ºÝºÝߣ

ºÝºÝߣShare a Scribd company logo
MODELING AND ANALYSIS OF
MANUFACTURING SYSTEMS
Session 7
FLEXIBLE
MANUFACTURING
SYSTEMS
E. Gutierrez-Miravete
Spring 2001
DEFINITION
A FLEXIBLE MANUFACTURING
SYSTEM (FMS) IS A SET OF
NUMERICALLY CONTROLLED
MACHINE TOOLS AND SUPPORTING
WORKSTATIONS CONNECTED BY AN
AUTOMATED MATERIAL HANDLING
SYSTEM AND CONTROLLED BY A
CENTRAL COMPUTER
ELEMENTS OF FMS
• AUTOMATICALLY
REPROGRAMMABLE MACHINES.
• AUTOMATED TOOL DELIVERY AND
CHANGING
• AUTOMATED MATERIAL HANDLING
• COORDINATED CONTROL
FMS FEATURES
• MANY PART TYPES CAN BE LOADED
• PARTS CAN ARRIVE AT MACHINES
IN ANY SEQUENCE
• PARTS IDENTIFIED BY CODES
• MANY MACHINES CAN BE INCLUDED
• SMALL FMS LEAD TO FLEXIBLE
CELLS
FMS FEATURES
• EXPENSIVE TO IMPLEMENT BUT
SAVINGS CAN BE SIGNIFICANT
• FLOOR SPACE REDUCIBLE BY 1/3
• EQUIPMENT UTILIZATION UP TO 85%
OR MORE
• DETAILED PRODUCTION SEQUENCE
NOT NEEDED WELL IN ADVANCE
FMS FEATURES
• REDUCED VARIABLE COSTS AND
THROUGHPUT TIME LEAD TO
ENHANCED MANUFACTURING
COMPETITIVENESS
• ELIMINATION OF STARTUP CYCLES
LEAD TO STANDARIZED
PERFORMANCE
• MODULAR DESIGN
FMS FEATURES
• REDUCED DIRECT LABOR COSTS
• THREE SHIFTS READILY FEASIBLE
• IDEAL FOR JIT
• CAN EASILY BE TURNED OVER TO
NEW SET OF PRODUCTS IF THE NEED
ARISES
MANUFACTURING
FLEXIBILITY
• BASIC
– MACHINE (VARIETY OF OPERATIONS)
– MATERIAL HANDLING (PART MOBILITY
AND PLACEMENT)
– OPERATION (VARIETY OF OPERATIONS
PRODUCING SAME PART FEATURES)
MANUFACTURING
FLEXIBILITY
• SYSTEM
– PROCESS (VARIETY OF PARTS
PRODUCIBLE WITH SAME SETUP)
– ROUTING (ABILITY TO USE DIFFERENT
MACHINES UNDER SAME SETUP)
– PRODUCT (CHANGEOVER)
– VOLUME (PRODUCTION LEVEL)
– EXPANSION (ADDED CAPACITY)
MANUFACTURING
FLEXIBILITY
• AGGREGATED
– PROGRAM (UNATTENDED RUNNING)
– PRODUCTION (RANGES OF PARTS,
PRODUCTS, PROCESSES, VOLUME,
EXPANSION)
– MARKET (COMBINATION OF PRODUCT,
PROCESS, VOLUME AND EXPANSION)
COMMENTS
DOES FLEXIBILITY REMOVE
VARIABILITY FROM THE
SYSTEM?
NO, BUT IT ENABLES IT TO
PERFORM EFFECTIVELY
COMMENTS
KEY ISSUE
CAN A SYSTEM BE DESIGNED WHICH
IS USEFUL OVER A SUFFICIENT TIME
HORIZON, PART MIX AND SMALL
CHANGEOVER TIMES SO AS TO
OFFER AN ALTERNATIVE TO
SIMULTANEOUS PRODUCTION OF
MEDIUM VOLUME PART TYPES?
