informs_poster_v9_sg
- 1. RESEARCH POSTER PRESENTATION DESIGN © 2015 www.PosterPresentations.com A1 G1 H1 A M1 AL 1 AK 1 A3 F3 E3B3 C3 D3 A2 F2 E2 B2 C2D2 A4 G4 C4 B4 F4E4 B1 F1 C1 E1 K2 J2 I2 H2 AJ 4 AK 4 AL 4 A M4 AH 3 AI3 AG 3 AK 3 T1 AT 1 AI4 AH 4 AF 3 AG 4 AB 3 AB 4 AA 3 AA 4 AC 3 AD 4 Y3 AC 4 AD 3 AE 4 AF 4 AE 3 AE 3 AG 4 AH 4 AF 3AN 1 AA 2 S1 D1 L2 M2 AP 1 N2 R1 O2 Q1 P2 T1 Z2 U1 Y2 G3 J3 H3 I3 AH 1 AG 1 I1 J1 AL 2 AK 2 A M2 AJ 2 L4 K4 D4 H4 J4 I4 AH 2 AI2 Y4 Z4 Q2 X2 X3 Z3 P1 W 1 K3 L3 K1 AF 1 AJ 1 AI1 AN 2 A M2 0.00065 0.0006 0.00037 0.00036 0.0017 0.0012 0.0012 4.115m 0.0003 0.0003 0.0003 0.0003 0.0003 0.0005 0.0006 0.0003 0.0017 0.00038 0.0015 0.0005 0.0006 0.0006 0.0006 0.0025 0.0014 0.00140.0013 0.0006 0.0006 0.0005 0.0005 0.0009 0.0004 0.0010 0.0003 0.0003 0.0003 0.0003 0.0010 0.0003 0.0003 0.0003 0.0009 0.0011 0.0009 0.0005 0.0005 0.0003 0.0011 0.0003 0.0003 0.0003 0.0003 0.00090.0009 0.0015 0.0010 0.0010 0.0003 0.0003 0.0003 0.0003 0.0003 0.0004 0.0009 0.0003 0.0003 0.0003 0.0003 0.0013 0.0003 0.0003 0.0006 0.0003 0.0003 0.0003 0.0003 0.0003 0.00036 0.0037 0.0006 0.00065 0.00038 0.0017 0.0015 0.0005 0.0005 0.00060.0006 0.0014 0.0025 0.0014 0.0005 0.0005 0.0015 0.0017 0.0006 0.00065 0.00060.0006 0.00038 0.0015 0.00038 0.00065 0.0006 0.00037 0.00037 0.0002 0.0003 0.0003 0.0003 0.0003 0.0006 0.0012 0.0006 0.0011 0.0003 0.0003 0.0003 0.0003 0.0005 0.0003 0.00036 0.0017 0.00036 0.0003 0.0003 0.0017 L1 AD 1 M1 AB 1 N1 AA 1 O1 X1 0.0003 0.0001 0.0005 0.0008 0.00080.0008 0.0008 0.0001 0.0006 AE 1 AC 1 Z1 Y10.0013 0.00130.00130.0013 0.0007 0.0004 0.0005 0.00050.0005 0.0004 0.00070.0007 0.0003 0.0015 BAY11 0.0005 BAY12 0.0006 BAY13 0.0006 BAY14 0.0001 BAY15 BAY16 BAY17 BAY180.0006 0.0006 0.0001 0.0003 BAY116 BAY115 BAY114 BAY113 BAY112 BAY111 BAY110 BAY190.0003 0.0005 0.0009 0.0006 0.0004 0.0006 0.0006 0.0005 M3 N3 Q3 R3 S3 T3 W 3 V3 0.0003 0.0001 0.0005 0.0008 0.00080.0008 0.0008 0.0001 0.0006 P3 O3 V3 U30.0013 0.00130.00130.0013 0.0007 0.0004 0.0005 0.00050.0005 0.0004BAY31 0.0003 BAY32 0.0006 BAY35 0.0006 BAY36 0.0001 BAY310 BAY312 BAY313 BAY3150.0006 0.0006 0.0001 0.0003 BAY33 BAY34 BAY37 BAY38 BAY39 BAY311 BAY314 BAY3160.0003 0.0005 0.0009 0.0006 0.0004 0.0006 0.00060.0005 M4 N4 R4 Q4 S4 T4 W 4 X4 0.00030.00010.0005 0.00080.0008 0.00080.0008 0.0001 0.0006 O4 P4 V4 U4 0.00130.0013 0.0013 0.0013 0.00070.0004 0.0005 0.0005 0.0005 0.0004 BAY4160.0005BAY4140.0006BAY4120.0006BAY4110.0001BAY47BAY45BAY44 BAY42 0.00060.00060.0001 0.0003 BAY415BAY413BAY410BAY49BAY48BAY46BAY43BAY41 0.0003 0.00050.00090.00060.00040.00060.0006 