to provide a computer-implemented method for scheduling a tree-like multi-cluster tool to generate a one-wafer cyclic schedule. The multi-cluster tool comprises plural single-cluster tools each having a robot for wafer handling, and is decomposable into a plurality of CTCs each having a serial topology.

The method comprises determining robot waiting times for the robots. In particular, the method makes use of Theorem 1 detailed below to determine the robot waiting times.

Preferably, a CTC-Check algorithm as detailed hereinafter computes candidate values of the robot waiting times for at least one individual CTC. The CTC-Check algorithm is configured to compute these candidate values such that the robot waiting time at a last processing step of each single-cluster tool in the individual CTC apart from the leaf tool therein is minimized.

With the use of the CTC-Check algorithm, the robot waiting times may be determined according to Algorithms 1 and 2 detailed below.

Other aspects of the present invention are disclosed as illustrated by the embodiments hereinafter.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 depicts an example of a tree-like multi-cluster tool.

FIG. 2 depicts a PN model for an individual cluster tool Ci as an illustrative example.

FIG. 3 depicts a PN model for a BM shared by adjacent clusters Ci and Ca as an illustrative example.

FIG. 4 shows a multi-cluster tool considered for analysis in Example 1.

DETAILED DESCRIPTION

Since a one-wafer cyclic schedule is easy to implement and understand by a practitioner, it gives rise to a question if there exists a one-wafer cyclic schedule for a tree-like multi-cluster tool and how such a schedule can be found if it exists. The present work provides a solution of schedulability and a method for scheduling the tree-like multi-cluster tool with the one-wafer cyclic schedule. Since the robot task time is much shorter than the wafer processing time in practice [Kim et al., 2003; and Lopez and Wood, 2003], it is reasonable to assume that the bottleneck tool of a multi-cluster tool is process-bound, or it is process-dominant. Hence, we address a process-dominant tree-like multi-cluster tool in this work. The key for scheduling a K-cluster tool is to coordinate the activities of the multiple robots in accessing the BMs. In scheduling single-cluster tools, Wu et al. [2008] and Wu and Zhou [2010a] find that it is crucial to determine when a robot should wait and how long it should wait if necessary. By following this idea, the present work develops a PN model for a tree-like multi-cluster tool with the robot waiting being explicitly modeled. With the model, a schedule for the system is parameterized by the robot waiting time. For a process-dominant tree-like multi-cluster tool, necessary and sufficient conditions under which a one-wafer cyclic schedule exists are derived. It is then shown that there is always a one-wafer cyclic schedule. An efficient algorithm is developed and disclosed herein to find the minimal cycle time and the one-wafer cyclic schedule.

A. System Modeling by Petri Net

A.1. Tree-like Multi-cluster Tool

The configuration of a tree-like multi-cluster tool is illustrated in FIG. 1. A tree-like multi-cluster tool composed of K single-cluster tools is called a K-tree-cluster tool. It follows from Chan et al. [2011b] that a K-tree-cluster tool can be described as follows. Let NK={1, 2, . . . , K} be the index set of the cluster tools that form a K-tree-cluster tool. The tool with the LLs in it is called the “root” and is denoted by C1. If Ci, i∈NK\{1}, has only one adjacent tool, it is called a leaf tool. From C1 to each leaf, there is a directed path. On such a path, if Ci is closer to C1 than Cj, Ci is the upstream tool of Cj.

As done in [Chan et al. 2011b], for a K-tree-cluster tool, we assume: 1) the root has two LLs; 2) a BM shared by two adjacent tools can hold one wafer at a time without a processing function; 3) a PM can process one wafer at a time; 4) all the wafers in a cassette should be processed in an identical sequence specified in the recipe and visit a PM no more than once (except for a BM); 5) the loading, unloading, moving time of the robots, and the processing time of a wafer at a PM are deterministic and known; 6) the loading and unloading time of a robot is same, and the robot moving time from a PM to another is also same regardless of whether it carries a wafer or not; and 7) there is no parallel module, or only one PM is configured for each processing step.

With two LLs, while one lot of wafers in one LL is being processed, the other LL can be used for loading/unloading another lot of wafers. In this way, a multi-cluster tool can work consecutively without interruption. Thus, it operates in a steady state at most of time and there are always wafers to be processed. This is equivalent to that the LLs have infinite capacity.

A.2. Operation Process

The wafer fabrication process can be illustrated by the 5-tree-cluster tool shown in FIG. 1. With the LL and BMs being treated as a step, let PSij denote the j-th step in Ci. Then, the processing route of a wafer is: PS10 (LL)→PS11→PS12(PS20)→PS21→PS22→PS23→PS24→PS20 (PS12)→PS13→(PS30)→PS31→PS32 (PS40)→PS41→PS42 (PS50)→PS51→PS52→PS53→PS54→PS50 (PS42)→PS43→PS44→PS40 (PS32)→PS33→PS34→PS30 (PS13)→PS14→PS10 (LL).

In Ci, i∈NK\{1}, the BM denoted as PMi0 can be viewed as a virtual LL for injecting wafers from an upstream tool, while the other BM(s) PMij (j>0) can be viewed as a virtual process module(s). The BM shared by adjacent tools Ci and Cj is called an outgoing buffer for Ci and an incoming one for Cj. The LLs in C1 can be treated as an incoming buffer with infinite capacity. There are a number of processing steps in each individual tool and a BM connecting two tools is treated as a processing step with processing time being zero. Let Ψ denote the index set of leaf tools and DI(Ci) the index set of the immediate downstream tools of Ci (i∉Ψ). We have DI(Ci)=Ø if i∈Ψ. Further, let PMi(b[i,k]) denote the BM shared by Ci and Ck, k∈DI(Ci), with PMk0 being the incoming buffer or virtual LL for Ck. It is called Step b[i,k] for Ci and Step 0 for Ck, respectively. Assume that, except the incoming buffer, there are n[i] steps in Ci. Then, there are n[i]+1 steps in Ci denoted as PSi0, PSi1, . . . , PSi(n[i]) with PSi0 and PSi(b[i,k]) being the incoming and outgoing steps. Let Ri denote the robot in Ci. As discussed above, at most of time, a multi-cluster tool operates in a steady state. Under such a state, for a process-dominant multi-cluster tool, a backward strategy [Dawande 2002] is optimal for each individual tool.

A.3. Modeling the Wafer Flow

We model a multi-cluster tool system by extending the resource-oriented PN (ROPN) proposed in [Wu, 1999; Wu et al., 2008a and 2008b; and Wu and Zhou, 2009]. It is a finite capacity PN=(P, T, I, O, M, K), where P is a finite set of places; T is a finite set of transitions, P∪T≠Ø, and P∩T=Ø; I:P×T→N={0, 1, 2, . . . } is an input function; O:P×T→N is an output function; M:P→N is a marking denoting the number of tokens in P with M0 being the initial marking; and K:P→N\{0} is a capacity function with K(p) being the number of tokens that p can hold at a time. The basic concept, transition enabling and firing rules of Petri nets (PN) can be found in [Zhou et al., 1998; Wu and Zhou, 2009].

To model a multi-cluster tool, one needs to model the individual tools and the BMs. The behavior of Step j in Ci is modeled as follows. The PM for Step j in Ci with i≠1 and j≠0 is modeled by timed place pij with K(pij)=1. Place p10 models the LLs in C1 with K(p10)=∞. Places zij and dij model Ri's holding a wafer (token) for loading into pij and moving from Steps j to j+1 (or Step 0 if j=n[i]), respectively. Place qij models Ri's waiting at Step j for unloading a wafer there. Transitions tij and uij model Ri's loading and unloading a wafer into and from PMij, respectively. Then, with arcs (zij, tij), (tij, pij), (pij, uij), (qij, uij), and (uij, dij), of the PN model for Step j in Ci is obtained as shown in FIG. 2.

