International Journal of Physical Distribution & Logistics Management

Distribution Requirements Planning: A Generalised System for Delivery Scheduling in a Multi-Sourcing

Logistics System
Chrwan‐jyh Ho,
Article information:
To cite this document:
Chrwan‐jyh Ho, (1990) "Distribution Requirements Planning: A Generalised System for Delivery Scheduling in a Multi‐
Sourcing Logistics System", International Journal of Physical Distribution & Logistics Management, Vol. 20 Issue: 2, pp.3-8,
https://doi.org/10.1108/09600039010144869
Permanent link to this document:
https://doi.org/10.1108/09600039010144869
Downloaded by UNIVERSITAETS UND LANDESBIBLIOTHEK DARMSTADT At 14:45 17 June 2019 (PT)

Downloaded on: 17 June 2019, At: 14:45 (PT)

References: this document contains references to 0 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 613 times since 2006*
Users who downloaded this article also downloaded:
(1995),"The integrated logistics management system: a framework and case study", International
Journal of Physical Distribution &amp; Logistics Management, Vol. 25 Iss 6 pp. 4-22 <a href="https://
doi.org/10.1108/09600039510093249">https://doi.org/10.1108/09600039510093249</a>
(1990),"Integrated Logistics Systems with Distribution Resources Planning", Logistics Information Management, Vol. 3 Iss 4
pp. 26-34 <a href="https://doi.org/10.1108/eb007519">https://doi.org/10.1108/eb007519</a>

Access to this document was granted through an Emerald subscription provided by emerald-srm:514619 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please visit
www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.


A
generalised version of a DRP system,
which enables evaluation of supplier
performance.
Distribution
Downloaded by UNIVERSITAETS UND LANDESBIBLIOTHEK DARMSTADT At 14:45 17 June 2019 (PT)
Requirements
Planning:
A Generalised System for
Delivery Scheduling in a
Multi-sourcing Logistics
System
Chrwan-jyh Ho
Introduction
Distribution requirements planning (DRP) has been
suggested as a scheduling method in the physical
distribution system[l,2]. DRP originated from its
manufacturing counterpart, material requirements planning
(MRP). A well-developed DRP system enables logistics
management to schedule product delivery better, more
effectively deploy inventories to meet customer
requirements, and be more responsive to uncertain events
occurring in the distribution network.
With the popularity of MRP as a production scheduling
and inventory control method in manufacturing firms, we
can reasonably expect that there will be more DRP users
in the near future. However, the slow adoption of DRP
in logistics system planning has been reflected by relatively
few studies in DRP in the recent literature in logistics.
This article aims to present a generalised version of a DRP
system which is flexible and responsive in the multi-
sourcing, multi-echelon distribution network.
A brief review of the relevant literature in logistics system
planning is given, followed by a presentation of the detailed
DISTRIBUTION REQUIREMENTS PLANNING 3
DRP mechanism. A three-echelon distribution system is
illustrated to elaborate the advantages of this DRP system.
Finally, future research issues are discussed.
Previous Research
Due to the changing technological environment,
distribution management faces new levels of complexity.
The significance of the interaction between the distribution
function with other organisational functions has been well
recognised[3]. Advances in information processing and
computer technology have brought about the development
of numerous practical planning models for analysing a
distribution network. A brief review of management
science and logistics literature identifies many such
models. The large-scale models include the use of such
management science technologies as linear programming
[4,5,6], and mixed-integer programming [7,8]. Long Range
Environmental Planning Simulator (LREPS)[9] typifies the
simulator model to deal with logistics system planning.
Bookbinder and Locke conducted a comparative study to
evaluate just-in-time (JIT) distribution[10]. Chakravarty and
Shtub presented a set of guidelines for safety stock
allocation in a two-echelon distribution system[11].
DRP was suggested as an integrated approach to
scheduling delivery and controlling inventory for a logistics
system[2]. A DRP schedule is used to address when and
how much to deliver. Vollmann, Berry, and Whybark[12]
suggested that DRP should be viewed as one part of
demand management, which provides a linkage between
the market, demand management and master production
schedule. Distribution resources planning (or DRP II) was
proposed to incorporate the planning of transportation
mode selection (e.g. co-ordination of trucks/rail cars from
plants to distribution centres), inventory control at various
channel levels, and effective allocation of distribution
resources (e.g. development of staffing/manpower
requirements) [1,13].
Practitioners in logistics introduced DRP processing logic
without detailed elaboration. Specifically, it deals with an
"arborescent" distribution network in which a given
channel member has a single source of supply. The
problem of multi-sourcing in logistics is not well explored
due to the general practice of using centralised distribution
centres to serve a host of retailers. Nevertheless, the
multi-sourcing of product supply will not only enhance the
delivery flexibility in logistics systems but also reduce the
risk of supply disruption in the single-sourcing situation.
A Generalised DRP System
The advocates of DRP maintain that DRP or DRP II is
an integrated approach to filling the gap between
manufacturing and physical distribution. DRP, based on
4 IJPD & LM 20,2

