H.--O.

Gnther
H.
Dept. of Production Management
Dept
Technical University of Berlin

Supply Chain Management and

Advanced Planning Systems
A Tutorial

SCM Workshop, TU Berlin, October 17-18, 2005

Outline
Introduction: The concept of SCM and APS
APS modules
Strategic Network Design
Demand
Pl
Planning
i
Supply Network Planning

External
Procurement

Production
Planning /
Detailed
Scheduling

Transportation
Planning /
Vehicle
Scheduling

Order
Fulfilment
ATP / CTP

Lesson 1

There is an ongoing
g g
change of paradigm
from ERP / MRP
towards APS.

Generations of PPC software

1960s

Predecessors of PPC systems

Focus on inventory control
Basic order processing

1970s

Material requirements
q
planning
p
g (MRP)
(
)
Bill of material files
Calculation of net requirements

1980s

Manufacturing resources planning (MRP II)

Enhanced planning functions
Integration
of financial accounting
functions
g
g and management
g

1990s

Integrated systems
CIM: Integration of manufacturing
ERP systems covering the whole enterprise

2000s

Advanced Planning and Scheduling systems

Integration into Supply chain management concept
Use of true optimization techniques

Lesson 2

SCM needs
d
sophisticated
planning
l
i systems.
t

Change of view

Plant specific view

Holistic view

More efficient network

network--wide planning needed
Development of APS

Supply chain management

Definition
(customer order or forecast)

Customers,
distribution centres

Transportation order

Transporters

Production order

Manufacturing
stages

Replenishment order

Tiers of
suppliers

Demand element

A supply chain
consists of all parties
involved, directly or
indirectly, in fulfilling
a customer request.
the supply chain
includes all functions
involved in receiving
and filling a customer
request.
(Chopra and Meindl, 2004)

Types of supply chains (networks)

Inter-company
Inters ppl chains
supply
Internal supply
chains
h i

Lesson 3

APS employ hierarchical

planning, consider the
availability of resources,
and apply optimization.

Characteristics of Advanced Planning

Analysis of the
decision problem
Procurement

Production

Distribution

Sales

Integral planning of the entire supply chain

True optimization

Hierarchical planning
e.g. APS from SAP, i2 technologies,
Manugistics, Oracle, Aspen Tech

Definition of the
objectives

Forecasting future
developments

Feasible
alternatives

Selection of
solution
(Fleischmann et al. 2002)

Architecture of Advanced Planning Systems

Procurement

long-term

mid-term

short-term

Production

Distribution

Sales

Strategic Network Design

Supply Network Planning

External
Procurement

Production
Planning /
Detailed
Scheduling

Transportation
Planning /
Vehicle
Scheduling

Demand
Planning

Order
Fulfilment
ATP / CTP

Companies rarely use the entire suite of modules.

Even APS from different software vendors are used in a company.
Various industry specific solutions offered

Classical MRP systems

P

M1

M2

Customer
orders

Material
requirements
q

M3

Capacity
requirements

Routing
A2
A1

A4
A3

Scheduling

Lesson 4

Forgett BOMs
F
BOM
and routings.
L
Learn
PPM
PPM.

Production-process-model
Operation:
Component C

Operation:
Sub-assembly
Sub
assembly S

Operation:
Product P
mat. M1
res. c

M1

act. P1
act. C1

M2

act. S1

mat. M3
res a
res.

act. P3

mat. M2
comp C
comp.
res. b

act. P2
res. d
res
sub-a. S

res. e
res f
res.

M3
C

M3

M2

M1

Assign resources and materials to each activity.

activity
Alternative modes (resources, routings) can be defined for an activity.

APS planning cycle

Strategic network design

Demand planning

Supply network planning

Transportation planning /
vehicle scheduling

Order fulfilment
and ATP / CTP

Production planning /
detailed scheduling

External
procurement

The APS planning cycle represents the logical order of planning tasks.
tasks
Planning tasks differ by the frequency by which they are called up.