With the model for a step as a module, tool Ci is modeled as follows. Place ri models Ri with K(ri)=1, meaning that, with only one arm, the robot can hold one wafer at a time. Timed yij with arcs (ri, yij) and (yij, qij) for connecting ri and qij, representing that Ri moves from Steps j+2 to j, j=0, 1, . . . , n[i]−2 (yi(n[i]−1) for moving from Steps 0 to n[i]−1, and yi(n[i]) for moving from Steps 1 to n[i]). Finally, xij, j=0, 1, 2, . . . , n[i]−1, are added between dij and zi(j+1) with arcs (dij, xij) and (xij, zi(j+1)); and xi(n[i]) is added between di(n[i]) and zi0 with arcs (di(n[i]), xi(n[i])) and (xi(n[i]), zi0). In this way, the modeling of Ci is completed as shown in FIG. 2.

A BM linking Ci and Ca, a∈DI(Ci), is denoted by PMi(b[i,a]) in Ci and PMa0 in Ca. Thus, pi(b[i,a]) and pa0 with pi(b[i,a])=pa0 are used to model the BM. Then, Step b[i,a] for Ci is modeled by {pi(b[i,a]), qi(b[i,a]), zi(b[i,a]), di(b[i,a]), ti(b[i,a]), ui(b[i,a])} and Step 0 for Ca by {Pa0, qa0, za0, da0, ta0, ua0}, respectively. Thereafter, when we refer to Step b[i,a] in Ci, we use pi(b[i,a]), whereas for Step 0 in Ca we use pa0. It should be pointed out that Step 0 in C1 (the LLs) is not shared with any other tool and there is no outgoing buffer in a leaf tool CL. The PN model for the BM linking Ci and Ca is shown in FIG. 3.

With the PN structure developed, marking M0 is set as follows. To describe the idle state of the system, a special type of token called V-token representing a virtual wafer is introduced. As mentioned above, there is a process-bound bottleneck tool in the system. Hence, in the steady state, if there is a wafer being processed in every PMij with j≠0, the system reaches its maximal throughput. Let χ denote the index set of non-leaf tools NK=χ∪Ψ with χ∩Ψ=Ø. Then, M0 is set as follows. For Ci, i∈NK, set M0(pij)=1 for j≠1 with a V-token, M0(pij)=0, j=1, and M0(p10)=n; M0(qij)=M0(zij)=M0(dij)=0 for any j; and M0(ri)=1. It is assumed that the token in pi(b[i,a]) enables ua0, where a∈DI(Ci).

It is easy to show that the obtained PN is deadlock-prone. To make it deadlock-free, let Ωn=Nn∪{0}, a control policy on yij's is defined as follows.

Definition 1: At marking M, yij, i∈NK and j∈Ωn[i]−1, is said to be control-enabled if M(pi(j+1))=0; y1(n[i]) is control-enabled if M(p1(n[i]))=1; and yi(n[i]), i∈NK\{1}, is control-enabled if M(pi0)=0.

With Definition 1, the PN for an individual tool is shown to be deadlock-free [Wu et al., 2008a]. Thereafter, it is assumed that the PN is controlled by the policy given in Definition 1.

In the PN model for a BM shared by Ci and Ca shown in FIG. 3, pi(b[i,a]) (pa0) has two output transitions ui(b[i,a]) and ua0, leading to a conflict. A token entering pi(b[i,a]) by firing ti(b[i,a]) in Ci should enable ua0 in Ca, while a token entering pa0 by firing ta0 in Ca should enable ui(b[i,a]) in Ci. Colors are introduced to make the model conflict-free as follows.

Definition 2: Define the unique color of transition ti as C(ti)={ci}.

By Definition 2, the colors for ui(b[i,a]) and ua0 are ci(b[i,a]) and ca0, respectively. Then, the color for a token in pi(b[i,a]) or pa0 is defined as follows.

Definition 3: For i∈χ, a∈DI(Ci), a token entering pi(b[i,a]) by firing ti(b[i,a]) in Ci has color ca0, while a token entering pa0 by firing ta0 in Ca has color ci(b[i,a]).

It is easy to check that, by Definitions 2 and 3, the model is conflict-free. The obtained model can also describe the behavior of initial and final transient processes as explained as follows. At M0, yi0 is enabled and fires leading to M1(qi0)=1. By starting from M1, transition u10 fires and W1 representing the first wafer is unloaded from p10. In this way, every time when u10 fires, a token for a real wafer is delivered into p11. With the transition enabling and firing rules, the model evolves to reach a marking M such that every V-token is removed from the model and is replaced by a token for a real wafer, or a steady state is reached. The evolution from M0 to M presents the initial transient process. Similarly, when the system needs to stop working, every time when u10 fires, a V-token is delivered into p11, the model can evolve back to marking M0, which describes the final transient process.

A.4. Modeling Activity Time

To describe the temporal aspect, time is associated with both transitions and places. It follows from Kim et al. [2003] and Lee and Lee [2006] that the time for the robot to move from one step to another is a constant, and so is the time for the robot to load/unload a wafer into/from a PM. The symbols and time durations for places and transitions are listed in TABLE 1.


TABLE 1
The time durations associated with transitions and places (i∈Nn, j∈Ωn[i]).
Transition or
Time
Symbol
place
Actions
required
λi
uij∈T
Robot Ri unloads a wafer from pij
Constant λi
tij∈T
Robot Ri loads a wafer into pij
μi
yij∈T
Robot Ri moves from a step to
Constant μi
another
xij∈T
αij
pij∈P
Wafer processing in pij
Constant αij
τij
pij∈P
Wafter sojourn in pij
τij
ωij
qij∈P
Robot Ri waits before unloading a
[0, ∞)
wafer from pij
zij, dij∈P
No time is associated
0

B. Properties Of Scheduling A K-Cluster Tool

This section recalls the properties of scheduling an individual tool in [Zhu et al., 2013]. For a process-dominant K-tree-cluster tool, since a backward strategy is optimal for each individual tool, with the PN model shown in FIG. 2, Ri's cycle time is given by [Wu et al., 2008]

ψi=2(n[i]+1)(λii)+Σj=0n[i]ωiji1i2, i∈NK,  (1)

where ψi1=2(n[i]+1)(λii) is Ri's task time and is a constant, and ψi2j=0n[i]ωij is the robot waiting time in a cycle.