the time-phased order point approach, is a scheduling Time-phased Replenishment Requirements

system which derives the master production schedules
at the plant level from the warehouses or distribution
centres. By using DRP schedules, the plant is capable
of reacting to changes in customer demand in the form
of rush orders or order cancellations.
A three-echelon logistics system, as shown in Figure 1,
is used to demonstrate the scheduling mechanism of DRP.
There are three retailers (Ri), two distribution centres
(DCi), and two plants (Pi) in this distribution system in
which the arc represents the material flow from one point
Downloaded by UNIVERSITAETS UND LANDESBIBLIOTHEK DARMSTADT At 14:45 17 June 2019 (PT)
Time-phased replenishment requirements represent
delivery schedules for a product on the planning horizon.
Table I indicates the time-phased requirements that show
the demand requirements of R2. For instance, the
schedule indicates that 30 units are needed at the start
of week 2 and another 50 units in week 3 and so on. The
replenishment quantities may come from different sources,
including customer orders, or sales forecasts. It should
be noted that it is not necessary to use weekly DRP
schedules. As the popularity of the concept of JIT spreads
to logistics, which requires the frequent receipts of small

of origin to one destination. There are three major inputs batches, it is up to the planners to determine an
for a DRP system: the time-phased replenishment appropriate time unit for the DRP system. For a system
requirements of retailers; the inventory records of all the with short delivery lead time, the daily DRP schedule
channel members; and the "bill of material" type of should provide more timely information than the weekly
distribution network structure, (a detailed description of schedule.
these information inputs can be found in[l]). A brief
discussion of DRP mechanisms is warranted here because Inventory Records
they are slightly different from the conventional MRP or
DRP inputs. Inventory records store information on the status of each
item by time period. The information includes in-transit
inventory and expected on-hand inventory. It also includes
other information such as the sources of supply,
Figure 1 An Example of a Distribution Network
transportation lead time, transportation mode, and lot-
sizing rule. In a DRP schedule, the expected on-hand
inventory of a channel member in a given time period is
calculated as follows.
OHt = OHt-1 + SRt - DRt (1)
where:
OHt = the on-hand inventory at the end of period t,
SRt = the schedule receipts in period t, and
DRt = the demand requirements in period t.
As shown in Table I, the expected on-hand inventory in
week 3 is calculated as: 10 + 0 - 50 = -40. Note that,
in the conventional MRP system, the on-hand inventory
is calculated cumulatively. For instance, the on-hand
inventory in week 5 is computed as: -40 + 0 - 20 =
-60. This negative inventory level does not indicate the
net requirement is 60 units in this period. Instead, the
net requirement is just 20 units in week 5. Suppose that

Table I. An Example of a DRP Schedule

Lead time: W1:l

Period 1 2 3 4 5 6 7 8 9 10 11 12

Demand requirements 30 50 20 40 50 60
Scheduled receipts
Expected on-hand inventory 40 40 10 -40 -40 -60 -100 -100 -100 -100 -150 -150 -210
Net requirements 40 20 40 50 60
Planned shipments 40 20 40 50 60