COMMENTS
THE PART TYPES ASSIGNED TO THE
FMS SHOULD HAVE SUFFICIENT
PRODUCTION VOLUMES TO MAKE
AUTOMATION ATTRACTIVE BUT
INSUFFICIENT TO JUSTIFY
DEDICATED PRODUCTION LINES
ORIGINS OF FMS
• LINK LINES (1960’S)
• NC MACHINES AND
CONVEYORS
• BATCH PROCESSING
ACRONYMS
• FMS
• NC
• DNC
• CNC
• AGV
• JIT
FMS PRIORITIES
• MEETING DUE DATES
• MAXIMIZING MACHINE
UTILIZATION
• MINIMIZE THROUGHPUT
TIMES
• MINIMIZE WIP LEVELS
FMS COMPONENTS
• MACHINES
• PART MOVEMENT SYSTEMS
• SUPPORTING
WORKSTATIONS
• SYSTEM CONTROLLER
MACHINES
• PRISMATIC VS ROTATIONAL
PARTS
• HORIZONTAL MACHINING
CENTERS (HMC) AND HEAD
INDEXERS (HI)
• TOOL MAGAZINES AND
AUTOMATIC TOOL CHANGERS
PART MOVEMENT
• CONVEYORS
• TOW CARTS
• RAIL CARTS
• AGV’S
SUPPORTING
WORKSTATIONS
• LOAD/UNLOAD STATIONS
• AUTOMATIC PART WASHERS
• COORDINATE MEASURING
MACHINES
CONTROLLER
• COMPUTER
• WORKER (ATTENDANT)
• TRACKING SYSTEM FOR
–PARTS
–MACHINES
PLANNING AND CONTROL
HIERARCHY
DECISION MAKING PROCESS
–WHICH INFORMATION SHOULD
BE COMMUNICATED?
–HOW DO SYSTEM
COMPONENTS
COMMUNICATE?
COMPONENTS OF THE
MANUFACTURING
FACILITY
– FACILITY
– SHOP
– CELL
– WORKSTATION
– EQUIPMENT
MULTILEVEL CONTROL
HIERARCHY
• TREE STRUCTURE OF THE
HIERARCHY
• INFORMATION FLOWS ONLY
BETWEEN ADJACENT LAYERS
• EACH LEVEL HAS ITS OWN
PLANNING HORIZON AND DECISION
TYPES
• Fig. 5.5 and Table 5.1 , p. 133
GENERIC CONTROL MODEL
• GENERIC CONTROL STRUCTURE USED
TO ACCOMPLISH PLANNING,
EXECUTION AND FEEDBACK
• COMMANDS ARE RECEIVED FROM THE
NEXT HIGHER LEVEL AND TASKS ARE
BROKEN INTO SUBTASKS
• SUBTASKS ARE ASSIGNED TO
COMPONENTS AT NEXT LOWER LEVEL
GENERIC CONTROL MODEL
• SUBTASK MONITORING PERFORMED
THROUGH RECEIPT OF STATUS
FEEDBACK FROM LOWER LEVEL
• TASK STATUS INFORMATION RELAYED
TO NEXT HIGHER LEVEL
• EACH CONTROLLER HAS A
PRODUCTION MANAGER RECEIVING
COMMANDS AND SCHEDULING TASKS
GENERIC CONTROL MODEL
• QUEUE MANAGER MAINTAINED FOR
EACH LOWER LEVEL COMPONENTS TO
MANAGE ASSIGNED SUBTASKS
• DISPATCH MANAGER RECEIVES
DISPATCH ORDERS AND MANAGES
SUBTASK EXECUTION FOR EACH QUEUE
MANAGER
• Fig. 5.6, p. 134
BASIC STEPS IN DECISION
HIERARCHY
• LONG TERM PLANNING OR SYSTEM
DESIGN (PART TYPES & EQUIPMENT
SELECTION)
• MEDIUM RANGE PLANNING OR
SETUP (DAILY DECISIONS ABOUT
PARTS & TOOLING)
• SHORT TERM OPERATION
(SCHEDULING & CONTROL)
SYSTEM DESIGN
• PROBLEM: SELECTING SYSTEM SIZE,
HARDWARE, SOFTWARE AND PARTS
FOR THE FMS
• SIZE & SCOPE ARE SELECTED
ACCORDING TO CORPORATE
STRATEGY
• HARDWARE & SOFTWARE SELECTED
TO FIT SCOPE
SYSTEM DESIGN
• PART SELECTION IS DONE
ACCORDING TO AN ECONOMIC
CRITERION & STRATEGIC
CONSIDERATIONS
• KNAPSACK PROBLEM: LOAD THE
FMS TO MAXIMIZE SAVINGS