With robot waiting as an event, the time taken for completing a wafer at Step j in Ci, i∈NK, is given by [Wu et al., 2008]

ξijij+4λi+3μii(j−1), i∈NK and j∈Nn[i],  (2)

and

ξi0=4λi+3μii(n[i]), i∈NK.  (3)

Note that, since a BM has no processing function, we have αi0i(b[i,a])=0, a∈DI(Ci). With αij, λi, and μi being constants and ωij's being variable, if ωij is set to be zero, the shortest time taken for completing a wafer at Step j in Ci is

ηijij+4λi+3μi, i∈NK and j∈Ωn[i].  (4)

A cluster tool may be scheduled such that a wafer stays in PMij for some time after its completion, or τij≥αij, where τij is the wafer sojourn time in PMij. For Step j∈{0, b[i,a]}, a∈DI(Ci), or a BM, the wafer unloaded from PSij by the f-th firing of uij is not the one that is loaded into PSij by the f-th firing of tij. However, for scheduling purpose, we concern only the time needed for completing the tasks in accessing a BM but not if they are the same wafer. Hence, the virtual wafer sojourn time for Step j∈{0, b[i, a]}, a∈DI(Ci), is introduced as follows. Let υ1 and υ2 be the end time point of the k-th firing of tij and the start time point of the k-th firing of uij, respectively. The virtual wafer sojourn time is defined as τij2−υ1. In this way, for all steps in Ci, τij has the same meaning. Then, by replacing αij with τij in (2) and (3), the cycle time at Step j in Ci is given by

θijij+4λi+3μii(j−1), i∈NK and j∈Nn[i],  (5)

and

θi0i0+4λi+3μii(n[i]), i∈NK.  (6)

Let Πi=max{ηi0, ηi1, . . . , ηi(n[i]), ψi1}. If Πii1, Ci is transport-bound, otherwise it is process-bound. The manufacturing process in Ci is a serial one and Ci should be scheduled such that

Θii0i1= . . . =θi(n[i]) and τij≥αij, i∈NK and j∈Ωn[i]  (7)

where Θi is the cycle time of Ci. To find a feasible schedule, Θi≥Πi must hold. Given Θi for Ci, Ri should be scheduled such that ψii. Thus, ψi2i−ψi1 is the Ri's waiting time in a cycle, or the schedule is a function of Ri's waiting time. This implies that a one-wafer cyclic schedule for a multi-cluster tool is determined by the robots' waiting time. The key here is how to determine the minimal Θi's.

Let Θ denote the cycle time of a K-tree-cluster tool and Θi the cycle time of Ci, i∈NK. From [Zhu et al., 2013b], it can easily show that, to obtain a one-wafer cycle schedule for a K-tree-cluster tool, Θ12= . . . =ΘK=Θ should hold. Let Ch, 1≤h≤K, be the bottleneck tool in a K-tree-cluster tool and Πh be its FP. Then, Θ12= . . . =ΘK=Θ≥Πh must hold. Hence, to obtain a one-wafer cycle schedule for a K-tree-cluster tool, every individual tool should have the same cycle time and, at the same time, the multiple robots should be coordinated such that all the individual tools operate in a paced way.

Assume that Ci, ∀i∈NK. It is scheduled to have a cycle time Θ. Examine Ci and Ca, i∉Ψ and a∈DI(Ci). If the system is scheduled such that, every time after Ri loads a wafer into PSi(b[i, a]) by firing ti(b[i, a]), ua0 fires immediately to unload this wafer by Ra, then Ci and Ca can operate in a paced way. If a K-tree-cluster tool is scheduled such that, any pair of Ci and Ca can operate in such a way, a one-wafer cyclic schedule is obtained. We then examine how Ci and Ca, i∉Ψ and a∈DI(Ci), can be scheduled to operate in a paced way. Let ϕi0 be the start time point for firing ui0 in Ci. Between the firings of ui0 and ti(b[i, a]), the following transition firing sequence is executed: σi=firing ui0 (with timeλi)→xi0i)→ti1i)→yi(n[i])i)→robot waiting in qi(n[i])i(n[i]))→ui(n[i])i)→xi(n[i])i)→ti0i)→ . . . →yi(b[i, a]−1)i)→robot waiting in qi(b[i, a]−1)i(b[i, a]−1))→ui(b[i, a]−1)i)→xi(b[i, a]−1)i)→ti(b[i, a])i). Given a schedule for Ci by setting ωij's, the time taken by σi is Δi={(n[i]+1)−(b[i, a]−2)}(2λi+2μi)−μij=b[i,a]−1n[i]ωij, if b[i, a]>1; and Δi=2λii, if b[i, a]=1. Then, ua0 starts to fire at ϕa0i0i, i∉Ψ. In other words, to make Ci and Ca operate in a paced way with a cycle time Θ, ua0 should fire Δi time units later than the firing of ui0.

With the PN model, the above requirement can be implemented by properly scheduling Ci and Ca as follows. At M0 that is set as above discussed, u10 fires and W1 representing the first real wafer released into the system is unloaded from p10. Then, for an m∈DI(C1), transition firing sequence σ1=firing u101)→x101)→t111)→y1(n[1])1)→robot waiting in q1(n[1])1(n[1]))→u1(n[i])1)→x1(n[1])1)→t101)→ . . . →y1(b[1, m]−1)1)→robot waiting in q1(b[1, m]−1)1(b[1, m]−1))→u1(b[1, m]−1)1)→x1(b[1, m]−1)1)→t1(b[1, m])1) is executed in C1. By firing t11, W1 is loaded into PS11. After firing t1(b[1,m]), um0 is enabled and fires. The time taken by σ1 is Δ1={(n[1]+1)−(b[1, m]−2)}(2λ1+2μ1)−μ1j=b[1,m]−1n[1]ωij such that ϕm0101. Then, in C1, t1(b[1, m]) fires again after a cycle. Meanwhile, in Cm, a cycle is also completed. Since the cycle time is Θ for both C1 and Cm, um0 fires again just after t1(b[1, m]) fires. During this cycle, W2 is released into the system.

By starting from the firing of um0, transition firing sequence σm=um0m)→xm0m)→tm1m)→ym(n[m])m)→robot waiting in qm(n[m])m(n[m]))→um(n[m])m)→xm(n[m])m)→tm0m)→ . . . →ym(b[m, d]−1)m)→robot waiting in qm(b[m, d]−1)m(b[m, d]−1))→um(b[m, d]−1)m)→xm(b[m, d]−1)m)→tm(b[m, d])m) is executed in Cm, d∈DI(Cm). After firing tm(b[m, d]), ud0 is enabled and fires. The time taken by σm is Δm such that ϕd0m0m. This process can be performed for every pair of Cm and Cd, d∈DI(Cm), such that every tool in the system acts in a paced way. Notice that, after each cycle in C1, a V-token is removed from the system. When all V-tokens are removed, the steady state is reached and the system operates according to the backward one-wafer cycle schedule.

Notice that, to perform the above process for Cm and Cd, d∈DI(Cm), it requires that: 1) when Rm arrives at pm(b[m, d]) for firing um(b[m, d]) (tm(b[m, d])), there is a token in pm(b[m,d]) (pm(b[m, d]) is emptied); and 2) when Rd arrives at pd0 for firing ud0 (td0), there is a token in pd0 (pd0 is emptied). To make these requirements satisfied, we have the following theorem.

Theorem 1: Given Θ≥Πh as cycle time for a process-dominant K-tree-cluster tool, a one-wafer cyclic schedule exists if and only if, for any pair of Ci and Ca, a∈DI(Ci), the following conditions are satisfied by determining ωij's and ωaf's:

θijaf=Θ, j∈Ωn[i] and f∈Ωn[a],  (8)

τi(b[i, a])≥4λa+3μaa(n[a])  (9)

and

τa0≥4λi+3μii(b[i, a]−1).  (10)

The proof of Theorem 1 is similar to that in [Zhu et al., 2013b] and omitted. By Theorem 1, to find a one-wafer cyclic schedule, the key is to determine Θ such that (8)-(10) are satisfied by setting ωij's and ωaf's. This issue is discussed next.