the lot-for-lot lot-sizing rule is used in the DRP. That is,
DRP planners order just enough to meet the net
requirements of a certain period. After offsetting the lead
time for the net requirements in each period, the planned
shipments (equivalent to the planned order releases in
MRP schedules) can then be determined.
Structure of the Distribution Network
The distribution network in Figure 1 needs to be
restructured into a typical "product structure", depicted
Downloaded by UNIVERSITAETS UND LANDESBIBLIOTHEK DARMSTADT At 14:45 17 June 2019 (PT)
in Figure 2, wherein retailers are treated as end items
in its manufacturing counterpart. Plants and distribution
centres are viewed as the components. The demand
requirements for plants are derived from distribution
centres or retailers. For instance, DC1 is supplied from
P1 while R2 is replenished by DC1 and DC2. This
extension of the bill of materials into a distribution network
allows the use of the standard MRP explosion process
to link the logistics system with all the manufacturing
planning systems.
The number in parentheses represents the percentage
split of order releases in the next echelon in Figure 2.
The percentage split is the allocation of product order to
a certain supplier in multi-sourcing situations. In the
example, R2 gets 20 per cent of the total supply from DC1
and 80 per cent from DC2. The determination of
percentage splits may depend on such factors as
transportation lead time, manufacturing capacity, supply
reliability, and delivery flexibility.
This DRP system considers the trans-shipment option
which delivers the products directly from the plant to
retailers. For example, in Figure 2, R3 may receive a
delivery directly from P2 if R3 runs out of stock due to
a surge in customer demand. Trans-shipment is a common
delivery practice to reduce the delivery time in the case
of rush orders.
Figure 2. A Distribution Network Structure
DISTRIBUTION REQUIREMENTS PLANNING 5
DRP Processing in Multi-Sourcing
DRP processing involves taking time-phased requirements
of channel members at the bottom level in the distribution
network and "exploding" them into the demand
requirements of members at the upper echelons. When
MRP logic is applied[14], a series of planned shipments
are used to derive the demand requirements of P1 and
P2. As shown in Table II, a DRP schedule forDC1indicates
that the demand requirements of DC1 are derived from
the planned shipments of its target distribution centres,
R1 and R2, by using the following equation:
DRj =α1 x {NR(R1j + LT)} + α2 x {NR(R2j + LT)} (2)
where:
DRj = demand requirements for period j,
LT = transportation lead time from DC1 to Ri,
NR(Rij+LT) = net requirements for Ri in period j+LT,
and
αi = percentage split of demand requirements of Ri
to DC1.
Using equation (2), the demand requirements forDC1in
week 2 is calculated as: 1 x 30 + (0.2) x 40 = 38. The
transportation lead time is then offset to derive the
demand requirements for DC1. When the "explosion"
process is completed, DRP schedules can be used to
address the questions of timing and quantity of delivery
from origin to destination. In Table II, it indicates that the
DRP planners at DC1 should order 23 units immediately
from P1. Suppose that no uncertain events occur in the
near future, the planners at P1 should ship 4, 38, 40, 50,
10, and 32 in weeks 3, 4, 6, 7, 8, and 10, respectively.
In the case of a multi-sourcing situation, multiple rows
of planned shipments are required. In Table II R2 has two
suppliers, DC1 and DC2. A net requirement of 40 units
is needed at R2 in week 3. Based on the percentage splits
of this product between these distribution centres and the
delivery lead times, the planned shipments from each
distribution centre are placed on separate rows, which are
used to derive the demand requirements forDC1or DC2.
It is shown that the planners at DC1 need to ship, 8, 4,
8, 10, and 12 units and those at DC2 need to ship 32, 16,
32, 40, and 48 units in weeks 2, 4, 5, 9 and 11, respectively.
With this added feature, the following advantages can be
expected.
Advantages of Generalised DRP Systems
In addition to the multi-sourcing problems addressed
previously, this generalised DRP system provides the
following advantages:
Handling Exceptional Situations
A DRP system should be flexible enough to accommodate
any exceptional situations. Suppose that a typical channel
 6 IJPD & LM 20,2

Table II. An Illustration of the DRP Explosion Process

Lead time: = 1 R1
Period 1 2 3 4 5 6 7 8 9 10 11 12

Demand requirements 60 30 40 50 20
Scheduled receipts
Expected on-hand inventory 30 30 30 -30 -30 -30 -60 -60 -100 -150 -150 -150 -170
Net requirements 30 30 40 50 20
Planned shipments 30 30 40 50 20
Downloaded by UNIVERSITAETS UND LANDESBIBLIOTHEK DARMSTADT At 14:45 17 June 2019 (PT)

Lead time: DC1 = 1; DC2 = 1 Percentage split = DC1 = 0.2; DC2 = 0.8 R2
Period 1 2 3 4 5 6 7 8 9 10 11 12

Demand requirements 30 50 20 40 50 60
Scheduled receipts
Expected on-hand inventory 40 40 10 -40 -40 -60 -100 -100 -100 -100 -150 -150 -210
Net requirements 40 20 40 50 60
Planned shipments DC1 8 4 8 10 12
Planned shipments DC2 32 16 32 40 48

Lead time = 2 R3
Period 1 2 3 4 5 6 7 8 9 10 11 12

Demand requirements 30 40 50 30 40 20
Scheduled receipts
Expected on-hand inve ntory 50 20 20 -20 -20 -20 -70 -100 -100 -140 -140 -160 -160
Net requirements 20 50 30 40 20
Planned shipments 20 50 30 40 20