Outline
Introduction: The concept of SCM and APS
APS modules
St t i network
t
k design
d i
Strategic
Supply network planning
Demand planning
External procurement
Production Planning / Detailed Scheduling
T
Transportation
t ti Planning
Pl
i / Vehicle
V hi l S
Scheduling
h d li
Order Fulfilment and ATP / CTP

Lesson 5

Strategic
St
t i network
t
k design
d i iis
a powerful, but the least
utilized
tili d module
d l off APS
APS.

Strategic network design

Strategic Network Design

Decisions
Number of plants and DCs
Locations and capacities
Assignment of products to plants
Assignment of locations to each other
e.g. customers to DCs
Determination of transportation links

Mathematical methods
He ristics MILP
Heuristics,
MILP, Cl
Clustering
stering techniq
techniques
es

Strategic network design

low

high

long

short

high

low

high

low

Planning frequency

Planning horizon

D
Degree
off aggregation
ti

Management level

Strategic network design

Exercise: Modelling Nutricias supply network design
by use of SAP APO 3
3.1
1 (based on Wouda et al
al., 2003)

Supplier
Supplier
Plant
Customer
Distribution
centre

Lesson 6

Forecasting is essential at
all planning levels. Pure
forecasts can cause ampliampli
fication of demand.

Demand planning
Use of forecast
Strategic level: design of the supply network
Operational level: production, distribution,
and procurement decisions

Demand
Planning

Short-term: update of production orders

Demand planning
Most essential in make-to-stock environment,
e.g. in the consumer goods industry
Collaborative forecasting between partners in the supply chain

Mathematical methods
Statistical forecasting techniques
Tools for incorporating human judgement

Lesson 7

Supply network planning

is a core module of SCM.
Huge cost savings
can be gained.

Supply network planning

Mathematical methods
LP and MILP,
MILP heuristics
Supply
pp y Network Planning
g

Decisions
Allocation of production quantities between plants
Supply from the plants to DCs and from the DCs to customers
Smoothing out seasonal cycles in demand
Consideration of production, transportation, and handling capacities
as hard constraints
Consideration of demand,
demand due
d e dates,
dates and safety
safet stocks
as soft constraints

Supply network planning

year

day

year

day

high

low

high

low

Planning frequency

Planning horizon

D
Degree
off aggregation
ti

Management level

Supply network planning

Supply chain engineer in SAP APO 3.1

S t
Set-up
off the
th network
t
k configuration
fi
ti
Integration of suppliers and transporters
Assign and modify master data, e.g. product portfolio and capacities,
costs, safety stocks, demand figures, modes of transportation etc

Lesson 8

Generic model formulations

show pros and cons.

Supply network planning

Generic model formulation
Storage and handling capacity at DCs
p y pjt SC j

Production capacity
p
y at p
plants

a pi x pijt PCit
pP (i ) j J (i )

p P (i )

p x pijt +

p P (i ) iI ( j )

p z pjkt HC j

p P (i ) k K ( j )

Transportation capacity per link

p x pijt
ijt TCijt
p P (i )

Aggregate demand per DC

z pjkt = b pkt

j J (k )

Storage and handling costs per DC

h p y pjt + c pjk z pjkt

pP jJ tT

pP jJ kK ( j ) tT

Production costs per plants

c pij x pijt

p P (i ) i I j J (i ) tT

Assigning attributes to pre-defined entities

Supply network planning

Planning books offer user-defined views of the solution

Additional features of SAP APO 3.1

Safety stock planning
Lot-sizing and production campaign planning

Lesson 9

PP/DS is the module

most difficult to adapt to the
application specific
application-specific
features.

Production Planning / Detailed Scheduling

Decisions

Mathematical methods

Generation of production orders

Allocation of resources according
to a finite scheduling policy

Genetic algorithms
Constraint programming
Rules and heuristics
Application specific
algorithms

Production
Planning /
Detailed
S h d li
Scheduling

Lot-sizing, sequencing, and procurement proposals

Consideration of the availability of resources as hard constraints
Consideration of due dates, time windows etc. as soft constraints

Production Planning / Detailed Scheduling

year

day

year

day

high

low

high

low

Planning frequency

Planning horizon

D
Degree
off aggregation
ti

Management level

Production Planning / Detailed Scheduling

Case study: Production of hair dyes

Final Product
Box

Dye Cream

Dispersion

Shampoo

Cream Liquid

Dispersion
Li id
Liquid

Shampoo
Li id
Liquid

Dispersion
Chemicals

Shampoo
Chemicals

Bottle

Tube

Cream

Processing
line
Chemicals
cream
Tube

Processing line
dispersion

Processing line
shampoo

Filling shampoo

Filling cream and dispersion

and packaging final product

Lesson 10

Most scheduling problems

are NP-hard.
However, practical solutions
are often easy to obtain.

