1

Operational models for customer –

specific variety

Prof. Bart MacCarthy

Operations Management Division

Nottingham University Business School

Bart.maccarthy@nottingham.ac.uk

2

Outline

1. Product variety in order fulfilment

2. A classification of fulfilment systems

3. Open pipeline approaches

–flexible and dynamic fulfilment strategies

4. When should we open the pipeline?

2

MaterialOrders

ShippingOrders

ManufacturingOrders - MRP

Forecast ofRequirements

ReplenishmentOrders

DistributionOrders

SalesOrders

RawMaterials

Suppliers

Transport

FactoryWIP

FinishedGoods

Transport

RegionalStores

Transport

LocalDepot

Customer

Information

ChainPhysical

Supply

Chain

Supermarket customers Supermarket customers

have a 50/50 chance of have a 50/50 chance of

finding all items on a 20 finding all items on a 20

item list. Inst. of Grocery item list. Inst. of Grocery

Distribution (IGD, UK) Distribution (IGD, UK)

�Inventory costs (many)

�Transport/ distribution costs

�SC Management costs

�Markdowns costs

�Lost sales

�Penalty costs

Ramdas K (2002)

Fisher & Ittner (1999)

Bils & Kennow (2001)

Er & MacCarthy (2006)

Alfaro & Corbett (2003)

Denton (2003)

Product

variety-

supply chain

perspective

4

Textbook models - MTS/ATO/MTO/ETO

MTS

C u s t o m e r

AssemblyFabrication PackDesign

Stock

Stock

C u s t o m e r

AssemblyFabrication PackDesign

ATO

order

order

Stock?

MTO

ETO

order

order

C u s t o m e r

C u s t o m e r

Stock

AssemblyFabrication PackDesign

AssemblyFabrication PackDesign

3

5

Stock

Form postponement and late point

differentiation

Decouple the system

Process

3

Process

2

Process

4

Process

1

Forecast-driven stable

production Customer order driven

responsive production

Upstream push Downstream pull

Move the decoupling point as

far downstream as possible?

Customer

Customer

limited variety high variety

van Hoek (2001)

Yang, Burns& Backhouse ( 2004)

6

Fixed

variety

MTS

Mass

Customization

ATO

Form postpon

ement

Late point

different.

Rapid &

reactive

NPI

Locate to

order LTO

Available to

promise (ATP)

Multi-mode

fulfilment

Open

pipeline

planning

BTO/MTO

fixed

variety

Superficial customisers

Personalisation/ aesthetics

Functional/ dimensional

customisers

ETO within

design

envelope Customize/ differentiate for major customers

1

2

3

4

5

6

Open

Systems

Mass Customisers

High clock

speed

Customisers

Stockists

Builders

Producer-led variety

Customer-led variety

Fulfilment by

Locating

Matching

Reconfiguring

4

7

Operational models for

customer-specific variety

� Classifications are approximate

� May be combined

� different products

� different markets

� product lifecycles

� Open pipeline systems use different

fulfilment mechanisms flexiblyAmaro, Hendry, Kingsman (1999)

MacCarthy, Brabazon, Bramham (2003)

MacCarthy, Brabazon (2008) – Order fulfilment models in high variety

production environments -Nottingham E-Print available

8

Multi-mode fulfilment strategies

Components

Part assembled products (Vanilla boxes)

Products

Stage 1 Stage 2

Customers

Part finished, common item

Products

Stage 1 Stage 2

Customer Raw materials

Slab store

Stage 1 Stage 2

Customers

Graman & Magazine (2002) Denton et al (2003)

Swaminathan & Tayur (1998, 1999)

Stage 1 Stage 2 Stage nC u s tom e rs

Meredith & Akinc (2007)

5

9

Open pipeline approach

Initiate to Forecast

C us tom er

Process stages

Unsold finished products

Parking of unsold, semi-finished products, awaiting a customer

Options for unsold products

Switch ‘to-order’

It could also incorporate some fixed decoupling points

Fulfil from anywhere in the system

• A generalisation of multi-mode fulfilment systems

•Combine Build-to-Forecast with some or all of Stock, ATO,

planning pipeline allocation & BTO

• Introduce more flexibility – reconfigure product in the pipeline

10

Multi-featured product

Interior Trim

Body shape: 2dr, 4dr

Exterior colour

Engine type: Diesel, petrol

Wheel type: e.g. alloys

Automatic/Manual

Engine size

A Feature can have several Options

Variant: a unique combination of Options

Why is an open pipeline attractive in automotive?