C. One-Wafer Cyclic Scheduling

C.1. Scheduling of a Cluster-tool-chain

By viewing an individual tool as a vertex, the topology of a K-tree-cluster tool is a tree (directed graph) [Cormen et al., 2001]. Let χ(i, j) denote the index set of the tools on the path from Ci to Cj, including i and j, and |χ(i, j)| be its cardinality. The length of the path from Ci to Cj is defined as the number of arcs connecting two adjacent tools on a path and is denoted by Len(Ci, Cj)=|χ(i, j)|−1. If there is a path from Ci to Cj and Len(C1, Ci)<Len(C1, Cj), Ci is an upstream tool of Cj and Cj is a downstream one of Ci. Let O-Deg(Ci) denote the number of downstream tools immediately adjacent to Ci. Ci is called a fork tool if O-Deg(Ci)≥2. For example, in the 5-tree-cluster tool shown in FIG. 1, Len(C1, C5)=3, O-Deg(C1)=2, and O-Deg(C3)=O-Deg(C4)=1.

A multi-cluster tool with serial topology can be seen as a cluster-tool-chain (CTC). Then, a K-tree-cluster tool can also be seen to be formed by a number of CTCs. A CTC that starts from Cf and ends at Cl is denoted as CTC[f, l]. Note that a CTC can be formed by a single tool and in this case we have f=l. Let UI(Ci) denote the index of the immediate upstream tool of Ci. Since C1 has no upstream tool, we set UI(C1)=0. Then, we divide a K-tree-cluster tool into a number of CTCs such that, for a CTC[f, l], we have O-Deg(Cj)≥2, j=UI(Cf) and f≠1; either O-Deg(Cl)≥2 or O-Deg(Cl)=0; and O-Deg(Cl)=1, i∈χ(i, j), i≠f, and i≠l. For a K-tree-cluster tool, the individual tools can be labeled such that the tools on each CTC[f, l] are labeled with consecutive numbers in an increasing order from f to l. That is, the immediate upstream and downstream tools of Ci are Ci−1 and Ci+1, respectively. For example, the 5-tree-cluster tool shown in FIG. 1 has three CTCs: {C2}, {C3, C4, C5}, and {C1}. Algorithm 2 is presented later to find such CTCs. Before presenting the method for scheduling a K-tree-cluster tool, we discuss how to schedule the CTCs first.

Given CTC[f, l], there is only one BM shared by Ci and Ci+1, i∈N[f, l−1]={f, f+1, . . . , l−1}. Hence, for the simplicity of presentation, for a CTC, we can use PSi(b[i]) for the BM in Ci instead of PSi(b[i,i+1]). Let φi(b[i])=4λi+3μii(b[i]−1) and φi0=4λi+3μii(n[i]). Examining the conditions given in Theorem 1, one can find that, to schedule CTC[f, l] from Cl to Cf such that these conditions are satisfied, one needs to set the robots' waiting time such that: 1) ωl(n[l]) is as small as possible; and 2) ωi(n[i]) and ωi(b[i]−1), i∈N[f, l−1], are as small as possible. Given cycle time Θ≥Πh, this can be implemented as follows: (1) set ωij=min{Θ−(4λl+3μll(j+1)), Θ−2(n[l]+1)(λll)−Σm∈Ωj\{j}ωlm}; j∈Ωn[l]−1; (2) ωl(n[l])=Θ−2(n[l]+1)(λll)−Σm=0n[l]−1ωlm; (3) τl0=Θ=φl0; and (4) set ωij's from i=l−1 to f according to

ωi(b[i]−1)=min{τ(i+1)0−(4λi+3μi), Θ−2(n[i]+1)(λii)},  (11)

ωij=min{Θ−(4λi+3μii(j+1)), Θ0−2(n[i]+1)(λii)−Σm∈Ωj\{j}ωim} for j∈Ωn[i]−1\{b[i]−1, n[i]},  (12)

ωi(n[i])=Θ−2(n[i]+1)(λii)−Σm=0n[i]−1ωim  (13)

and

τi0=Θ−φi0.  (14)

If ∀i∈N[f, l−1], ωi(b[i]−1)≥0, the conditions given in Theorem 1 are satisfied. That is, given Θ≥Πh, a one-wafer cyclic schedule is found for CTC[f, l] by setting the ωij's. However, i∈N[f, l−1], ωi(b[i]−1)<0 may happen. Since Θ−2(n[i]+1)(λii)≥0 always holds. ωi(b[i]−1)<0 implies that τ(i+1)0−(4λi+3μi)<0, leading to τ(i+1)0<(4λi+3μi), i.e., the conditions in Theorem 1 are violated. Thus, the above procedure can be used to not only schedule a CTC if a feasible one-wafer schedule exists for the given cycle time Θ, it also can be used to check the existence of a one-wafer cyclic schedule for CTC[f, l] and this procedure is called CTC-Check.

Note that, in the above procedure, for i∈N[f, l−1], ψi2 is assigned ωij's, j∈Ωn[i]−1} as much as possible. In this way, ωi(n[i]) and φi0=4λi+3μii(n[i]) are minimized.

C.2. Minimal Cycle Time of a CTC

Given Θ≥Πh for CTC[f, l], if ωi(b[i]−1)<0, i∈N[f, l−1], happens by using CTC-Check, one cannot find a one-wafer schedule with cycle time Θ. However, a one-wafer cyclic schedule still exists and such a schedule can be found by increasing Θ. The key is how to determine the minimal cycle time Θ and this issue is discussed as follows.

For CTC[f, l] and given Θ, τ(k+1)0<(4λk+3μk), k∈N[f, l−1], we assume that τ(i+1)0≥(4λi+3μi), i∈N[k+1, l−1], that is, the conditions given in Theorem 1 are satisfied for Ci, ∀i∈N[k+1, l−1], but not satisfied for Ck. Since 4λk+3μk is a constant, to make τ(k+1)0≥(4λk+3μk) satisfied, the only way is to increase τ(k+1)0. Since τ(k+1)0=Θ−(4λk+1+3μk+1(k+1)(n[k+1])) and (4λk+1+3μk+1) is a constant, to increase τ(k+1)0, one needs to increase Θ and reduce ω(k+1)(n[k+1]). By examining the CTC-Check procedure, one can find that, by increasing Θ, one can assign more part of ψ(k+1)2 into ω(k+1)j, j∈Ωn[k+1]−1\{n[k+1]}, such that ω(k+1)(n[k+1]) is reduced. In this way, τ(k+1)0 can be increased by increasing Θ and reducing ω(k+1)(n[k+1]) simultaneously.

By increasing Θ, to obtain a one-wafer cyclic schedule, the cycle time for any Ci in a K-tree-cluster tool should be increased as well such that every tool has the same cycle time. Hence, by increasing Θ for Ci, i∈N[k+2, l], ωi(n[i]) is reduced. Then, by the CTC-Check procedure, this leads to the decrease of ω(i−1)(n[i−1]). Finally, the increase of Θ for Ci, i∈N[k+2, l−1], can result in the decrease of ω(k+1)(n[k+1]). This implies that, with the increase of Θ, ω(k+1)(n[k+1]) can be reduced by scheduling Ck+1 and its downstream tools. However, given Θ, if CTC[f, l] is scheduled by the CTC-Check procedure such that ωi(n[i])=0, i∈N[k+2, l], the increase of Θ for Ci and its downstream tools has no effect on ω(k+1)(n[k+1]), since ωi(n[i]) cannot be reduced any more. Thus, the effect of scheduling downstream tools on ω(k+1)(n[k+1]) should be taken into account in finding the minimal cycle time.

Note that Cl is either a leaf or a fork. If it is a fork, it has multiple outgoing BMs. This means that CTC[f, l] has multiple downstream CTCs. As above discussed, with the increase of Θ, scheduling tool Ci in a downstream CTC of CTC[f, l] can result in the reduction of ω(k+1)(n[k+1]). Also, given Θ, if a schedule for tool Ci in a downstream CTC is obtained by the CTC-Check procedure such that ωi(n[i])=0, then, with the increase of Θ, scheduling Ci and its downstream tools has no effect on ω(k+1)(n[k+1]).

