IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Forecasting Methods and Implementation of DRP (Distribution

Requirement Planning) Methods in Determining the Master Production
Schedule
To cite this article: Riana Magdalena and Theresia Suli 2019 IOP Conf. Ser.: Mater. Sci. Eng. 528 012049

View the article online for updates and enhancements.

This content was downloaded from IP address 91.241.164.115 on 12/06/2019 at 18:07

ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

Forecasting Methods and Implementation of DRP

(Distribution Requirement Planning) Methods in Determining
the Master Production Schedule

Riana Magdalena1,2 and Theresia Suli1

1
Industrial Engineering, Engineering Faculty,Atma Jaya Indonesian Catholic
University, Jln. Raya Cisauk Lapan No. 10, Sampora, Cisauk, Tangerang, Banten
15345, Indonesia.
2
riana.magdalena@atmajaya.ac.id

Abstract. PT ABC President Indonesia is a manufacturing company engaged in the food sector
by producing ABC noodles and Nu Greentea. At PT ABC President Indonesia, the production
scheduling process is one of the main processes carried out to determine the number of
products that must be produced. Based on the problems that occur regarding stock, it is
proposed that the best forecasting method is based on the time series of each Distribution
Centre (DC) and processing Distribution Centre (DC) requests using the Distribution
Requirement Planning (DRP) method. Where the DRP method is used, so that the company
can meet the demand of every DC so that there is no shortage of products and the production
process continues to run without the company experiencing losses. In addition, with the use of
DRP method there will be additional stock so that if there is a defect, the company still has
stock remaining so that it can still meet the customer demands for each DC and the product
distribution process runs more regularly.
Keywords: Forecasting Method, Distribution Requirement Planning Method, Stock.

1. Introduction
In this era, the industrial world is growing rapidly. This can be seen, where many manufacturing
companies are engaged in various industrial fields. To be able to survive and compete with other
manufacturing companies, a company must be able to produce products that have good quality and are
able to maintain the level of satisfaction of customers towards the products produced by the company.
One way to maintain and compete with other companies is to create a good production system and
one which is mutually integrated by considering many things. One of them considers the production
scheduling carried out and the distribution flow carried out in a company. Production scheduling is
one important aspect of a company where a good production scheduling strategy can minimize losses
and increase customer satisfaction with the product being produced. One of the losses that can happen
is that there are several requests from customers that cannot be fulfilled due to delay in the distribution
process to customers. So, if there is an unfulfilled demand, it will reduce the level of corporate service
to customers and can reduce the level of customer satisfaction with the product being produced. This
could cause the possibility that customers will switch to competing products.
PT ABC President Indonesia is a manufacturing company that produces noodles and beverages
such as Nu Greentea. In the world of the food industry, there are often uncertainties in the time and
number of products ordered by customers. This results in an excess or lack of stock which was

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

sometimes experienced by PT ABC President Indonesia where PT ABC President Indonesia

implemented the make to stock system. The existence of excess or lack of stock sometimes results in
losses such as the existence of slow moving (products that are not distributed), non-fulfillment of
customer demand, or the existence of products that cannot be consumed again. The importance of
strategies in production scheduling can be used as a solution to overcome the problems that occur.
Therefore, in this study, the topic raised was about production scheduling planning by
implementing the DRP method (Distribution Requirement Planning). The DRP method is a
distribution planning method using several control parameters to calculate product requirements at
each distribution point. The advantage of the DRP method is that the DRP method is centered on the
demand of each depot or agent so that the production process will be scheduled based on requests
from each depot or agent. Where using the DRP method can provide recommendations on the
production master schedule and can overcome the stock issues at PT ABC President Indonesia.

2. Methods
In this study, DRP (Distribution Requirement Planning) was used as a research model. DRM is
Distribution resource planning (DRP) is a method used in business administration for planning orders
within a supply chain. DRP enables the user to set certain inventory control parameters (like a safety
stock) and calculate the time-phased inventory requirements. The method in this study consisted of:
(1) preparation, (2) research design, and (3) data collection (4) data processing.

2.1. Preparation
In the preparation phase in this study, identification of real conditions was conducted at PT ABC
President Indonesia through observation (direct research in the field) as well as the interview process
with parties from several. In addition, in this phase literature studies from books and research
journals are carried out to discuss method used in this study, namely the forecasting techniques, DRP
method, and MPS.

2.2. Research Design

DRP is carried out to identify problems in the questions related to the topic of the issue to be discussed
which focused on the problems regarding the existence of undistributed products experienced by PT
ABC President Indonesia. By using DRP, product production is regulated to meet the demand until the
next product production arrives for each distribution center. In making MPS using the DRP method, a
forecast is needed from each DC to estimate the number of products that must be produced. With the
determination of good forecast methods, the forecast has a lesser error value so there is no excess
stock.