SUBJECT TO FMS CAPACITY
KNAPSACK PROBLEM
P = PRODUCTIVE TIME PER PERIOD
AVAILABLE ON BOTTLENECK FMS
RESOURCE
pi= TIME PER PERIOD REQUIRED FOR
PART i
si= SAVINGS PER PERIOD IF PART TYPE
i
KNAPSACK PROBLEM
maximize Σi si Xi
subjectto
Σi pi < P
SOLVING THE KNAPSACK
PROBLEM
• GREEDY HEURISTIC
• Example 5.1, p. 136
• OPTIMIZATION
• Example 5.2, p. 138
SYSTEM SETUP
• ASSIGNMENT OF OPERATIONS AND
ACCOMPANYING TOOLING TO
MACHINES
• PART SELECTION PROBLEM: BATCH
FORMATION
• LOADING PROBLEM: SEQUENCING
AND ROUTING OF PARTS
PART SELECTION
• GOAL: PLACE REQUIRED PARTS INTO
COMPATIBLE BATCHES SUCH THAT
• EACH BATCH USES ALL MACHINES
• REQUIRE A LIMITED NUMBER OF
TOOLS ON EACH MACHINE
• HAVE SIMILAR DUE DATES FOR
PARTS IN THE BACTH
PART SELECTION
• GREEDY HEURISTIC: FORM BATCHES
BY ARRANGING PART ORDERS BY
DUE DATES
• PART ORDERS ARE SEQUENTIALLY
ADDED TO CURRENT BATCH
WITHOUT VIOLATING CONSTRAINTS
• BATCH IS THEN READY FOR LOADING
• Example 5.3, p. 140
Part Selection as a Mixed-Integer
Program
• Time phased set of part orders Dit for part
i in time t
• Time available in machine j , Pj
• Time required by product i in machine j
pij
• Number of parts of type i made in time t
xit
• Number of tool slots in machine j , Kj
Part Selection as a Mixed-Integer
Program
• Number of tool slots required by tool l in
machine j , klj
• Set of tools l required on machine j to
produce part i , l j(i)
• Holding cost per period for part i hi
• Formulation: p. 142
Part Selection as a Mixed-Integer
Program
• Goal: Minimize inventory cost while
meeting due dates
• Example 5.4 , p. 142
Incremental Part Selection
• Several part types in process at any time
• System operates almost continuously
• Goal: Minimize makespan to complete all
available part orders
• Procedure: Minimize idle time by balancing
work loads subject to part demand and tool
magazine capacity
• Formulation: p. 144
LOADING PROBLEM
• BATCH TO BE PROCESSED IS KNOWN
• OBJECTIVES REQUIRED
• LOADING SOLUTION MUST BE
ROBUST AND FLEXIBLE
• SOLUTION METHODOLOGIES
– MATHEMATICAL PROGRAMMING (p.145)
– HEURISTIC APPROACHES (p. 148)
LOADING PROBLEM:
HEURISTIC APPROACH
• PHASE I : ASSIGN OPERATIONS TO
MACHINE TYPES
• PHASE II:
– OPERATIONS COMBINED INTO
CLUSTERS TO REDUCE TRANSFERS
– MACHINE GROUPS FORMED
– OPERATIONS AND TOOLS ASSIGNED
TO GROUPS
SCHEDULING AND
CONTROL
• BASIC PROBLEM AREAS
– SEQUENCING AND TIMING OF PART
RELEASES TO THE SYSTEM
– SETTING OF INTERNAL PRIORITIES IN
THE SYSTEM
– ABILITY OF SYSTEM TO TAKE
CORRECTIVE ACTION WHEN
COMPONENTS FAIL
Flexible Assembly Systems
• For the combination of raw materials and
components into products with functional
characteristics.