0.0005 AG 2 AF 2 AB 2 AC 2 U2 T2 V2 W 2 0.00030.00010.0005 0.00080.0008 0.00080.0008 0.0001 0.0006 AE 2 AD 2 R2 S2 0.00130.0013 0.0013 0.0013 0.00070.00040.0005 0.0005 0.0005 0.0004 BAY2160.0005BAY2130.0006BAY2120.0006BAY2110.0001BAY27BAY26BAY24BAY22 0.00060.00060.00010.0003 BAY215BAY214BAY210BAY29BAY28BAY25BAY23BAY21 0.00030.00050.00090.00060.00040.00060.0006 0.0005 AQ 1 0.0004 AR 1 0.0003 A M3 AL 3 0.0003 0.0003 0.0004 AO 4 AN 4 0.0003 0.0003 0.0004 AO 2 AN 2 0.0003 0.0003 0.0004 AQ 2 AP 2 0.0003 0.0003 0.0004 AQ 4 AP 4 0.0003 0.0003 0.0004 AO 3 AN 3 0.0003 0.0003 0.0004 AS 1 AT 1 0.0003 0.0003 0.0004 D1D1 0.0003 0.0003 AT 1 0.0003 T1 0.0003 48 secs i. Linear (current practice): 𝑒𝑒 𝑛𝑛 = 𝑛𝑛 � 10 + 𝑡𝑡 ii. Piece-wise Linear: 𝑒𝑒 𝑛𝑛 = � 𝑛𝑛 � 10 + 𝑡𝑡, 𝑛𝑛 � 20 + (𝑡𝑡−10) � 𝑘𝑘, iii. Big M: 𝑒𝑒 𝑛𝑛 = � 𝑛𝑛 � 10 + 𝑡𝑡, 𝑀𝑀, n – number of vehicles on an edge t – time taken to travel an empty edge e(n) – time to traverse jammed edge with n vehicles •Routing of automated material handling systems (AMHS) vehicles in wafer fabs. •Increasing number of FOUP pickup/ delivery requests are causing: Increase in number of vehicles Increase in vehicle congestion Increase in wait time for FOUPs 1. Homogeneous Simulations: Equally probable pickup/delivery requests. 2. Heterogeneous Simulations: Doubly probable pickup/delivery from any one bay. Pickup Protocols continued: c. Reservation with ETA Policy: Every scheduling cycle, busy vehicles with estimated time of arrival under some threshold, get reserved by lots in queue that are nearest to their current location (via current destination). Any idle reserved vehicle is re-reserved by nearer lots in queue. d. Reservation without ETA Policy: Same as Reservation with ETA except there’s no ETA threshold. INTRODUCTION TOOLS OBJECTIVES •15 server (vehicle) stylized system •Stable pickup request arrival rate - determined by trial and error •Stability - vehicle utilization ≈98% •Graphical system stability verification: a. FIFO: Idle vehicle picks up a lot from the queue that has waited the longest. b. Nearest Lot Policy: Pickup lot that’s nearest in terms of the time an idle vehicle will take to travel to it. METHODOLOGY (1/3) METHODOLOGY (3/3) ONGOING WORK •Python – simulate & test algorithms •AutoCAD – design stylized fab •Microsoft Visio – - Toy lattice networks for building & testing simulator: - Network representation of stylized fab METHODOLOGY (2/3) John J Hasenbein a, Shreya Gupta b a: Associate Professor, jhas@mail.utexas.edu; b: PhD Candidate, shreya.gupta@utexas.edu; OR/IE Group, Department of