MRP vs. Advanced Supply Chain Planning

1
MRP vs. Advanced Supply Chain Planning

• NCOAUG
• March 3, 2003
• Presenter
• Dan Vanden Brink

2
Outline

• Review MRP/MRP II
• Review JIT/Toyota Production System
• Advanced Production System
• Constraint Based
• Linear Programming
• Simulation Based

3
Advanced Planning and Scheduling

• Current Planning Techniques
• MRP/MRP II
• Toyota Production System
• Problems with Current Systems
• Constraint Based Planning Systems
• Linear Programming Based System
• Simulation based planning
• The Future...

4
MRP

• Developed as an offshoot of Financial and
  Inventory Transaction Software
• Bill of Material Driven
• Lead Time Based Calculations
• Push System
• Time Buckets

5
MRP
6
MRP Problems

• Capacity Infeasibility
• Long Lead Times
• Lead Times Fixed
• Planners cover themselves with long lead times
• Schedule Instability

7
System Instability - Example

• Reduction in demand actually switches a feasible
  MRP run into an infeasible run
• Reduce demand in period 2 from 24 to 23
• Fixed Order Period Lot Sizing of 5 periods
• One unit of A requires one unit of B
• A has a lead time of 2 periods
• B has a lead time of 4 periods

8
System Instability - Example
9
System Instability - Example
10
MRP II

• Reduces Problems seen in MRP
• Capacity
• Rough Cut Capacity Planning
• Capacity Resource Planning
• Time Fences

11
Capacity Requirements Planning
12
Pros of MRP (II)

• Easy to use/understand
• Many packages available
• Workforce well training in the Mechanics of MRP
• Excellent at aggregating Material Requirements

13
Cons of MRP (II)

• Reduces, but does not eliminate problems seen in
  MRP
• Still assumes fixed lead times
• Capacity Planning does not offer solution or
  feasible production plan
• Not interactive - no what-if planning is
  possible
• Long processing time

14
Cons w/ MRP (II)

• Lead time assumption
• Lead times are always variable and dependant on
  other jobs on factory floor
• Planners increase lead times to insure that
  product is available
• Increased lead times increases inventory and WIP
• Increased / Inventory / WIP clogs production and
  increases actual lead times

15
Toyota Production System

• Developed by Taiichi Ohno in 1950s for Toyota
• System wide approach, not just scheduling
• Just in time material/production
• Autonomation
• Automated
• Foolproof

16
TPS Goals

• Reduce variability
• Zero Defects
• Zero Surging
• Zero Set Ups
• Zero Breakdowns
• Zero Handling
• Zero Lead Time
• Zero Lot Size
• Results in Inventory

17
TPS Goals

• Capacity Buffers also required for Zero
  Inventories

18
TPS vs. MRP
19
Pros of JIT/TPS

• Complete System Improvements
• Scheduling only one aspect of TPS
• Improvements in Quality
• Improvements in Machine Uptime
• Improvements in Morale
• No large IT system required
• Data integrity for inventory and BOMs not as
  important

20
Problems with TPS

• Does not deal well with volatility in
• Demand - MPS must be level loaded within 10
• Diverse Product Line
• Short Product Life Cycle
• Requires Capacity Slack, Not meant for industries
  with high fixed capacity costs.
• A COMPLETE MANUFACTURING SYSTEM NOT JUST A
  SCHEDULING TOOL

21
Advance Production Scheduling Systems

• Eliminates many of the shortcomings of MRP II
  without the limitations of TPS
• Interactive
• Variable lead times
• Capacity constrained schedules
• Feasible Schedule
• Made possible with advances in computer power
• Memory Resident

22
Types of APS

• Constraint based scheduling
• Based on E. Goldratts book The Goal.
• Linear Programming
• Simulation

23
Advantages to all APS Systems

• Fast (Generally)
• Less than 10 of the time required to run MRP
• What if analysis
• Better solution than MRP
• Better user interfaces (generally)
• Better access to data
• Minimize trade off between capacity, inventory
  and customer service

24
Advantages to all APS Systems
25
Constraint Based Scheduling

• Offers feasible schedule by using Gantt chart
  like production plans
• Recommends alternatives to eliminate production
  constraints
• Schedules inventory to come in only when needed
• Uses proprietary heuristics to develop solutions