Lead time = 1 DC,

Period 1 2 3 4 5 6 7 8 9 10 11 12

Demand requirements 38 4 38 40 50 10 32
Scheduled receipts
Expected on-hand inventory 15 15 -23 -23 -27 -65 -65 -105 -155 -165 -165 -197 -197
Net requirements 23 4 38 40 50 10 32
Planned shipments 23 4 38 40 50 10 32

Lead time = 1 DC2

Period 1 2 3 4 5 6 7 8 9 10 11 12

Demand requirements 20 32 66 62 40 60 48
Scheduled receipts
Expected on-hand inventory 60 40 8 8 -58 -120 -120 -160 -160 -220 -220 -268 -268
Net requirements 58 62 40 60 48
Planned shipments 58 62 40 60 48

Lead time = 2 P1
Period 1 2 3 4 5 6 7 8 9 10 11 12

Demand requirements 23 4 38 40 50 10 32
Scheduled receipts
Expected on-hand inventory 25 2 2 -2 -40 -40 -80 -130 -140 -140 -172 -172 -172
Net requirements 2 38 40 50 10 32
Planned shipments 2 38 40 50 10 32

Table II. continued
Lead time = 2
Period 1 2 3 4
Demand requirements 58
Scheduled receipts
Expected on-hand inventory 30 30 30 -28
Net requirements 28
Planned shipments 28 62
Downloaded by UNIVERSITAETS UND LANDESBIBLIOTHEK DARMSTADT At 14:45 17 June 2019 (PT)
supply is shipped from P2 to DC2, and then R2, as shown
in Figure 1. If there is a sudden surge in customer demand
at R2 and there is insufficient stock at DC2 to cover it,
then it is logical to deliver from P2 to meet the customer
demand at R2. Therefore, the stockout duration would
be reduced. Furthermore, Figure 3 illustrates a situation
in which there may be approximately equal distance
between distribution centres and retailers; R1 and R2 can
actually be served by either DC1 or DC2 due to its
geographic proximity to both distribution centres. In this
situation, the generalised DRP system can provide time-
phased requirements for the suppliers.
| Figure | 3. | An Illustration of Geographic Proximity
Better Integration of Distribution and Manufacturing
Functions
As with any DRP system, the DRP system described here
helps integrate distribution and manufacturing with master
production schedules at the plant level. Instead of
forecasting the demand for thefinalproducts at the plant
level, production planners can obtain inputs from various
channel members in the distribution network with a DRP
system. Furthermore, this generalised DRP can
accommodate emergency orders and "peg" the sources
of these orders in multi-sourcing situations. Production
planners can then determine the timing of expediting action
effectively.
DISTRIBUTION REQUIREMENTS PLANNING 7
P2
5 6 7 8 9 10 11 12
62 40 60 48
-90 -130 -130 -130 -190 -190 -238 -238 -238
62 40 60 48
40 60 48
Built-in Evaluation System
On-time delivery is an extremely important factor in
operational scheduling in JIT logistics systems. If a channel
member has control over which supplier to order from,
then a reward system can be established to provide the
incentive to encourage suppliers to deliver on time. It is
not an uncommon practice even for a business organisation
which owns the distribution network, such as an oil
company, to treat each channel member as an individual
profit centre. A larger percentage split may be given as
a reward to the supplier who consistently performs well
in on-time delivery. With separate rows of planned
shipments for each supplier as shown in Table II, it is easy
to keep track of the performance of each supplier.
Future Research Issues
Operational problems exist in this DRP approach which
may concern logistics practitioners. The following research
issues are identified as attempts to study the effectiveness
of DRP in dealing with distribution problems:
Conducting a Feasibility Study on Multi-Sourcing
Strategy in a Logistics System
In distribution, single-sourcing is a commonly used
strategy. However, there are advantages to adopting a
multi-sourcing strategy, such as responsiveness and
flexibility, as discussed previously. The generalised DRP
system outlined here provides an experimental vehicle to
compare these two distribution strategies under such
operating environments as demand variability, uncertainty
in delivery lead time, or structural change in logistics
network. Furthermore, the impact of risk-pooling in multi-
sourcing can also be evaluated. This involves issues
regarding how risk should be spread, and whether the
change in percentage splits is warranted. To address these
issues, the percentage splits used in the DRP system must
be carefully examined.
Determination of Percentage Splits
A retailer receives goods from a distribution centre with a
