The Applications of Ordering Materials using Time Series

Forecasting with CB-Predictor

Tri Pujadi

, Witarsyah

, Haris Setia Budi

, Wahyu Sardjono

, Bachtiar H. Simamora

and Ximing Ruan

Information Systems Department, School of Information Systems, Bina Nusantara University, Jakarta, Indonesia

Department of Management, Binus Business School, Bina Nusantara University, Jakarta, Indonesia

MSc-International Management, Bristol Business School, University of the West of England, U.K.

ximing.ruan@uwe.ac.uk

Keywords: Inventory, Forecast, CB Predictor, Safety Stock, Unified Modeling Language.

Abstract: Shortage of raw materials may occur in manufacturing companies, which is caused by inaccurate orders for

raw materials, and lack of raw material supplies. This problem causes inefficient costs due to the production

process, the possibility of having to emergency procurement to fulfill customer orders. The solution to this

problem is to develop a web-based system that supports ordering of raw materials. The calculation based on

the estimated time series with the CB-Predictor. The methodology in the calculation is (1) collecting historical

data on the use of raw materials, step (2) forecasting raw material needs, step (3) calculating the order quantity

based on forecasting data, by comparing the deterministic method and the probabilistic method. For

calculation of safety stock for each raw material, for situations outside normal conditions, for example

increasing orders. The design method, the system to be developed uses the Unified Modeling Language

(UML) modeling language based on the concept of Object-Oriented Analysis and Design (OOAD). The result

is a web-based system application model to support a more efficient and accurate calculation of ordering raw

materials. With the proposed application of information systems, the company can estimate raw material

needs more quickly and accurately and can determine the quantity of orders that are tailored to the needs. So

that the costs associated with ordering and storing raw materials can be minimized.

1 INTRODUCTION

An important factor that influences serving customers

is the availability of products. For this reason, it is

necessary to order raw materials for production in the

right quantity. If raw materials can be ordered in

sufficient quantity and time, the production process

can run smoothly because the raw materials are

available, and the costs associated with inventory can

also be minimized. The cost of inventory consists of

the cost of ordering, storage costs and backorder costs

(Undersander et all 2017).

Unavailability of raw materials, resulting in loss

of production processes, expensive order costs, and

reduced customer trust, may have to pay inefficient of

inventory costs. For all these increase costs, because

the company cannot fulfill customer orders on time.

According to Mart et all (2013) timelines must be

measured correctly by the company from the

beginning of the order being recorded, production is

carried out, until the goods are delivered to the

customer. Important factors (Citra et all, 2013)that

support the punctuality of time are the right quantity

of raw material orders, and the time to place orders

for these raw materials. If the raw materials can be

ordered in the right quantity and at a time, the

production process can run smoothly because the raw

materials are always available. So that inventory costs

can be minimized (Irmayanti et all, 2019).

This study aims to develop a material ordering

process information system model that can assist in

arranging and managing raw material stocks quickly

and precisely so that the company will be able to

fulfill customer orders on time and can maintain

credibility and trust in the eyes of its customers.

The purposes for making this ordering model

include:

1. Make a planning regarding forecasting raw

material requirements in accordance with

historical production data.

2. Provide suggestions regarding the method of

ordering raw materials in the right quantity,

1316

Pujadi, T., Witarsyah, ., Setia Budi, H., Sardjono, W., H. Simamora, B. and Ruan, X.

The Applications of Ordering Materials using Time Series Forecasting with CB-Predictor.

DOI: 10.5220/0010964400003260

In Proceedings of the 4th International Conference on Applied Science and Technology on Engineering Science (iCAST-ES 2021), pages 1316-1321

ISBN: 978-989-758-615-6; ISSN: 2975-8246

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

considering the comparison of storage costs and

the cost of ordering raw materials. As well as

providing suggestions when to order raw

materials, in accordance with the forecasting of

raw material needs that have been done

previously?

3. Making suggestions regarding the safety stock

quantity of each type of raw material.

Propose an information system model to support the

process of recording raw material stocks that are out

and received, the calculation of forecasting raw

material requirements, and the calculation of the

quantity of raw material orders

The results of implementing the system model

that have been made will be useful for companies to:

a. Improve the accuracy of determining raw material

requirements in future periods.

b. Improve the accuracy of recording raw material

stocks in the warehouse.

c. Increase the accuracy of determining the quantity

of raw material orders.

d. Minimizing storage costs and ordering costs of

raw materials.

e. Increase the work efficiency of employees in the

warehouse of raw materials and the purchasing

department because with the system they can

work faster.

f. Assisting the warehouse in calculating the raw

material requirements for each incoming order. g)

Increase customer satisfaction by providing on

time delivery of orders.

g. Increase the company's credibility in the eyes of

customers.