With the above discussion, we present how to find the minimal cycle time and schedule the system as follows. Given Θ, for CTC[f, l], every time if ωk(b[k]−1)<0, k∈N[f, l−1] is returned by the CTC-Check procedure, the cycle time is increased by Δ, i.e. Θ←Θ+Δ, such that τ(k+1)0=(4λk+3μk). The key is how to determine Δ. To do so, define D[i]={PMij|j∈Ωn[i]−1\{b[i]−1, n[i]}} be the set of steps in Ci and B[i]=|D[i]|, and B*[i]=B[i] if i≠l, and B*[l]=n[l]−1.

Given Θ, for CTC[f, l], when ωk(b[k]−1)<0, k∈N[f, l−1], is returned by the CTC-Check procedure, for every Ci, i∈N[k, l], or in the downstream CTCs of CTC[f, l], the robot waiting time is set by the CTC-Check procedure. To find Δ, for Ci, i∈N[k, l], or in the downstream CTCs of CTC[f, l], let Aijij and Λijij, j∈Ωn[i], and Φ=4λk+3μk−Λ(k+1)0=4λk+3μk−(Θ−(4λk+1+3μk+1(k+1)(n[k+1])))=4λk+3μk+4λk+1+3μk+1(k+1)(n[k+1])−Θ.

To take the effect of tools in the downstream CTCs of CTC[f, l] into account, we use DsI(Ci), i∉Ψ, to denote the index set of all the downstream tools of Ci and LeafDsI(Ci), the index set of leaf tools in DsI(Ci). Further, define E(k)={g∈∪χ(k, m)|m∈LeafDsI(Ck), ωj(n[j])>0, j∈∪χ(h, g) and h=DI(Ck), and ω(g+1)(n[g+1])=0, g∉LeafDsI(Ck)}. Further, let Y(k)={∪χ(h, s)|s∈E(k) and h=DI(Ck)}. Then, Δ is determined as follows. For g∈Y(k), let δgg(n[g])j∈Y(g)B*[j], δk=Φ/(1+Σj∈Y(k)B*[j]), and Q(k)={a∈∪χ(k, d)|d∈E(k)}. Then, let Δ=minh∈Q(k)δh and there are two cases.

Case 1: If Δ=δk, let Θ←Θ+Δ. For ∀i∈DsI(Ca) and ∀a∈E(k), set ωi0=Ai0+Δ, τi0i0+Δ, and τi(m−1)i(m−1)+Δ with m∈BI(Ci), when Ci is not a leaf tool. For ∀i∈Y(k)\E(k), ωij=Aij+Δ, j∈Ωn[i]\{{j−1|j∈BI(Ci)}∪{n[i]}}, ωi(b[m, d]−1)=Ai(b[m, d]−1)+Δ×Σj∈{∪χ(d,a)|a∈E(i)}B*[j]+Δ, m∈BI(Ci) and d=DI(Ci); ωi(n[i])=Ai(n[i])−Δ×Σj∈{∪χ(i,a)|a∈E(i)}B*[j]; and τi0i0+Δ+Δ×Σj∈{∪χ(i,a)|a∈E(i)}B*[j]. Then, with the increased Θ and the reset ωij's, it is easy to verify that Conditions (9)-(10) are satisfied for i∈DsI(Ck).

Case 2: If Δ≠δk, let Δ=δg=minh∈Q(k)δh, Θ←Θ+Δ. Then, for ∀s∈E(k) and ∀i∈DsI(Cs), ωi0=Ai0+Δ, τi0i0+Δ, and τi(m−1)i(m−1)+Δ, m∈BI(Ci), when Ci is not a leaf tool. For ∀i∈Y(k), ωij=Aij+Δ, j∈Ωn[i]\{{j−1|j∈BI(Ci)}∪{n[i]}}, ωi(m−1)=Ai(m−1)+Δ×Σj∈∪Y(i)B*[j]+Δ, m∈BI(Ci); ωi(n[i])=Ai(n[i])−Δ×Σj∈{∪χ(i,a)|a∈E(i)}B*[j]; and τi0i0+Δ+Δ×Σj∈{∪χ(i,a)|a∈E(i)}B*[j]. By doing so, it follows from (5)-(6) that Conditions (9)-(10) are satisfied for i∈DsI(Ck). Due to δk>Δ=δg, we may have τ(k+1)0<(4λk+3μk) and Ag(n[g])=0, g∈∪χ(k+1, a), a∈E(k). In this case, let Θ←Θ+Δ. Then, repeat the above procedure for χ(k+1, g−1) until Case 1 occurs such that Δ=δk or A(k+1)(n[k+1])=0.

Finally the cycle time is increased by Δ such that τ(k+1)0=(4λk+3μk) and Conditions (9)-(10) for any pair of Ci and Ci+1, i∈N[f, l], are satisfied. In addition, since τ(k+1)0=(4λk+3μk), a cycle time increase that is less than Δ must result in τ(k+1)0<(4λk+3μk). For a CTC without downstream CTCs, a method is presented to find the minimal cycle time and its optimal one-wafer cyclic schedule in [Zhu et al., 2013b]. Since such method is somehow similar to the one disclosed herein, we do not go into too much detail and please refer to [Zhu et al., 2013b] for detail.

In the above development, each individual tool is scheduled such that, for any Ci, ωi(n[i]) is minimized. Thus, it follows from Θ12= . . . =ΘK=Θ that φi0, i∈N[f, l], in CTC[f,l] is minimized. The above development is summarized as Algorithm 1, where inputs φ and Θ≥Πh represent the index set of the individual cluster tools in a CTC and the given cycle time, respectively. Assume that all αij, λi, μi of the K-cluster tool are known constants to Algorithm 1. Its output Θ is the CTC's minimal cycle time.


