Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Unlike many other approaches and techniques, material requirements planning works which is its best recommendation.
http://factory-physics.com
History
Begun around 1960 as computerized approach to purchasing and production scheduling. Joseph Orlicky, Oliver Wight, and others. Prior to MRP, production of every part and end item was triggered by the inventory falling below a given level (reorder point). APICS launched MRP Crusade in 1972 to promote MRP.
http://factory-physics.com
Key Insight
Independent Demand finished products Dependent Demand components
http://factory-physics.com
Assumptions
1. Known deterministic demands. 2. Fixed, known production leadtimes. 3. In actual practice, lead times are related to the level of WIP. Flow Time = WIP / Throughput Rate (Littles Law)
Idea is to back out demand for components by using leadtimes and bills of material.
http://factory-physics.com
http://factory-physics.com
MRP Procedure
1. Netting: net requirements against projected inventory --Gross Requirements over time (e.g., weekly buckets) --Scheduled Receipts and current inventory --Net Requirements 2. Lot Sizing: planned order quantities 3. Time Phasing: planned orders backed out by leadtime 4. BOM Explosion: gross requirements for components
6
http://factory-physics.com
Inputs
Master Production Schedule (MPS): due dates and quantities for all top level items Due dates assigned to orders into time buckets (week, day, hour, etc.) Bills of Material (BOM): for all parent items Inventory Status: (on hand plus scheduled receipts) for all items Planned Leadtimes: for all items Components of leadtime Move Setup Process time Queue time (80-90% total time)
Wallace J. Hopp, Mark L. Spearman, 1996, 2000
http://factory-physics.com
Example - Stool
Indented BOM Stool Base (1) Legs (4) Bolts (2) Seat (1) Bolts (2) Graphical BOM
Stool
Level 0
Base (1)
Seat (1)
Level 1
Legs (4)
Bolts (4)
http://factory-physics.com
Note: bolts are treated at lowest level in which they occur for MRP calcs. Actually, they might be left off BOM 8 altogether in practice.
Example
Item: Stool (Leadtime = 1 week) Week Gross Reqs Sched Receipts Proj Inventory Net Reqs Planned Orders 0 1 2 3 4 5 120 -100 100 6
20
20
20
20
20 100
-100
Item: Base (Leadtime = 1 week) Week Gross Reqs Sched Receipts Proj Inventory Net Reqs Planned Orders
Wallace J. Hopp, Mark L. Spearman, 1996, 2000
0 100
-100
-100
http://factory-physics.com
Example (cont.)
BOM explosion
Item: Legs (Leadtime = 2 weeks) Week Gross Reqs Sched Receipts Proj Inventory Net Reqs Planned Orders 0 1 2 200 0 3 400 -200 200 4 5 6
0 200
-200
-200
-200
http://factory-physics.com
10
Terminology
Level Code: lowest level on any BOM on which part is found Planning Horizon: should be longer than longest cumulative leadtime for any product Time Bucket: units planning horizon is divided into Lot-for-Lot: batch sizes equal demands (other lot sizing techniques, e.g., EOQ or Wagner-Whitin can be used) Pegging: identify gross requirements with next level in BOM (single pegging) or customer order (full pegging) that generated it. Single usually used because full is difficult due to lot-sizing, yield loss, safety stocks, etc.
11
http://factory-physics.com
Pegging and Bottom up Planning Table 3.7 MRP Calculations for Part 300 Part 300 Required from B Required from 100 Gross Requirements Scheduled Receipts Adjusted Scheduled Receipts Projected on-hand Net Requirements 50 125 100 100 50 25 25 -65 65 65 15 15 1 2 35 90 3 4 30 60 90 15 5 6 15 7 8
Planed order receipts 65 15 Planed order releases On-hand = 40 Scheduled releases = none Lot-for-lot LT = 1 week
Wallace J. Hopp, Mark L. Spearman, 1996, 2000
http://factory-physics.com
12
Bottom Up Planning
Reference Figure 3.7 Assume that the scheduled receipt for week 2 for 100 units is not coming in! Have 50 units of inventory, with requirements of 125. Implications I can fill 50 units of demand from part 100 or 50 units from part B. If I go with part 100, that will allow me to fill some of the demand of part A (100 needed for part A). Can I ship 50 units to the customer now and 50 units later? What if I cover the demand of part B? Correct choice depends on the customers involved.