Production Planning / Detailed Scheduling

Case study: Production of hair dyes
Solutions
S
l ti
obtained
bt i d ffrom th
the PP/DS module
d l off SAP APO 3
3.1
1
within 2 minutes
GA superior
i tto CP
Alternative solvers can be
i t
integrated
t d via
i O
Optimization
ti i ti
extended workbench

Additional features
of SAP APO 3.1
Characteristics based planning
Shelf-life consideration
Model mix planning

Production Planning / Detailed Scheduling

Complete pegging (dynamic or fixed)
10

Customer orders

20

10

10 20

10 40

40

-10

60

50
Deficit

30

T
Transportation
t ti orders
d

100
Surplus

30

Production orders

20

50

50

50

30

50

+20
Surplus

50

Purchase orders

80

30

20

80

50

+10

Lesson 11

Procurement requires a
high degree of collaboration
between partners in the
supply chain.

External procurement
Decisions
Determination of the purchase quantity
Selection of the supplier

Mathematical methods
Rule-based and heuristic
procedures
d
External
Procurement

In-house or external supply?

Release of deliveries for JIT goods

MILP

Lesson 12

ATP / CTP provide simple,

but veryy effective tools.

Order fulfilment and ATP / CTP

Decisions
Matching customer orders
against available quantities
Quick response to customer requests
(Available to Promise: ATP)
Capacity check for new or enhanced
production orders in response to
customer requests
(C
bl tto P
i
(Capable
Promise:
CTP)

Mathematical methods
Rule-based ATP
Batch mode ATP (MILP)

Order
Fulfilment
ATP / CTP

Order fulfilment and ATP / CTP

Rule-based ATP: Bayer AG, Germany
Query from customer EXC1, for
product P1, amount Q1, date T1

Preliminary quick check with

availability chart for P1, T1

Query for commitment

from inventory for
EXC1, P1, Q1, T1

Yes

Commit order for

EXC1, P1, Q1, T1

No
Earliest date available for P1, 1 T2

EXC1 satisfied?

Commit order for EXC1, P1, Q1, T2

using query for commitment from
inventory

Yes

No

Query for commitment

from inventory or
production for EXC1,
P1,, Q1,
Q , T1
No

Yes

Commit order for EXC1, P1, Q1, T1;

create campaign(s) for production

Place order for EXC1, P1, Q1, T1;

commitment after optimization run if
capable to produce

Order fulfilment and ATP / CTP

ATP decision cube of SAP APO 3.1

Original
order
d

Customer order /
end item

Same product/
alternative location

Alternative product/
Alternative

alternative location

product

Production

Change of location

Alternative component

Lesson 13

Customer satisfaction
finally depends on the
timelyy execution of
transportation activities.

Transportation planning and vehicle scheduling

Decisions
Choice of the transportation mode (e
(e.g.
g truck or rail)
Integration of external logistics service providers
and offering of own logistics services (e.g. VMI)
Determination of regular freight
frequencies between locations and
of the unit transportation size
Vehicle loading and scheduling
considering
g time windows for delivery
y
Collaboration between locations and
with external logistics service providers

Mathematical methods
Heuristics
Local Search

Transportation
Planning /
Vehicle
Scheduling

Concluding remarks
Huge cost savings can be gained through efficient use of APS.
APS have been adopted in many industries.
Collaboration between partners in the supply chain including
share of information and transparency of business processes
is seen as a major driver of SCM performance.
No global optimize SCM button provided by APS.
Expertise needed.
Optimization models often require large computational effort.
APS, especially at the detailed scheduling level, do not
sufficiently consider application specific features.
APS are most successful for intra
intra-company
company supply chains
with centralized logistics control.