Algorithm 1: Scheduling CTC[f, l] for given Θ denoted as SCH(CTC[f, l], Θ)
Input:Θ, N[f, l].
Output: ωij for Ci, i∈ N[f, l] and in the downstream CTCs, j∈Ωn[i], φf0, Θ, ε.
1.
Initialization:
1.1. Ñ← l−f+1, AD ←Υ← UI(f)
2.
Set ωij, j∈Ωn[l] for Rl as:
2.1. For j← 0 to n[l]− 1 do
2.2. ωlj← min{Θ− (4λl +3μll(j+1)),Θ− 2(n[l] + 1)(λl + μl) −Σm=0j−1 ωlm}
2.3. EndFor
2.4. ωl(n[l]) ←Θ− 2(n[l] + 1)(λl + μl) −Σm=0n[l]−1 ωlm
2.5. τl0 ←Θ− (4λl+ 3μl+ ωl(n[l])), i←l − 1, a←Ñ− 1
3.
If AD = 1 Then update ωmj(m∈ DsI(Ci)) as
3.1. For each j∈E0
3.2. ωj0 ←ωj0 + Δ, τj0←τj0 + Δ
3.3. τj(m−1)←τj(m−1) + Δ, for m∈BI(Cj)
3.4. EndFor
3.5. If E(i)≠Ø Then
3.6. For each j∈∪χ(i+1, g)|g∈E(i) do
3.7. ωjm←ωjm + Δ, for m∈Ωn[j] \ ({i−1|j∈BI(Ci)}∪{n[j]})
3.8. β←Σm∈Y(j) B*[m]
3.9. ωj(m−1) ←ωj(m−1)+ Δ×β+ Δ, for m∈BI(Cj)
3.10. ωj(n[j]) ←ωj(n[j]) −Δ×β
3.11. τj0 ←Σj0+Δ + Δ×β
3.12. EndFor
3.13. EndIf
3.14. a← |χ(f,i)|
4.
Set ωij for Ri (i∈χ(f, l−i), j∈Ωn[i]) as:
4.1. While a≥1
4.2. Ifτ(i+1)0< (4λi + 3μi) Then
4.3. Φ←4λi + 3μi −τ(i+1)0
4.4. If ω(i+1)(n[i+1])= 0 Then Δ←Φ, E(i)←Ø, E0←DsI(Ci)
4.5. Else
4.6. Find E(i)
4.7. δk ←ωk(n[k])/ Σm∈Y(k) B*[m], for each k∈Y(i)
4.8. δi←Φ/(1 + Σm∈Y(i) B*[m])
4.9. Δ←min{δi, δk }with k∈Y(i)
4.10. E0←∪DsI(Cg)| g ∈E(i)
4.11. EndIf
4.12. Θ←Θ+ Δ
4.13. AD ← 1
4.14. Goto 3.
4.15. EndIf
4.16. ωi(b[i]−1)← min{τ(i+1)0− (4λi + 3μi),Θ−2(n[i] + 1)(λi i)}
4.17. For j← 0 to n[i] and j≠b[i]−1 do
4.18. ωij← min{Θ− (4λi +3μii(j+1)),Θ− 2(n[i] + 1)(λi + μi) −Σm∈Ωj\{j} ωim }
4.19. EndFor
4.20. τi0←Θ− (4λi+ 3μi + ωi(n[i])), i← i− 1, a ←a− 1
4.21. EndWhile.
5.
Set φf0 and Θi as:
5.1. If Υ≠ 0 and 4λr + 3μr f0 Then
5.2. ←4λr + 3μr −τf0
5.3. If ωf(n[f])= 0 Then Δ←Φ, E(i)←Ø, E0←DsI(Ci)
5.4. Else
5.5. Find E(f)
5.6. δk ←ωk(n[k])/ Σm∈Y(k) B*[m], for each k∈Y(f)
5.7. δf←Φ/(1+ Σm∈Y(f) B*[m])
5.8. Δ←min{δf, δk }with k∈Y(f)
5.9. E0←∪DsI(Cf)|f∈E(f)
5.10. EndIf
5.11. Θ←Θ+ Δ, Adjusting← 1
5.12. GoTo 3
5.13. EndIf
5.14. ε←Θ−Θ0
5.15. φf0 ← 4λf+ 3μf + ωf(n[f])
6.
End

C.3. Scheduling K-Tree-Cluster Tool

As above discussed, a K-tree-cluster tool is composed of a number of CTCs, to obtain the minimal cycle time of the K-tree-cluster tool, it needs to minimize φi0=4λi+3μii(n[i]) for every Ci in each CTC, or minimize ωi(n[i]). This is done as follows. Let L-CTC be the set of CTC[f, l]s with Cl being a leaf tool of the K-tree-cluster tool and assume that |L-CTC|=g. Given Θ=Πh, the lower bound of cycle time for a K-tree-cluster tool, as the initial cycle time, each CTC in the L-CTC is scheduled by using Algorithm 1. Assume that the minimal cycle time obtained for these g CTCs are Θ1, . . . , Θg, respectively. Then, given Θ=max{Θ1, . . . , Θg} as the cycle time, for any Ci in any CTC that is in L-CTC, ωij's are set by the CTC-Check procedure. After that, given Θ=max{Θ1, . . . , Θg} as initial cycle time, we schedule the CTC∉L-CTC that is not scheduled and whose downstream CTCs are all in L-CTC. By repeating such a process, one can find a one-wafer cyclic schedule for the entire K-tree-cluster tool by minimizing φi0=4λi+3μii(n[i]) for every individual tool Ci. If finally the obtained minimal cycle time is Θ=Πh, the lower bound of cycle time is reached by a one-wafer cyclic schedule. We present Algorithm 2 as follows to implement above sketchy procedure.


Algorithm 2: Schedule a K-tree-cluster tool.
Input: αij, λi, μi (i ∈ NK,j∈Ωn[i])
Output: ωij, Θ
1. Initialization: Πh← max(Πi), i ∈ NK, Ψ←Ø, Ψ1 ←Ø
2. For each i∈ NK do
3. If (∀j∈Nn[i], αij>0) Then Ψ←Ψ∪{i} EndIf
4. EndFor
5. While Ψ≠Ø
6. For each e∈Ψ do
7.   ← {e}, m←UI(Ce)
8. While m ≠ 0 and O-Deg(Cm) = 1 do  ←  ∪{m}, m←UI(Cm)
  EndWhile
9. f← min(  )
10. i← UI(Cf), j← the step index of PMi(b[i,f]) in Ci
11. Ψ1←Ψ1∪ {i}
12. Πnew← max(Πh, max(Πm)|m∈  )
13. Call SCH(CTC[f, l], Πnew) to set φf0, Θ, εij
14. αij←φf0
15. EndFor
16. Ψ←Ø
17. For each i∈Ψ1 do
18. If ∀j∈BI(i), αij > 0 Then
19. ε← max(εij), j∈BI(i), Ψ1←Ψ1\{i}, and Ψ←Ψ∪ {i}
20. For ∀j∈BI(i) and εij
21. ωm0←ωm0 + (ε−εij), τm0←τm0 + (ε−εij), m∈DsI(Cj)
22. EndFor
23. EndIf
24. EndFor
25. EndWhile

By Algorithm 2, the CTCs in a K-tree-cluster tool are identified and they are scheduled by calling Algorithm 1 to find the minimal cycle time and set the robots' waiting time. This is done from the leaves to the root in a backward way as above discussed. Also, by Algorithm 2, if Θ=Πh is returned, it implies that τ(i+1)0<(4λi+3μi), ∀i∈NK, does not occur and a one-wafer cyclic schedule with the lower bound of cycle time is obtained. Therefore, Algorithm 2 can be used to decide whether the K-tree-cluster tool can be scheduled to reach the lower bound of cycle time. Thus, we have Theorem 2 immediately.

Theorem 2: For a process-dominant K-tree-cluster tool, there is a one-wafer cyclic schedule such that its cycle time is the lower bound Πh if and only if, by Algorithm 2, Θ=Πh is returned.

Moreover, by the proposed method, if Θ>Πh is returned, the obtained one-wafer cyclic schedule is optimal as stated by the following theorem.

Theorem 3: For a process-dominant K-tree-cluster tool, by Algorithm 2, if Θ>Πh is returned, the obtained one-wafer cyclic schedule is optimal in terms of cycle time.

Proof. By applying Algorithm 2 to a K-tree-cluster tool, every time if τ(i+1)0<(4λi+3μi), i∈NK, occurs, the cycle time Θ is increased by Δ as Θ←Θ+Δ such that τ(i+1)0=(4λi+3μi). Hence, by Algorithm 2, finally a one-wafer cyclic schedule is obtained with cycle time Θ>Πh, this implies that there is at least an i∈NK such that τ(i+1)0=(4λi+3μi). In this case, any decrease of Θ would result in τ(i+1)0=(4λi+3μi), leading to the violation of Conditions (9)-(10). Thus, Θ>Πh obtained is the minimal cycle time.

Note that, for a process-dominant K-tree-cluster tool addressed in this work, each individual tool is scheduled by using a backward scheduling strategy. This implies that a backward scheduling strategy is optimal for such a K-tree-cluster tool if the proposed method is applied.