6

11

Dealer Stock

Automotive build strategies

Producer forecast

Sales Order Bank Prod.

Order Bank In

Transit

Dealer Orders

Customer

Sell

from

Stock

High

stocks

dispersed Short lead-

times ex-

stock +

discounts

February 2003

Build

to

Order

Scheduling,

complex

searching +

allocation

Scheduling

The VBTO system opens the pipeline

12

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.0001 0.001 0.01 0.1 1 10 100 1000

Variety / Pipeline Ratio

Proportion of customers (Cumulative) Stock (pipe 64)

Stock (pipe 128)

Stock (pipe 256)

Stock (pipe 512)

Stock (pipe 1024)

Stock (pipe 2048)

BTO (all)

Stock

BTO

Behaviour of a conventional

system

Customer

Stock

Pipeline

New order queue

Forecast feed

Conventional closed pipeline

7

13

Behaviour of an open system

(without reconfiguration)

Customer

Stock

Pipeline

New order queue

Open pipeline VBTO

Forecast feed

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.0001 0.001 0.01 0.1 1 10 100 1000

Variety / Pipeline Ratio

Proportion of customers (Cumulative)

BTO

Pipe

Stock

Pipeline

Stock

BTO

Brabazon and MacCarthy (2006), IJPE

14

Comparison of stock levels

Customer

Stock

Pipeline

New order queue

Forecast feed

Conventional - closed

Customer

Stock

Pipeline

New order queue

VBTO - open

Forecast feed

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.0001 0.001 0.01 0.1 1 10 100 1000

Variety / Pipeline Ratio

Stock level as a proportion of pipeline length

Conventional

VBTO

8

15

The effect of V/P and S/V on

BTO fulfilment

16

The effect of V/P and S/V on

stock fulfilment

9

17

The effect of V/P and S/V on

pipeline fulfilment

18

Pipeline reconfiguration flexibility

Effect of different levels of reconfiguration flexibility on lead time

0

10

20

30

40

50

60

0.001 0.01 0.1 1 10 100 1000

Variety / Pipeline

Customer Leadtime

No Flex

Flex A

Flex A,B

Flex A,B,C

Flex A,B,C,D

Flex A,B,C,D,E

Increasing

flexibility

None

Full

10

19

Designing and managing

open pipeline systems

� Many issues in managing and controlling

and open pipeline system

� Intelligent release

� Revenue management

� Initial conditions

� Reconfiguration in the pipeline

� How does each stakeholder group fare ?

producer, customers, dealers?

� Operational constraints

20

Opening the pipeline

� Exploits the variety in the pipeline to

meet customer requirements more

precisely

ManufacturePlanning Distribution

•Scale

•Complexity

•Time horizons

•Costs & efficiency

????

11

21

Opening the pipeline – when

should you do it?

� Valuable when � significant levels of variety, significant planning

pipeline

� heterogeneous customer base + a significant

proportion require matched specifications

� products can be allocated from the planning pipeline

� flexible reconfiguration possible at reasonable costs

� pure MTS or M/BTO unattractive - costs or

response times

� Machinery, capital equipment, instrumentation,

computer servers, automotive……

22

The dynamic open pipeline

Constraints tightening

Focus on allocation,

scheduling and fulfilment

Focus on forecasting,

planning and trade–offs

Decision cut-off points

�Extensive

product

ranges

� Complex

product

structures

�Diverse

customer

base

� Specification

�Waiting

� Price

UPSTREAM DOWNSTREAM

�Business models are changing in many sectors,

�Many new operating models involve greater dynamism and flexibility in the use of future available productive capacity‘

12

23

In conclusion

� Form postponement requires right combination of

product architecture, manufacture and assembly +

careful design

� Open systems exploit the variety in the planning

pipeline to meet customer requirements more precisely

� Open systems may offer significant improvements in

fulfilment levels in the presence of high variety,

particularly when combined with flexible reconfiguration

� Waiting and stock characteristics are more complex

than conventional systems

24

� Thanks to the research team and our

industrial collaborator, Ford

� Questions?