2.3. Data Collection

This research was conducted at PT ABC President Indonesia by means of observation (direct research
in the field) as well as the interview process with parties from several divisions on questions related to
the topic of the issue to be discussed which focused on the problems regarding the existence of
undistributed products experienced by PT ABC President Indonesia. The data collected consists of
historical data requests, lead times of each distribution center, stock from each DC and available
distribution centers.

2.4. Data Processing

The first data processing is to determine the time series plot of each DC. The time series plot
determination aims to find out the exact demand forecasting method in accordance with the available
data. Then the forecasting process is carried out using the selected method. After forecasting process,
the process of determining the proposed lot size for each DC is appropriate and continued with the
calculation of the Master Production Schedule with the DRP method by considering the safety stock

2
 ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

and the remaining stock. Where the results of the data processing are the MPS proposal for the 2018
period for the production of Alfamart private label noodles.

3. Result and Discussion

The method used in the study is the DRP method but is closely related to the DRP method with
forecasting. Because the data obtained from the company is only in the form of historical data requests
per DC so that forecasting needs to be done first so that the existing data can be used for subsequent
DRP calculations.The data collected was the data from distribution centre at PT ABC President
Indonesia for Alfamart ABC private label noodles, and lead time data and inventory for each DC
spread on the island of Java.

Table 1. Data Demand from each DC at Java Island

DC
Bulan DC 1 DC 2 DC 3 DC 4 DC 5 DC 6 DC 7 DC 8 DC 9 Total
10
January 1636 795 392 840 448 400 336 1060 1976 736 8619
February 1674 1078 672 756 728 220 560 860 2228 1060 9836
March 2215 404 336 504 112 414 0 1260 2264 732 8241
April 1400 514 504 196 616 612 168 780 1274 898 6962
May 1588 1404 322 728 728 580 952 100 2382 2170 10954
June 3450 846 868 1008 504 0 168 1940 2072 420 11276
July 2276 1010 224 84 700 220 784 1080 1803 916 9097
August 1872 1538 560 468 476 220 504 760 1760 620 8778
September 1160 794 210 264 434 384 420 625 1386 328 6005
October 1244 630 336 416 0 212 56 251 2218 534 5897
November 908 1102 126 340 42 280 196 500 2575 1156 7225
December 1334 772 252 290 156 360 796 880 1348 202 6390
The above data was obtained from requests for each DC Alfamart for 2017.

Table 2. Lead Time Data

Lead Time Lead Time
DC DC
(month) (month)
1 1 5 2
2 2 6 1
3 2 7 1
4 2 8 1
9 1

The data above was the ordering time data from each DC to the time the product arrived at each DC.

Table 3. Project Data on Hand for Every DC

DC Project On Hand DC Project On Hand
1 160 6 160
2 200 7 160
3 80 8 200
4 200 9 360
5 160 10 360

3
ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

The data above shows the project on hand data that is owned by each DC where the data above is
obtained from the product remaining in each DC at the end of 2017 (December 2017). Besides that,
the other data needed to calculate the DRP method is that the lot used is Fixed Order Quantity (FOQ)
of 40 pcs or multiples thereof for each order.

3.1. Forecasting
3.1.1. Time Series Plot
Time Series is used to find out the graphic form of each DC to determine the correct forecasting
method.

Figure 1. Time Series Plot DC 1

Based on the picture above, it can be seen that the form of the DC 1 time series plot graph shows the
form of non-trendy data due to the absence of rising or falling and non-seasonal demand trends due to
the absence of a repetitive pattern in the request data. This means forecasting methods that must be
used are Moving Average and Single Exponential.

Figure 2. Time Series Plot DC 2

Based on the picture above, it can be seen that the form of the DC 2 time series plot graph shows the
form of non-trendy and seasonal data with a period of 3 periods as in the 1-2-3 period and 10-11-12
period. Seasonal said despite having the same pattern at the end of the data, due to limited data taken
only 12 periods so that there is a possibility of a repeat pattern that is not visible on this graph. This
means that the forecasting method that must be used is Decomposition of Additive and Multiplicative
and Winter Additive and Multiplicative.

4
ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

Figure 3. Time Series Plot DC 3

Based on the picture above, it can be seen that the shape of the DC 3 time series plot graph shows the
form of trend (declining) data which can be seen in odd periods (1-3-5-7-9-11) which tends to
decrease and seasonal with interval 2 period seen in the pattern of periods 1-2-3, 3-4-5, 5-6-7, 7-8-9,
9-10-11. This means that the forecasting method that must be used is Decomposition of Additive and
Multiplicative and Winter Additive and Multiplicative.