• Automated vs manned systems
• Example: Vibratory bowl feeders and vision
systems
• Role of Design for Assembly

More Related Content

Mams07

  • 1. MODELING AND ANALYSIS OF MANUFACTURING SYSTEMS Session 7 FLEXIBLE MANUFACTURING SYSTEMS E. Gutierrez-Miravete Spring 2001
  • 2. DEFINITION A FLEXIBLE MANUFACTURING SYSTEM (FMS) IS A SET OF NUMERICALLY CONTROLLED MACHINE TOOLS AND SUPPORTING WORKSTATIONS CONNECTED BY AN AUTOMATED MATERIAL HANDLING SYSTEM AND CONTROLLED BY A CENTRAL COMPUTER
  • 3. ELEMENTS OF FMS • AUTOMATICALLY REPROGRAMMABLE MACHINES. • AUTOMATED TOOL DELIVERY AND CHANGING • AUTOMATED MATERIAL HANDLING • COORDINATED CONTROL
  • 4. FMS FEATURES • MANY PART TYPES CAN BE LOADED • PARTS CAN ARRIVE AT MACHINES IN ANY SEQUENCE • PARTS IDENTIFIED BY CODES • MANY MACHINES CAN BE INCLUDED • SMALL FMS LEAD TO FLEXIBLE CELLS
  • 5. FMS FEATURES • EXPENSIVE TO IMPLEMENT BUT SAVINGS CAN BE SIGNIFICANT • FLOOR SPACE REDUCIBLE BY 1/3 • EQUIPMENT UTILIZATION UP TO 85% OR MORE • DETAILED PRODUCTION SEQUENCE NOT NEEDED WELL IN ADVANCE
  • 6. FMS FEATURES • REDUCED VARIABLE COSTS AND THROUGHPUT TIME LEAD TO ENHANCED MANUFACTURING COMPETITIVENESS • ELIMINATION OF STARTUP CYCLES LEAD TO STANDARIZED PERFORMANCE • MODULAR DESIGN
  • 7. FMS FEATURES • REDUCED DIRECT LABOR COSTS • THREE SHIFTS READILY FEASIBLE • IDEAL FOR JIT • CAN EASILY BE TURNED OVER TO NEW SET OF PRODUCTS IF THE NEED ARISES
  • 8. MANUFACTURING FLEXIBILITY • BASIC – MACHINE (VARIETY OF OPERATIONS) – MATERIAL HANDLING (PART MOBILITY AND PLACEMENT) – OPERATION (VARIETY OF OPERATIONS PRODUCING SAME PART FEATURES)
  • 9. MANUFACTURING FLEXIBILITY • SYSTEM – PROCESS (VARIETY OF PARTS PRODUCIBLE WITH SAME SETUP) – ROUTING (ABILITY TO USE DIFFERENT MACHINES UNDER SAME SETUP) – PRODUCT (CHANGEOVER) – VOLUME (PRODUCTION LEVEL) – EXPANSION (ADDED CAPACITY)
  • 10. MANUFACTURING FLEXIBILITY • AGGREGATED – PROGRAM (UNATTENDED RUNNING) – PRODUCTION (RANGES OF PARTS, PRODUCTS, PROCESSES, VOLUME, EXPANSION) – MARKET (COMBINATION OF PRODUCT, PROCESS, VOLUME AND EXPANSION)
  • 11. COMMENTS DOES FLEXIBILITY REMOVE VARIABILITY FROM THE SYSTEM? NO, BUT IT ENABLES IT TO PERFORM EFFECTIVELY
  • 12. COMMENTS KEY ISSUE CAN A SYSTEM BE DESIGNED WHICH IS USEFUL OVER A SUFFICIENT TIME HORIZON, PART MIX AND SMALL CHANGEOVER TIMES SO AS TO OFFER AN ALTERNATIVE TO SIMULTANEOUS PRODUCTION OF MEDIUM VOLUME PART TYPES?