http://factory-physics.com 13
More Terminology
Firm Planned Orders (FPOs): planned order that the MRP system does not automatically change when conditions change --- can stabilize system Service Parts: parts used in service and maintenance --- must be included in gross requirements Order Launching: process of releasing orders to shop or vendors --- may include inflation factor to compensate for shrinkage Exception Codes: codes to identify possible data inaccuracy (e.g., dates beyond planning horizon, exceptionally large or small order quantities, invalid part numbers, etc.) or system diagnostics (e.g., orders open past due, component delays, etc.)
14
http://factory-physics.com
Before, EOQ represented a basic trade-off between inventory costs and setup costs. Goldratt: If I am not at capacity on an operation requiring a setup, the time spent on a setup is a mirage. Make setups occur as frequently as possible (smaller lot sizes) as long as capacity is available Produce only when inventory level reaches zero (Wagner Whitin) is not optimal when capacity is a constraint Authors know of no commercial MRP system that uses WW.
Wallace J. Hopp, Mark L. Spearman, 1996, 2000
http://factory-physics.com
15
http://factory-physics.com
16
http://factory-physics.com
17
Problem Formulation
t= A time period, t = 1 to T, where T is the planning horizon. Dt = Demand in time period t (in units). Ct = Unit production costs ($/unit), excluding setup or inventory cost in period t. At = Setup (order) cost to produce (purchase) a lot in period t ($). Ht = Holding costs to carry a unit of inventory from period t to period t+1 ($/unit). e.g., if holding costs consists entirely of interest on money tied up in inventory, where i is the annual interest rate and periods correspond to weeks, then
h=
I * Ct 52
It =Inventory (units) left over at the end of period t. Qt =The lot size (units) in period t; this is the decision variable
http://factory-physics.com
18
Problem Objective
Satisfy all demands at minimum cost (production, setup, and holding costs) All the demands must be filled, only the timing of production is open to choice. If the unit production cost does not vary with t, then production cost will be the same regardless of timing and can be dropped from consideration. Look at an example: assume setup costs, production costs, and holding costs are all constant over time. Thus, need only to consider setup costs and holding costs.
Wallace J. Hopp, Mark L. Spearman, 1996, 2000
http://factory-physics.com
19
http://factory-physics.com
20
10
t Dt
6 7
9 10
20 50 10 50 50 10 20 40 20 30
ct 10 10 10 10 10 10 10 10 10 10 At 100 100 100 100 100 100 100 100 100 100 ht 1 1 1 1 1 1 1 1 1 1
http://factory-physics.com
21
Lot-for-lot
Simply produce in period t the net requirements for period t. Minimizes inventory carrying costs and maxim izes total setup costs. It is simple and it is consistent with just in time. Tends to generate a more smooth production schedule. In situations where setup costs are minim al (assembly lines), it is probably the best policy to use.
http://factory-physics.com 22
11
A = 100 h =1 300 D= = 30 10
Lot-for-Lot: $1000
http://factory-physics.com
23
Q=
2 AD = h
2 x100 x 30 = 77 1
7 20
8 9 40 20 77 100 41 21 58
38
http://factory-physics.com
24
12
http://factory-physics.com
25
15 45 60
25 15 20 15 60 15
Let P = 3. Skip first period, no demand. Sixty units covers demand for periods 2, 3, and 4. Period 5 is skipped because there is no demand. Sixty units covers demand in periods 6, 7, 8. Fifteen units covers demand in period 9, which is the last period in the planning Horizon.