Mechanical Engineering, The University of Texas at Austin Improving Scheduling and Control of the OHTC Controller in Wafer Fab AMHS Systems Lot 10, 40 secs Lot 11, 10 secs Lot 12, 33 secs Lot n, 27 secs (Lots in queue Time from vehicle 3) , Sequence in which vehicles become idle 3 8 5 7 Lot 10, NA Lot 11, NA Lot 12, NA Lot n, NA (Lots in queue Time from vehicle 3) , Sequence in which vehicles become idle 3 8 5 7 Lot 10, 300 secs Lot 11, 10 secs Lot n, 27 secs (Lots in queue Time from vehicle 3) , Sequence in which vehicles become idle 3 8 7 1st reservation of vehicles with ETA under 4 mins ETA, 3.3 mins ETA, 3.7 mins ETA, 1.0 min Scheduling Cycle 1 (Cycle time = 3mins) Lot 10, 28 secs Lot 11, reserved Lot n, 5 secs (Lots in queue Time from vehicle 8) , Sequence in which vehicles become idle 3 8 7 2nd reservation of vehicles with ETA under 4 mins ETA, reserved ETA, 3.7 mins ETA, 1.0 min Lot 10, 400 secs Lot 11, 10 secs Lot 12, 330 secs Lot n, 27 secs (Lots in queue Time from vehicle 3) , Sequence in which vehicles become idle 3 1 9 7 1st reservation of vehicles with no ETA threshold Scheduling Cycle 1 (Cycle time = 3mins) Lot 13, 120 secs 8 Additional vehicles available for reservation between vehicles 3 & 8 No. of vehicles threshold Time No. of vehicles threshold Time Bay 1 25% Bay 2 25% Bay 4 25% Bay 3 25% Bay 1 50% Bay 2 16.67% Bay 3 16.67% Bay 4 16.67% PICKUP PROTOCOLS PARAMETER ESTIMATION ROUTING ALGORITHMS Symmetric & unidirectional Asymmetric & bidirectional No. of vehicles threshold M Time 𝑛𝑛 ≤ 𝑘𝑘 𝑛𝑛 > 𝑘𝑘 𝑛𝑛 ≤ 𝑘𝑘 𝑛𝑛 > 𝑘𝑘 ∀ 𝑛𝑛 B C A D B Sub-optimal route decided using the actual distance between nodes D & B. Optimal route decided using state dependent time (or cost) between nodes D & B. From To A B 18 secs 0 B C 18 secs 0 C D 18 secs 0 D B 18 secs 3 B A 18 secs 1 Network State Edges Edge Travel Time 48 secs 36 secs 18 secs State Dependent Edge Travel Time 18 secs 18 secs A D C 36 secs 0.035 0.016 A B C D H G F E I J K L 0.005 0.015 0.007 0.019 0.013 0.010 0.006 0.005 0.023 0.002 0.007 0.012 0.014 0.020 0.007 0.014 0.013 0.004 0.022 0.0140.014 0.0030.011 0.0020.008 0.005 0.0120.020 0.0250.013 0.0080.015 A B C D H G F E I J K L 0.005 0.005 0.005 0.0050.005 0.005 0.0050.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 FAB CONTROLLER MCS OHTC 1 2 N •Improve throughput via better routing algorithms for the over hoist transport control (OHTC) system. •Enhance state dependent cost functions used in Dijkstra's algorithm. State of the network at time t is the number of vehicles on the various edges of the network. vehicles