  • 13. COMMENTS THE PART TYPES ASSIGNED TO THE FMS SHOULD HAVE SUFFICIENT PRODUCTION VOLUMES TO MAKE AUTOMATION ATTRACTIVE BUT INSUFFICIENT TO JUSTIFY DEDICATED PRODUCTION LINES
  • 14. ORIGINS OF FMS • LINK LINES (1960’S) • NC MACHINES AND CONVEYORS • BATCH PROCESSING
  • 15. ACRONYMS • FMS • NC • DNC • CNC • AGV • JIT
  • 16. FMS PRIORITIES • MEETING DUE DATES • MAXIMIZING MACHINE UTILIZATION • MINIMIZE THROUGHPUT TIMES • MINIMIZE WIP LEVELS
  • 17. FMS COMPONENTS • MACHINES • PART MOVEMENT SYSTEMS • SUPPORTING WORKSTATIONS • SYSTEM CONTROLLER
  • 18. MACHINES • PRISMATIC VS ROTATIONAL PARTS • HORIZONTAL MACHINING CENTERS (HMC) AND HEAD INDEXERS (HI) • TOOL MAGAZINES AND AUTOMATIC TOOL CHANGERS
  • 19. PART MOVEMENT • CONVEYORS • TOW CARTS • RAIL CARTS • AGV’S
  • 20. SUPPORTING WORKSTATIONS • LOAD/UNLOAD STATIONS • AUTOMATIC PART WASHERS • COORDINATE MEASURING MACHINES
  • 21. CONTROLLER • COMPUTER • WORKER (ATTENDANT) • TRACKING SYSTEM FOR –PARTS –MACHINES
  • 22. PLANNING AND CONTROL HIERARCHY DECISION MAKING PROCESS –WHICH INFORMATION SHOULD BE COMMUNICATED? –HOW DO SYSTEM COMPONENTS COMMUNICATE?
  • 23. COMPONENTS OF THE MANUFACTURING FACILITY – FACILITY – SHOP – CELL – WORKSTATION – EQUIPMENT
  • 24. MULTILEVEL CONTROL HIERARCHY • TREE STRUCTURE OF THE HIERARCHY • INFORMATION FLOWS ONLY BETWEEN ADJACENT LAYERS • EACH LEVEL HAS ITS OWN PLANNING HORIZON AND DECISION TYPES • Fig. 5.5 and Table 5.1 , p. 133
  • 25. GENERIC CONTROL MODEL • GENERIC CONTROL STRUCTURE USED TO ACCOMPLISH PLANNING, EXECUTION AND FEEDBACK • COMMANDS ARE RECEIVED FROM THE NEXT HIGHER LEVEL AND TASKS ARE BROKEN INTO SUBTASKS • SUBTASKS ARE ASSIGNED TO COMPONENTS AT NEXT LOWER LEVEL
  • 26. GENERIC CONTROL MODEL • SUBTASK MONITORING PERFORMED THROUGH RECEIPT OF STATUS FEEDBACK FROM LOWER LEVEL • TASK STATUS INFORMATION RELAYED TO NEXT HIGHER LEVEL • EACH CONTROLLER HAS A PRODUCTION MANAGER RECEIVING COMMANDS AND SCHEDULING TASKS
  • 27. GENERIC CONTROL MODEL • QUEUE MANAGER MAINTAINED FOR EACH LOWER LEVEL COMPONENTS TO MANAGE ASSIGNED SUBTASKS • DISPATCH MANAGER RECEIVES DISPATCH ORDERS AND MANAGES SUBTASK EXECUTION FOR EACH QUEUE MANAGER • Fig. 5.6, p. 134
  • 28. BASIC STEPS IN DECISION HIERARCHY • LONG TERM PLANNING OR SYSTEM DESIGN (PART TYPES & EQUIPMENT SELECTION) • MEDIUM RANGE PLANNING OR SETUP (DAILY DECISIONS ABOUT PARTS & TOOLING) • SHORT TERM OPERATION (SCHEDULING & CONTROL)
  • 29. SYSTEM DESIGN • PROBLEM: SELECTING SYSTEM SIZE, HARDWARE, SOFTWARE AND PARTS FOR THE FMS • SIZE & SCOPE ARE SELECTED ACCORDING TO CORPORATE STRATEGY • HARDWARE & SOFTWARE SELECTED TO FIT SCOPE
  • 30. SYSTEM DESIGN • PART SELECTION IS DONE ACCORDING TO AN ECONOMIC CRITERION & STRATEGIC CONSIDERATIONS • KNAPSACK PROBLEM: LOAD THE FMS TO MAXIMIZE SAVINGS SUBJECT TO FMS CAPACITY