3.1.2. Forecasting
Forecasting is done using MINITAB 17 software. Here are the results of forecasting:

Figure 4. Double Moving Average for DC 1

Forecasting results above, which must be considered, is the magnitude of the error value listed where
the error values listed are MAPE, MAD, and MSD.

3.1.3. Accuracy
After forecasting, all the resulting error values are recapitulated and compared to what method has the
smallest error value.

5
ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

Table 4. Accuracy Table from DC 1

Based on the recapitulation of DC 1 accuracy level it can be seen that the best forecasting method used
for DC 1 is the Double Moving Average method because the error value of the Double Moving
Average method is the smallest error value indicating that the error of the forecasting method is
smaller than the 2 methods others.

Table 5. Results of Forecasting Every DC

Period DC 1 DC 2 DC 3 DC 4 DC 5 DC 6 DC 7 DC 8 DC 9 DC 10
13 1229 723 115 350 198 292 331 549 1934 608
14 1229 1306 192 350 198 292 331 549 1934 608
15 1229 694 69 350 198 292 331 549 1934 608
16 1229 723 97 350 198 292 331 549 1934 608
17 1229 1306 24 350 198 292 331 549 1934 608
18 1229 694 2 350 198 292 331 549 1934 608
19 1229 723 0 350 198 292 331 549 1934 608
20 1229 1306 0 350 198 292 331 549 1934 608
21 1229 694 0 350 198 292 331 549 1934 608
22 1229 723 0 350 198 292 331 549 1934 608
23 1229 1306 0 350 198 292 331 549 1934 608
24 1229 694 0 350 198 292 331 549 1934 608

The table above is a request recapitulation table obtained from the MINITAB software according to
the chosen forecasting method. For the period, it starts with the period 13-24 which indicates that the
forecasting is forecasting for the period January - December 2018.

3.1.4. Distribution Requirement Planning (DRP)

Table 6. DRP Calculation Table for DC 1

6
ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

Calculation example:

• Gross Requirement: obtained from DC forecasting results 1

• Project on Hand: obtained from production at the end of 2017 for DC 1 for period 1
• Planned Order Receipt - Net Requirements = 1080 - 1069 = 11
• Net Requirement: Gross Requirement - Project on Hand = 1229 - 160 = 1069
• Planned Order Receipt: FOQ = 40 then multiples of 40 where 40 x 27 = 1080
• Planned Order Release: Based on lead time (1 month).

Data in Table 6 describes the calculation of Distribution Requirement Planning for DC 1, where the
data needed in this calculation is the gross requirement data, receipt schedule, and project on hand.
Gross requirement is the number of requests from each DC obtained from the prediction that has been
done previously with the chosen forecasting method. For the receipt schedule, a product that has been
scheduled will be ordered by the company. In this case the receipt schedule is 0 because the company
produces its own product without the presence of other companies to produce the noodle product. As
for the project on hand, the stock available to the company is taken from the remaining production in
the previous month. The results from Table 6 are scheduling of each distribution for each month by
considering things like gross requirement data, receipt schedule, and project on hand and lot size of
each distribution.

3.1.5. Master Production Schedule (MPS)

Table 7. MPS Table

MPS calculation is obtained from recapitulation of period 1 to period 12 by adding up each DC.
Data in Table 7 is an MPS table where the MPS value is obtained from the combination of each DC
where the merge is obtained by planned order release. The relation from tables 6 and 7 is that after the
calculation of table 6, it is obtained that the planned order release from each DC is combined so that it
becomes the total that must be produced in one month.

3.1.6. Analysis
Based on observations made at PT ABC President Indonesia, it can be seen that the current production
scheduling process is still not optimal. It is said that it is not optimal because there are still several
distribution centres that place orders simultaneously so that there is a bottleneck in the product
distribution process and there are some DCs that are not fulfilled; the demand caused by the existing
process has not been able to collect in line with existing requests.
Based on the problems at PT ABC President Indonesia above, the authors recommend Proposed
Forecasting Methods and Implementation of the DRP Method (Distribution Requirement Planning) in
Determining Production Master Schedules. In the forecasting method, the time series plot is carried
out in advance on the request data to determine the trend pattern of demand data where there are 2
types of data plots, namely seasonal trend and data. For trend data are time series data where there is a
tendency for demand to rise or fall. As for seasonal data, the data that is formed has a tendency to
repeat over a period of time.
After determining the demand pattern using a time series plot, it can be known that the forecasting
method is in accordance with the time series plot of each DC. For DC 1,4,5,6,7,8,9 and 10 because it
has a non-trendy and non-seasonal pattern, the Moving Average and Single Exponential Smoothing
forecasting methods are used. For DC 2 and 3 using the Decomposition and Winter methods.