Wallace J. Hopp, Mark L. Spearman, 1996, 2000
http://factory-physics.com
26
13
Part-Period Balancing
Combines the assumptions of WW with the mechanics of the EOQ. One of the assumptions of the EOQ is that it sets the average setup costs equal to the average inventory carrying costs. Part-period. The number of parts in a lot times the number of periods they are carried in inventory. Part-period balancing tries to make the setup costs as close to the carrying costs as possible.
http://factory-physics.com
27
1 2
5 6
8 9
15 45
25 15 20 15
Part-periods 0 15 x 1 = 15 15 + 20 x 2 = 55 55 + 15 x 3 = 100
Fixed order quantity method (without modifications) tends to work better than WW when dealing with multi-level production systems with capacity limitations.
Wallace J. Hopp, Mark L. Spearman, 1996, 2000
http://factory-physics.com
28
14
Nervousness
Item A (Leadtime = 2 weeks, Order Interval = 5 weeks) Week 0 1 2 3 4 5 6 7 Gross Reqs 2 24 3 5 1 3 4 Sched Receipts Proj Inventory 28 26 2 -1 -6 -7 -10 -14 Net Reqs 1 5 1 3 4 Planned Orders 14 50 8 50 -64 50
Component B (Leadtime = 4 weeks, Order Interval = 5 weeks) Week 0 1 2 3 4 5 6 7 Gross Reqs 14 50 Sched Receipts 14 Proj Inventory 2 2 2 2 2 2 -48 Net Reqs 48 Planned Orders 48
http://factory-physics.com
29
http://factory-physics.com
30
15
Reducing Nervousness
Reduce Causes of Plan Changes: Stabilize MPS (e.g., frozen zones and time fences) Reduce unplanned demands by incorporating spare parts forecasts into gross requirements Use discipline in following MRP plan for releases Control changes in safety stocks or leadtimes Alter Lot-Sizing Procedures: Fixed order quantities at top level Lot for lot at intermediate levels Fixed order intervals at bottom level Use Firm Planned Orders: Planned orders that do not automatically change when conditions change Managerial action required to change a FPO
31
http://factory-physics.com
Handling Change
New order in MPS Order completed late Scrap loss Engineering changes in BOM
Regenerative MRP: completely re-do MRP calculations starting with MPS and exploding through BOMs. Net Change MRP: store material requirements plan and alter only those parts affected by change (continuously on-line or batched daily). Comparison: Regenerative fixes errors. Net change responds faster to changes (but must be regenerated occasionally for accuracy.
http://factory-physics.com 32
16
Rescheduling
Top Down Planning: use MRP system with changes (e.g., altered MPS or scheduled receipts) to recompute plan can lead to infeasibilities (exception codes) Orlicky suggested using minimum leadtimes bottom line is that MPS may be infeasible Bottom Up Replanning: use pegging and firm planned orders to guide rescheduling process pegging allows tracing of release to sources in MPS FPOs allow fixing of releases necessary for firm customer orders compressed leadtimes (expediting) are often used to justify using FPOs to override system leadtimes
http://factory-physics.com
33
Safety Leadtimes:
inflate production leadtimes in part record used as hedge against time uncertainty (e.g., delivery delays)
http://factory-physics.com
34
17
100 120
http://factory-physics.com
35
4 0 10
5 30 -20 20
40
1 20 20 50
40
2 40 50 30
3 20 10 10
4 0 10
5 30 -20 30
36
http://factory-physics.com
18
Safety Leadtime = 1 week Week Gross Reqs Sched Receipts Proj Inventory Net Reqs Planned Orders 0 1 20 20 2 40 50 30 50 3 20 10 4 0 10 5 30 -20 20
40
http://factory-physics.com
37
http://factory-physics.com
38
19
http://factory-physics.com
http://factory-physics.com
20
http://factory-physics.com
41
http://factory-physics.com
42
21
http://factory-physics.com
43
Conclusions
Insight: distinction between independent and dependent demands Advantages:
General approach Supports planning hierarchy (MRP II)
Problems:
Assumptions --- especially infinite capacity Cultural factors --- e.g., data accuracy, training, etc. Focus --- authority delegated to computer
http://factory-physics.com
44
22