D. Illustrative Examples

In this section, we use two examples to show how to apply the proposed method and its effectiveness. In the examples, the time is in seconds and is omitted for simplicity.

Example 1

It is from [Chan et al., 2011b]. As shown in FIG. 4, it is a 3-tree-cluster tool composed of three single-arm cluster tools, where PS10 is the LLs, PS20 and PS30 are incoming buffers for C2 and C3, and, PS12 and PS13 are outgoing buffers for C1. The activity time is as follows: for C1, (α10, α11, α12, α13, α14; λ1, μ1)=(0, 10, 0, 0, 10; 5, 2); for C2, (α20, α21, α22, α23, α24; λ2, μ2)=(0, 70, 70, 58, 5; 3, 1); and for C3, (α30, α31, α32, α33, α34; λ3, μ3)=(0, 5, 58, 70, 70; 3, 1). The wafer processing route is LL→PS11→PS12 (PS20)→PS21→PS22→PS23→PS24→PS20 (PS12)→PS13 (PS30)→PS31→PS32→PS33→PS34→PS30 (PS13)→PS14→LL. It contains three CTCs: CTC[2, 2], CTC[3, 3], and CTC[1, 1]. For this example, a multi-wafer cyclic schedule is obtained with cycle time 102.8 in [Chan et al., 2011b].

By ηijij+4λi+3μi, ψi1=2(n[i]+1)(λii), and Πi=max{ηi0, ηi1, . . . , ηi(n[i]), ψi1}, for C1, η1010+4λ1+3μ1=0+4×5+3×2=26, η11=36, η12=26, η13=26, η14=36, and ψ11=2(n[1]+1)(μ11)=2×(4+1)×(5+2)=70. We have Π1=70 and C1 is transport-bound. For C2, we have that η20=15, η21=85, η22=85, η23=73, η24=20, and ψ21=40. Thus we have Π2=85 and C2 is process-bound. For C3, we have that η30=15, η31=20, η32=73, η33=85, η34=85, and ψ31=40. Then we have Π3=85 and C3 is process-bound. Since Π231, and C2 and C3 are process-bound, the 3-tree-cluster is process-dominant.

With the lower bound of cycle time being Πh23=85, set Θ=Πh=85 as initial cycle time, we have ψ12=Θ−ψ11=15, ψ22=Θ−ψ21=40, and ψ32=Θ−ψ21=40. By Algorithm 2 that calls Algorithm 1, ωij's are set as ω2021=0, ω22=12, ω23=33, ω24=0, ω30=45, ω31323334=0, ω10111213=0 and ω14=15 such that Θ=Πh=85 is returned. Hence, a one-wafer cyclic schedule with the lower bound of cycle time Θ=85 is obtained, which verifies Theorem 2.

With this topology of this example, by changing the activity time, four other cases are generated and studied in [Chan et al., 2011b] by using the method proposed there. For each of these cases, a multi-wafer cyclic schedule is obtained. These cases are also studied by the proposed method and a one-wafer cyclic schedule is obtained for each of the cases with the cycle time being less than that obtained in [Chan et al., 2011b].

Example 2

It is a 5-tree-cluster tool composed of five single-arm cluster tools as shown in FIG. 1, where PS10 is the LLs, PS20, PS30, PS40, and PS50 are incoming buffers for C2, C3, C4, and C5, PS12 and PS13 are outgoing buffers for C1, and PS32 and PS42 are outgoing buffers for C3 and C4. The activity time is as: for C1, (α10, α11, α12, α13, α14; λ1, μ1)=(0, 10, 0, 0, 10; 5, 2); for C2, (α20, α21, α22, α23, α24; λ2, μ2)=(0, 70, 70, 58, 5; 3, 1); for C3, (α30, α31, α32, α33, α34; λ3, μ3)=(0, 34, 0, 31, 4; 8, 1); for C4, (α40, α41, α42, α43, α44; λ4, μ4)=(0, 100, 0, 100, 98; 1, 1); and for C5, (α50, α51, α52, α53, α54; λ5, μ5)=(0, 100, 100, 100, 100; 1, 1). The wafer processing route is given in Section II. It contains three CTCs: CTC[3, 5], CTC[2, 2], and CTC[1, 1].

By ηijij+4λi+3μi, ψi1=2(n[i]+1)(λii), and Πi=max{ηi0, ηi1, . . . , ηi(n[i]), ψi1}, for C1, η1010+4λ1+3μ1=0+4×5+3×2=26, η11=36, η12=26, η13=26, η14=36, and ψ11=70. We have Π1=70 and C1 is transport-bound. For C2, we have that η20=15, η21=85, η22=85, η23=73, η24=20, and ψ21=40. Then we have Π2=85 and C2 is process-bound. For C3, we have that η30=35, η31=69s, η32=35, η33=66, η34=39, and ψ31=90. Thus we have Π3=90 and C3 is transport-bound. For C4, we get η40=7s, η41=107, η42=7s, η43=107, η44=105s, and ψ41=20. Therefore, we have Π4=107 and C4 is process-bound. For C5, it is found that η50=7s, η51=107, η52=107, η53=107, η54=107, and ψ51=20. Hence we have Π5=107 and C5 is process-bound. Since Π45≥Πi, i∈N3, and C4 and C5 are process-bound, the 5-tree-cluster is process-dominant.

With the lower bound of cycle time being Π45=107, set Θ=107 as initial cycle time. By ψi2=Θ−ψi1, we have ψ12=37, ψ22=67, ψ32=17, ψ42=87, and ψ52=87. By applying Algorithm 2 to CTC[3, 5], the cycle time is increased as Θ←Θ+Δ←Θ+((4λ3+3μ3)−τ40)/(1+B*[4]+B*[5])=107+(35−21)/(1+3+3)=107+2=109. Then, Θ=109 is set for scheduling CTC[2, 2] and CTC[1, 1]. Then, given Θ=109, by applying Algorithm 2 to CTC[2, 2] and CTC[1, 1] sequentially, the cycle time Θ=109 is returned and ωij's are set as: ω30=19, ω31323334=0, ω40=2, ω41=14, ω42=2, ω43=4, ω44=67, ω50515253=2, ω54=81, ω10=39, ω11121314=0, ω20=24, ω21=22, ω22=23, ω23=0, and ω24=0. Hence, an optimal one-wafer cyclic schedule with cycle time Θ=109 is obtained.

E. The Present Invention

The present invention is developed based on the theoretical development in Sections A-C above.

An aspect of the present invention is to provide a computer-implemented method for scheduling a tree-like multi-cluster tool to generate a one-wafer cyclic schedule. The multi-cluster tool comprises K single-cluster tools Ci's. The multi-cluster tool is decomposable into a plurality of CTCs each having a serial topology.

Exemplarily, the method makes use of the finding given by Theorem 1. In particular, the method comprises determining values of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], such that (8)-(10) are satisfied for any pair of Ci and Ca, in which i, a∈NK, Ca being an immediate downstream tool of Ci.

In determining the aforesaid values of ωij, preferably a CTC-Check algorithm for computing candidate values of ωij under a constraint on Θ. The computation of the candidate ωij values is performed for at least one individual CTC. The CTC-Check algorithm for CTC[f, l] is configured to compute the candidate ωij values such that ωi(n[i]) is minimized for each value of i selected from N[f, l−1]. In one embodiment, the CTC-Check algorithm as used in the disclosed method is the one given by Section C.1.