shorter delivery lead time, more speedily. Thus, it seems
 8 IJPD & LM 20,2
reasonable that a larger percentage split should be
allocated to such a warehouse. However, this practice may
contradict the emphasis on suppliers on-time delivery,
discussed earlier. Whether a reliable supplier is better
than a speedy supplier is the question to be addressed
here. Therefore, the selection of decision criteria to
determine percentage splits among the suppliers is another
interesting research issue. A simulation experiment can
be conducted to test alternative decision criteria to
determine percentage splits.
Downloaded by UNIVERSITAETS UND LANDESBIBLIOTHEK DARMSTADT At 14:45 17 June 2019 (PT)
Sensitivity Analysis in Percentage Splits
The change in a channel member's percentage splits
should affect DRP performance because it affects the
order quantity from suppliers. Since a lot-sizing decision
has a great impact on DRP performance [15], the lot-sizing
performance can further be confounded by the
determination of percentage splits. A sensitivity analysis
in percentage splits can be conducted to evaluate the
performance of such commonly used lot-sizing rules in
DRP as lot-for-lot or economic order quantity.
Summary
This article presents a generalised version of a DRP
system aimed at increasing delivery flexibility and
enhancing responsiveness to customer demand in the
distribution network. This modified DRP is demonstrated
to deal with an operating environment in which multi-
sourcing is a desirable distribution strategy. This system
also enables a channel member to evaluate the
performance of its suppliers using the criterion of on-time
delivery. Finally, related research issues are discussed.
Attempts to address such issues should help accelerate
the adoption of DRP in the area of logistics.
References
1. Martin, A.J., Distribution Resource Planning, Prentice-
Hall, Englewood Cliffs, NJ, 1983.
2. Stenger, A.J. and Cavinato, J.L., "Adapting MRP to the
Outbound Side-Distribution Requirements Planning",
Chrwan-jyh Ho is Associate Professor of Operations Management at Oklahoma State University, USA.
Production and Inventory Management, Vol. 20 No. 4,
1979, pp. 1-14.
3. Bowersox, D.J., Logistical Management, Macmillan
Publishing Company, Inc., New York, 1974.
4. Markland, R., "Analyzing Multi-Commodity Distribution
Networks Having Milling-in-Transit Features",
Management Science, Vol. 21, 1975, pp. 1405-16.
5. Rees, L.P., Clayton, E.R. and Taylor, B.W. III, "A Linear
Programming Model of a Multi-period, Multi-commodity
Network Flow Problem", Journal of Business Logistics,
Vol. 8 No. 1, 1987, pp. 117-38.
6. Tomlin, J., "Minimum Cost Multi-Commodity Network
Flows", Operation Research, Vol. 14, 1966, pp. 45-51.
7. Geoffrion, A.M. and Graves, G.W., "Multicommodity
Distribution System Design by Bender Decomposition",
Management Science, Vol. 20, 1974, pp. 17-41.
8. Geoffrion, A.M. and Master, R.E., "Integer
Programming Algorithms: A Framework and State-of-the-
Art Survey", Management Science, Vol. 18, 1972, pp.
465-91.
9. Bowersox, D.J., "Planning Physical Distribution
Operations with Dynamic Simulation", Journal of
Marketing, January 1972, pp. 17-25.
10. Bookbinder, J.H. and Locke, T.D., "Simulation Analysis
of Just-in-Time Distribution'', International Journal of
Physical Distribution & Materials Management, Vol. 16
No. 7, 1986, pp. 31-45.
11. Chakravarty, A.K. and Shtub, A., "Simulated Safety
Stock Allocation in a Two-echelon Distribution System'',
International Journal of Production Research, Vol. 24 No.
5, 1986, pp. 1245-53.
12. Vollmann, T.E., Berry, W.L. and Whybark, D.C.,
Manufacturing Planning and Control Systems, Richard
D. Irwin, Homewood, Illinois, 1987.
13. Dube, W.R., "Closed Loop for Manufacturing and
Distribution", International Journal of Physical
Distribution & Materials Management, Vol. 16 No. 1,
1986, pp. 5-13.
14. Orlicky, J.J., Material Requirements Planning, McGraw-
Hill, New York, 1975.
15. Bookbinder, J.H. and Health, D.B., "Replenishment
Analysis in Distribution Requirement Planning", Decision
Sciences, Vol. 19 No. 3, 1988, pp. 477-89.
 This article has been cited by:

1. S.T. Enns, Pattita Suwanruji. 2004. Work load responsive adjustment of planned lead times. Journal of Manufacturing
Technology Management 15:1, 90-100. [Abstract] [Full Text] [PDF]
2. Winfried Krieger. Literaturverzeichnis 203-213. [Crossref]
Downloaded by UNIVERSITAETS UND LANDESBIBLIOTHEK DARMSTADT At 14:45 17 June 2019 (PT)