26
Solver Technologies Heuristics

• Using Heuristics
• Rules are defined that limit or constrain the
  domain of possible solutions
• Examples of limits modeled through heuristics
• Preferred Routings
• Batch Sizes
• Minimum Run Lengths

• Load Line 1,
• Then Line 3

27
Iterations

• When an algorithm runs through a series of
  solutions and chooses one, the cycle of steps is
  one ITERATION
• Generally, you must run many iterations to solve
  a whole planning or scheduling problem

• Search

• Find

• Implement

28
How Does It Solve?

• Algorithms

• Algorithms are steps the software goes through to
  find and solve problems.
• Example
• 1. Find Problems Forecasts, Orders, Safety
  Stocks...
• 2. Prioritize Problems Fulfill Orders, then
  Forecast, then Safety Stock
• 3. Pick a Problem to solve 1 Order due tomorrow
• 4. Determine possible solutions is inventory
  available, or should production be scheduled?
  Line 1 or Line 2?
• 5. Prioritize solutions 1st - Starting
  inventory 2nd - Line 2
• 3rd - Line 1
• 6. Choose a solution either low cost or least
  cost

29
Constraint Based Scheduling

• Pros
• Can model large complex manufacturing systems
• Fastest of APS solutions
• Detail schedules
• Cons
• Schedule solution can be sub-optimal
• Difficult to debug schedule
• Complex

30
Constraint Based Scheduling

• Industries
• Discrete Manufacturing
• High Tech
• Engineer to Order
• Manufacture to Order
• Assemble to Order

31
Linear Programming Based Scheduling

• Uses Linear Programming to optimize production
  schedule
• Generally used for Master Scheduling

32
Mathematical Optimization

• Using Mathematical optimization
• Problems are modeled as mathematical equations
  (gt, lt, )
• Examples of problems modeled through optimization
• Profitability
• Capacity limitations
• Crewing Requirements

• X Y lt 40

33
Linear Programming Based Scheduling

• Pros
• Offers optimal solution
• Easy to implement if you stay within predefined
  templates
• Cons
• Difficult in complex environments
• Does not produce detailed schedules
• Difficult to debug schedule

34
Linear Programming Based Scheduling

• Industries
• Logistics
• Process
• CPG
• Build to Stock

35
Simulation Based Scheduling

• Uses statistical and mathematical models to model
  factory
• Queuing Theory
• Actually steps through each operation in
  manufacturing to model plant
• Used for production, but commonly used for
  process improvement and capital decisions.

36
SimulateCalculate

• As an algorithm proceeds through a series of
  solutions, it must evaluate each solution it
  tries
• When the software determines the quality of a
  potential solution, it is performing a
  simulation or calculation of the appropriate
  costs or penalties
• In general, lower cost better solution

37
Simulation Based Scheduling

• Pros
• Accurately models the stochastic nature of a
  manufacturing plant
• Extremely accurate representation of production
• Detailed schedules
• Cons
• Schedule solution can be sub-optimal
• Difficult to implement
• Difficult to debug schedule
• Complex

38
What APS doesnt do...

• APS systems can not execute your business
• APS systems will recommend a course of action -
  it is up to the user to commit to a plan or
  schedule and then to publish it
• Once the plan or schedule is published, the rest
  of the world can act on it by ordering raw
  materials, and manufacturing the products.

39
APS and the Future

• As APS tools become more advanced, the number of
  decisions requiring human intervention will
  continue to decrease.
• APS will improve ability to display and
  communicate information to planners and factory
  floor.
• Think of this Manufacturing could evolve into a
  completely automated process where an order gets
  entered by the customer and the factory will know
  how, where and when to make the order.

40
Should you go with APS

• Are you focused on Capacity or Material?
• Many operations
• Constrained Operations
• Do you have accurate BOM, inventory, routing, and
  costing information?
• Are you in a highly variable environment?
• New Products
• Volatile Demand
• Supply Volatility

41
Should you go with APS

• In most cases, APS will improve your planning.
• Training is essential!

42
Questions?