7
ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

Furthermore, data processing is performed using forecasting methods that are in accordance with the
time series.
In the forecasting process, MINITAB software was used. In processing data by using this
software, in addition to obtaining forecasting results, an error value from the forecasting is also
obtained. This error value indicates the magnitude of the forecasting error or non-conformity that is
carried out with the conditions that will occur in the future. Therefore, in choosing the best forecasting
method, we need to pay attention to the error value where from all DCs look for a method that has the
smallest error value where the error value used is MAPE, MAD and MSD which will show a more
accurate level of data accuracy than forecasting data with use other methods. After forecasting it will
proceed to the calculation using the DRP method.
The use of the DRP method is intended to collect all requests from each DC by considering the
lead time, lot and the number of requests from each DC so that the company can meet the existing
demand without stock out or stock shortage. Forecasting data on data processing is used as a gross
requirement (GR) because the DRP method is assessed for the number of products to be produced for
the next 1 year period where this assessment uses historical data one year earlier.
The period in DRP starts from period 0-12 but in reality the period starts in period 1 while the use
of period 0 in the DRP method is due to the existence of project on hand which a stock before the
commencement of production is. In DRP there is also a safety stock where the stock is a stock of
deposits. However, in this case, the safety stock is assumed to be zero because the product produced is
a food product whose storage period cannot be too long so that if used as a safety stock can cause risks
such as expired products, product damage, and large storage costs.
The lot sizing method used for each DC is a fixed order quantity of 40 pcs. Where this
determination is based on the agreement of the company with Alfamart where the order is at least 1
box where 1 box contains 40 pcs of ABC noodles. So the ordering process can be done by ordering 40
pcs of noodles or multiples. The results of this DRP are Production Master Schedule or MPS which is
obtained from the sum of planned order releases for each DC in the same period.
Based on the research and analysis that has been done, the evaluation that must be done is the
importance of the accuracy of forecasting methods and the use of the DRP method can affect the
amount of product production each period. With the forecasting method and the proposed DRP
method, it can minimize unmet demand for each DC.

4. Conclusion
The best forecasting method to do on DC 1, 4, and 5,6,7,8,9,10 is Double Moving Average, DC 2 is
Decomposition Additive and DC 3 is Multiplicative Decomposition. In the calculation of MPS with
the DRP method, the results show that the number of production in period 1 is 6320 pcs, period 2 is
6240 pcs, period 3 is 6840 pcs, period 4 is 6200 pcs, period 5 is 6280 pcs, period 6 is 6760 pcs, period
7 is 6160 pcs, period 8 is 6240 pcs, period 9 is 6760, period 10 is 6200, and period 11 is 4960 pcs. By
scheduling production using the DRP method, there is a stock due to the use of FOQ lot size so as to
minimize the risk of not fulfilling the request if a defect occurs. Companies should forecast each DC
so that more optimal forecasting is done. Production scheduling should be done per DC so that there is
no miscalculation of product production. The limitations of the data in this study are due to historical
data collection that was carried out only for 12 months. Suggestions for further research, it is better to
take historical data for at least 2 years in order to see more clearly the time series plots available so
that the results will be more optimal.

5. References
[1] Abdillah, Adib Fahrozi 2009 Perencanaan dan Penjadwalan Aktivitas Distribusi Hasil
Perikanan dengan Menggunakan Distribution Requirement Planning (DRP). (Jawa Timur:
Universitas Pembangunan Nasional “Veteran”)

8
ISIEM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 528 (2019) 012049 doi:10.1088/1757-899X/528/1/012049

[2] Ong, Johan Oscar 2013 Implementasi Distribusi Requirement Planning dan Saving Matrix
untuk Minimimasi Total Biaya Distribusi di Industri Bahan Kimia. (Cikarang: President
University)
[3] Pasaribu, Theresia Oshin Rosmaria 2014 Penentuan Metode Peramalan Sebagai Dasar
Penentuan Tingkat Kebutuhan Persediaan Pengaman Pada Produk Karet Remah SIR 20.
(Depok: Universitas Gunadarma)
[4] Setyawan, Eko 2016 Analisis Peramalan (Forecasting) Produksi Karet (Hevea Brasiliensis) di
PT Perkebunan Nusantara IX Kebun Sukamangli Kabupaten Kendal. (Semarang:
Universitas Wahid Hasyim)
[5] Supriyadi 2016 Penjadwalan Produksi IKS-Filler pada Proses Ground Calcium Carbonate
menggunakan Metode MPS di Perusahaan Kertas. (Serang: Universitas Serang Raya)
[6] Surya, Regina Steven. 2013 Implementasi Metode Distribution Requirements Planning (DRP)
pada CV Karya Mandiri Sejahtera di Surabaya. (Surabaya: Universitas Surabaya)
[7] Wardah, Siti 2016 Analisis Peramalan Penjualan Produk Keripik Pisang Kemasan Bungkus.
(Tembilahan: Universitas Islam Indragi)