Based on the CTC-Check algorithm, the values of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], are determined as follows according to one embodiment of the present invention. First, the candidate values of ωij for the single-cluster tools in an individual CTC are determined by performing the CTC-Check algorithm under the constraint that Θ is equal to a lower bound of cycle time of the multi-cluster tool. This lower bound is a FP of a bottleneck tool among the K single-cluster tools, and is given by Πh. If the candidate ωij values for the individual CTC are non-negative, then the candidate ωij values are set as the determined values of ωij for the individual CTC. If at least one of the candidate ωij values for the individual CTC is negative, a minimum cycle time for the individual CTC is first determined and then the values of ωij for the individual CTC are determined by performing the CTC-Check algorithm under a revised constraint that Θ is the determined minimum cycle time for the individual CTC.

More particularly, it is preferable and advantageous that the values of ωij, i=1, 2, . . . , K, and j=0, 1, . . . , n[i], are determined according to the procedure given in Section C.3, where the procedure is embodied in Algorithms 1 and 2 detailed above. A summary of this procedure is given as follows. First identify an L-CTC set consisting of first individual CTCs each having a leaf tool therein. It follows that one or more second individual CTCs not in the L-CTC set are also identified. For each of the first individual CTCs, a minimum cycle time is determined, followed by determining the values of ωij by performing the CTC-Check algorithm under the constraint that Θ is equal to the determined minimum cycle time. For each second individual CTC, candidate values of ωij for each single-cluster tool therein are determined by performing the CTC-Check algorithm under an initial constraint that Θ is equal to a maximum value of the minimum cycle times already determined for the first individual CTCs. If the candidate ωij values for the second individual CTC are non-negative, then the candidate ωij values are set as the determined values of ωij for the second individual CTC. If, on the other hand, at least one of the candidate ωij values for the second individual CTC is negative, then the following two-step procedure is performed. First, determine a minimum cycle time for the second individual CTC. Second, determine the values of ωij for the second individual CTC by performing the CTC-Check algorithm under a revised constraint that Θ is equal to the determined minimum cycle time for the second individual CTC.

The embodiments disclosed herein may be implemented using general purpose or specialized computing devices, computer processors, or electronic circuitries including but not limited to digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA), and other programmable logic devices configured or programmed according to the teachings of the present disclosure. Computer instructions or software codes running in the general purpose or specialized computing devices, computer processors, or programmable logic devices can readily be prepared by practitioners skilled in the software or electronic art based on the teachings of the present disclosure.

In particular, the method disclosed herein can be implemented in a tree-like multi-cluster cluster tool if the multi-cluster tool includes one or more processors. The one or more processors are configured to execute a process of generating a one-wafer cyclic schedule according to one of the embodiments of the disclosed method.

The present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiment is therefore to be considered in all respects as illustrative and not restrictive. The scope of the invention is indicated by the appended claims rather than by the foregoing description, and all changes that come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

Read more
PatSnap Solutions

Great research starts with great data.

Use the most comprehensive innovation intelligence platform to maximise ROI on research.

Learn More

Patent Valuation

$

Reveal the value <>

28.0/100 Score

Market Attractiveness

It shows from an IP point of view how many competitors are active and innovations are made in the different technical fields of the company. On a company level, the market attractiveness is often also an indicator of how diversified a company is. Here we look into the commercial relevance of the market.

25.0/100 Score

Market Coverage

It shows the sizes of the market that is covered with the IP and in how many countries the IP guarantees protection. It reflects a market size that is potentially addressable with the invented technology/formulation with a legal protection which also includes a freedom to operate. Here we look into the size of the impacted market.

75.0/100 Score

Technology Quality

It shows the degree of innovation that can be derived from a company’s IP. Here we look into ease of detection, ability to design around and significance of the patented feature to the product/service.

44.0/100 Score

Assignee Score

It takes the R&D behavior of the company itself into account that results in IP. During the invention phase, larger companies are considered to assign a higher R&D budget on a certain technology field, these companies have a better influence on their market, on what is marketable and what might lead to a standard.

15.0/100 Score

Legal Score

It shows the legal strength of IP in terms of its degree of protecting effect. Here we look into claim scope, claim breadth, claim quality, stability and priority.

Citation

Title Current Assignee Application Date Publication Date
Scheduler for processing system ASM IP HOLDING B.V. 29 November 2012 29 May 2014
Method of operating a lithographic processing machine, control system, lithographic apparatus, lithographic processing cell, and computer program ASML NETHERLANDS B.V. 30 September 2004 12 May 2005
Scheduling Start-up Process for Time-constrained Single-arm Cluster Tools MACAU UNIVERSITY OF SCIENCE AND TECHNOLOGY 20 October 2015 23 March 2017
Method and apparatus for scheduling wafer processing within a multiple chamber semiconductor wafer processing tool having a multiple blade robot APPLIED MATERIALS, INC. 10 February 2000 01 May 2001
System and method for determining an optimized schedule of a production line MACAU UNIVERSITY OF SCIENCE AND TECHNOLOGY 22 October 2015 27 April 2017
See full citation <>

More like this

Title Current Assignee Application Date Publication Date
Production planning system and method TRUMPF WERKZEUGMASCHINEN GMBH + CO. KG 10 March 2017 21 September 2017
Device real-time status-based material scheduling method and system BEIJING NMC CO., LTD. 30 December 2015 22 June 2017
Power tool drop arm with offset ribbing ROBERT BOSCH GMBH,DOUMANI, ROBERT,SZWEDA, TIMOTHY,TALESKY, MARK,TAYLOR, BRIAN 11 March 2016 15 September 2016
デュアル脱気/冷却装填ロッククラスタツール ティーガル コーポレイション 02 July 2001 17 October 2012
Wafer processing tool having a micro sensor APPLIED MATERIALS, INC. 19 January 2017 14 September 2017
Caliper tool with positive pivoting arm PROBE HOLDINGS, INC. 29 April 2016 10 November 2016
Method of managing electricity providing in a computers cluster BULL SAS 16 June 2016 21 December 2017
System and method for self-contained self-calibrating modular manufacturing tool MOBILE TOOL MANAGEMENT, INC. 22 September 2016 13 April 2017
Fast generation adjustment algorithm for energy management system THE UNIVERSITY OF HONG KONG 13 November 2014 19 May 2016
Scheduled universal serial bus (USB) low-power operations QUALCOMM INCORPORATED 01 April 2016 10 November 2016
Systems and methods for detecting the existence of one or more environmental conditions within a substrate processing system APPLIED MATERIALS, INC. 25 January 2017 03 August 2017
End of arm tooling MAGNA INTERNATIONAL, INC.,DUCLOS, DONALD,INGRAM, JOHN,WANG, ZONGXUN 01 April 2015 08 October 2015
Robotic arm with a tool interface comprising an electronically controllable tool attachment AUTOMATA TECHNOLOGIES LIMITED,CHANDRA, SURYANSH,ELSAYED, MOSTAFA 07 October 2016 13 April 2017
Method and device for determining an energy-efficient operating point ZF FRIEDRICHSHAFEN AG 23 May 2016 29 December 2016
Methods & apparatus for controlling an industrial process ASML NETHERLANDS B.V. 21 September 2016 13 April 2017
See all similar patents <>

PatSnap Solutions

PatSnap solutions are used by R&D teams, legal and IP professionals, those in business intelligence and strategic planning roles and by research staff at academic institutions globally.

PatSnap Solutions
Search & Analyze
The widest range of IP search tools makes getting the right answers and asking the right questions easier than ever. One click analysis extracts meaningful information on competitors and technology trends from IP data.
Business Intelligence
Gain powerful insights into future technology changes, market shifts and competitor strategies.
Workflow
Manage IP-related processes across multiple teams and departments with integrated collaboration and workflow tools.
Contact Sales