  • 31. KNAPSACK PROBLEM P = PRODUCTIVE TIME PER PERIOD AVAILABLE ON BOTTLENECK FMS RESOURCE pi= TIME PER PERIOD REQUIRED FOR PART i si= SAVINGS PER PERIOD IF PART TYPE i
  • 32. KNAPSACK PROBLEM maximize Σi si Xi subjectto Σi pi < P
  • 33. SOLVING THE KNAPSACK PROBLEM • GREEDY HEURISTIC • Example 5.1, p. 136 • OPTIMIZATION • Example 5.2, p. 138
  • 34. SYSTEM SETUP • ASSIGNMENT OF OPERATIONS AND ACCOMPANYING TOOLING TO MACHINES • PART SELECTION PROBLEM: BATCH FORMATION • LOADING PROBLEM: SEQUENCING AND ROUTING OF PARTS
  • 35. PART SELECTION • GOAL: PLACE REQUIRED PARTS INTO COMPATIBLE BATCHES SUCH THAT • EACH BATCH USES ALL MACHINES • REQUIRE A LIMITED NUMBER OF TOOLS ON EACH MACHINE • HAVE SIMILAR DUE DATES FOR PARTS IN THE BACTH
  • 36. PART SELECTION • GREEDY HEURISTIC: FORM BATCHES BY ARRANGING PART ORDERS BY DUE DATES • PART ORDERS ARE SEQUENTIALLY ADDED TO CURRENT BATCH WITHOUT VIOLATING CONSTRAINTS • BATCH IS THEN READY FOR LOADING • Example 5.3, p. 140
  • 37. Part Selection as a Mixed-Integer Program • Time phased set of part orders Dit for part i in time t • Time available in machine j , Pj • Time required by product i in machine j pij • Number of parts of type i made in time t xit • Number of tool slots in machine j , Kj
  • 38. Part Selection as a Mixed-Integer Program • Number of tool slots required by tool l in machine j , klj • Set of tools l required on machine j to produce part i , l j(i) • Holding cost per period for part i hi • Formulation: p. 142
  • 39. Part Selection as a Mixed-Integer Program • Goal: Minimize inventory cost while meeting due dates • Example 5.4 , p. 142
  • 40. Incremental Part Selection • Several part types in process at any time • System operates almost continuously • Goal: Minimize makespan to complete all available part orders • Procedure: Minimize idle time by balancing work loads subject to part demand and tool magazine capacity • Formulation: p. 144
  • 41. LOADING PROBLEM • BATCH TO BE PROCESSED IS KNOWN • OBJECTIVES REQUIRED • LOADING SOLUTION MUST BE ROBUST AND FLEXIBLE • SOLUTION METHODOLOGIES – MATHEMATICAL PROGRAMMING (p.145) – HEURISTIC APPROACHES (p. 148)
  • 42. LOADING PROBLEM: HEURISTIC APPROACH • PHASE I : ASSIGN OPERATIONS TO MACHINE TYPES • PHASE II: – OPERATIONS COMBINED INTO CLUSTERS TO REDUCE TRANSFERS – MACHINE GROUPS FORMED – OPERATIONS AND TOOLS ASSIGNED TO GROUPS
  • 43. SCHEDULING AND CONTROL • BASIC PROBLEM AREAS – SEQUENCING AND TIMING OF PART RELEASES TO THE SYSTEM – SETTING OF INTERNAL PRIORITIES IN THE SYSTEM – ABILITY OF SYSTEM TO TAKE CORRECTIVE ACTION WHEN COMPONENTS FAIL
  • 44. Flexible Assembly Systems • For the combination of raw materials and components into products with functional characteristics. • Automated vs manned systems • Example: Vibratory bowl feeders and vision systems • Role of Design for Assembly