OMG 402 - Operations ManagementSpring 1997

CLASS 10:

Process and Inventory Control (2)

Harry Groenevelt

March 1997 2

Agenda

• Recap

• Set-up Times and Just-In-Time

• Production Control Techniques– Production Control Decisions– Push Systems and MRP– Pull Systems and Kanban

• Summary

March 1997 3

Recap

• Philosophy of JIT: “produce only when products are needed,

in the quantities needed”• Goals:

– small batch sizes– ‘smoothed’ production for all products (internal production

and supplier’s production)– low inventories– problem visibility and continuous improvement – elimination of waste

March 1997 4

Set-up Times and Just-In-Time (JIT)

• But smaller batches also imply less production per unit time (remember the traffic light?)

• How small can batches become before we no longer meet demand?

• The same question in some production environments: how short can a ‘production cycle’ be?

March 1997 5

Set-up Times and Just-In-Time (JIT)

• Example: A Juice Bottler– five types of juice– eight hour set-ups between batches of each type– steady demand:

1600 (six-)packs/week for each juice– steady production: 9600 packs/week

• Questions:– How small can batches be and still meet demand? – How long between production runs of apple juice?

Orange juice? ...

March 1997 6

Example: 5 juices

• Total Demand (D) = 5*1600 packs/wk = 8000 packs/wk

• Single Machine: Production rate (C) = 9600 packs/wk– Changeover (setup) time between juices

= 8 hrs. (0.2 wks in a 40-hour week)

applesix-pack

cranberrysix-pack

tropicalsix-pack

grapesix-pack

orangesix-pack

7

Cyclic Production Schedule

• A production cycle consists of a setup followed by a production run for each of the products

• The cycle time (T) is the time required to complete a production cycle

Note: in practice, “cycle time” has many definitions; this is just one of them!

Juice 1

setup

setup Juice 2 Juice 3

setup

setup Juice 4 Juice 5

setup

setup Juice 1 Juice 2

setup

setup

0.2 wk0.2 wk 0.2 wk

cycle time T

March 1997 8

Cyclic Production Schedule

S = total setup time per cycle = ________

In terms of the cycle length T:demand for each juice per cycle = ________

total demand per cycle = ________

production time needed to fill demand per cycle = _______

Juice 1

setup

setup Juice 2 Juice 3

setup

setup Juice 4 Juice 5

setup

0.2 wk0.2 wk 0.2 wk

cycle time T0.2 wk0.2 wk

March 1997 9

D = total demand (packs/wk)C = production rate (packs/wk)S = total setup time/cycle (wk)T = cycle time (wk)

JIT: Minimum Cycle Time

We can satisfy demand as long as:cycle time time to fill demand during cycle

+ total set-up time

or

or

SC

DTT

March 1997 10

wk wk.*S

Twkpackswkpacks

CD = 6

1

205

1/9600/8000min

JIT: Minimum Cycle Time

minimum batch size for one juice= (1600 packs/wk)(Tmin) = 9600 packs

average inventory for one juice< 9600/2 packs = 4800 packs (why “<”)

average total inventory< 5*4800 packs = 24,000 packs

March 1997 11

JIT: Product Cycling

inventory for juice (with 6-week cycle)

0

1000

2000

3000

4000

5000

6000

7000

8000

0 1 2 3 4 5 6

Time (weeks)

Inve

nto

ry (

6-p

ack

s)  

Apple Cranberry Tropical Grape Orange

March 1997 12

Total Set-Up Time (S) Demand / Prod. rate (D/C)

minimumT

1

JIT: Intuition for Tmin

minimumT

D = total demand (packs/wk)C = production rate (packs/wk)S = total setup time/cycle (wk)T = cycle time (wk)

March 1997 13

Inv (packs)juice (packs)(packs)(wks)(hrs)Setup Time Cycle Time Lot Size Avg Inv Per Total Avg

8 6 9600 4000 20000

4 3 4800 2000 10000

3 2.25 3600 1500 7500

2 1.5 2400 1000 5000

0.5 0.375 600 250 1250

0.25 0.1875 300 125 625

JIT: Minimum Cycle Times

Impact of Setup Time Reductions

March 1997 14

Insights for Product Cycling

Insight 1: set-up time reduction enables reductions in cycle length and inventory, increases flexibility and reduces lead times.

Insight 2: Cycle length (and average inventory) explode as the ratio of demand to production rate (D/C) nears 1.

March 1997 15

Production Control TechniquesOutline

• Push vs. Pull

• Example of Push: MRP

• Example of Pull: Kanban

March 1997 16

Deciding ‘when’ and ‘how much’:Push vs. Pull

Push: production follows a schedule to meet demand– examples: making lasagna for a dinner party...

Pull: production triggered by demand-driven change in system status– examples: purchasing milk and cereal...

March 1997 17

Bill of Materials(BOM)

Master ProductionSchedule

Lead Times Inventory Status

plannedorders

orderreleases

noticesreports

Example of Push:Material Requirements Planning (MRP)

March 1997 18

Example of Push:Materials Requirements Planning (MRP)

• Generates coordinated material plans– issues work orders to the shop(s)– issues purchase orders to vendors– tracks and projects inventories– generates plans for future orders– flags projected delays or early arrivals

A

C (1)

C (1) B (2)

Bill of Materials(BOM):

Need to begin producing10 product As in week 5

lead timesfor B: 2 weeksfor C: 3 weeks

begin production of ____ B’s in week ____

begin production of ____ C’s in week ____

begin production of ____ C’s in week ____

level 0

level 1

level 2

Material Requirements Planning Example

One ‘A’ built from two B’s and one C. Each B built from one C:

March 1997 20

Material Requirements Planning

• What are MRP’s Strengths?– planning for demand fluctuations– coordinates complex activities and materials

requirements

• Successors to MRP:– MRP II, SAP (and others) coordinate resources,

inventory, financial reporting, and other functions

March 1997 21

Material Requirements Planning

• What are MRP’s Weaknesses?– lead time inflation– codifies inefficiencies– costs (hardware, software, time)– data glutton– lack of flexibility

March 1997 22

station 1 station 2

(see Schonberger article for more details)

Example of Pull: Kanban

• Developed by Toyota• Named after cards (Kanbans) used to control

inventory• Simplest version (HP Video):

“build and move to the hole”– station 1 only moves material

when station 2 empty;

– station 1 only builds when station 1 empty.

March 1997 23

Uses of the Kanban System

• Authorization for production is the demand from downstream

• Tight control over inventory by limiting:– number of Kanban cards

– size of standard container associated with each card (a card could represent 1, 2, 10 or 100 parts)

• Balances flows in production system: everyone works at the bottleneck rate

March 1997 24

Kanban at Work

• What happens when...– a problem halts production upstream?

– a problem halts production downstream?

– downstream demand increases?

March 1997 25

Example of Pull: Kanban

• What are Kanban’s Strengths?– Simple to design and run

(but not necessarily simple to implement!)– Reduces impact of demand fluctuations– Makes queues of work visible– Exposes problems– Provides incentives for continuous

improvement

March 1997 26

Example of Pull: Kanban

• What are Kanban’s limitations?– Does not plan for demand fluctuations– Only works for standard products

(not with “design to order”)– Difficult to coordinate complex tasks– Must be accompanied by worker

empowerment, training, JIT environment

informationmaterial

MRP

Kanban

Summary

• MRP is a push system, Kanban a pull system

• Many systems combine push and pull elements(i.e. push for aggregate plan, pull on shop floor)

• Either ‘push’ or ‘pull’ can be JIT.• Hybrid systems can avoid the weaknesses of each