Inventories are classified into raw materials, work

in process, finished goods, supporting materials,

complementary materials, components stored in

anticipation of demand. Inventory control is a very

important managerial function, because the majority

of companies involve large investments in this aspect

(20% to 60%). This is a dilemma for the company.

When inventory is excess, storage costs and the

required capital increase. The excess supply also

makes capital stagnate, the capital should be invested

in other sectors that are more profitable (opportunity

cost). Conversely, if the inventory is reduced, it can

cause out of raw materials (stock out). If the company

does not have sufficient supplies, emergency

procurement costs will be more expensive, another

impact is consumer disappointment with the

company(Sukmawati et all, 2009)..

The method of ordering raw materials can be

classified as Fig. 1 below (Chen et all, 2009):

Figure 1: Classification of method an ordering material.

To select the method of ordering raw materials

according to the demand data pattern. Demand

patterns are grouped into:

 Static data, if demand is a stationary data pattern,

or with a tendency to be constant / stable.

 Dynamic data, which is demand with data patterns

that fluctuate or tend to move, also known as

"lumpy demand"

Ordering is the procurement of goods or the

purchase of goods and services for companies that

have been regulated in their supply chain. There are

two types of procurement of goods, namely direct

procurement, and indirect procurement. This is

related to the purpose of procurement, whether to

support the production process (production related) or

non-production related. Good procurement

management will improve ordering services which in

turn result in budget savings and simplify the

procurement process so that it will be more efficient

(Nadella et all 2020).

There are three areas that will be supported by

online procurement of goods and services, namely in

the order transaction process, inventory management

and support in marketing. Thus, the use of e-

procurement will impact four business to business

(B2B) activities, namely search and identifying of

right products, order processing, monitoring &

control as well as coordination between the company

and its partners.

Forecasting is the prediction of the value of a

variable based on the known value of that variable or

related variables. Forecasting can be based on

appraisal expertise, which in turn is based on

historical data and experience. The main reason for

forecasting is because of the grace period between

awareness of future events or needs and the events

themselves (Salais-Fierro et all 2020). If the grace

period is zero or small, planning and forecasting are

not required, whereas if the grace period is long and

the outcome of the event depends on known factors,

The Applications of Ordering Materials using Time Series Forecasting with CB-Predictor

1317

forecasting is needed to determine when an event will

occur or arise, then appropriate action can be taken

done.

According to Subramanian and Render,

(Voulgaris (2019) forecasting is the art and science of

predicting future events with some form of

mathematical model, it can be a subjective or intuitive

prediction about the future or it can also include a

combination of mathematical models that are adapted

to good judgment by managers.

Forecasts are usually classified based on the

underlying future time horizon:

a. Short-term forecasting, usually used to plan

purchases, work scheduling, number of workers,

assignments, and production levels, and the time

span reaches one year but generally less than three

months.

b. Forecasting is medium-term, typically of three

months to three years, and is very useful in sales

planning, production planning and budgeting,

cash budgeting, and analyzing various operating

plans.

c. Long-term forecasting, usually spanning three

years or more, is used in planning new products,

capital expenditure, facility locations, or

expansion and research and development

Today, much software are available that support

forecasting calculations, so that users can more easily

perform forecasting calculations

CB Predictor is a type of software that can support

forecasting (Goldman 2002). CB Predictor is a

program built into Crystal Ball. This software is a

program for predicting data that will occur in the

future by analyzing past data. In the gallery of this

program, there are eight forecasting models (Glasser

1969):

a. Single Moving Average is a forecasting method

used for stationary data (does not contain seasonal

or trend elements).

b. Double Moving Average is a forecasting method

used for data containing trend elements.

c. Single Exponential Smoothing, is a forecasting

method used for stationary data (does not contain

seasonal or trend elements)

d. Double Exponential smoothing is a forecasting

method used for data that contains trend elements.

e. Seasonal Additive is a forecasting method used

for data containing seasonal elements.

f. Holt Winter’s Additive is a forecasting method

used for data that contains elements of seasonality

and trends.

g. Seasonal Multiplicative, is a forecasting method

used for data containing seasonal elements. This

method is the best method for data with the

highest aggregation, such as product sales and

data on raw material requirements.

The calculation of the Seasonal Multiplicative

method is influenced by

and

. The first step is

determine the seasonal elements of n periods, to

determine the seasonal term.

The block replacement process is a replacement

action that is carried out at a fixed interval (Dekker et

all, 1991). This method is applied by replacing the

damage that occurs at intervals (0, tp) by ignoring any

replacements that occur during that time interval, as

well as making preventive changes at each interval tp

but constantly (Bahtera, 2017).

Block replacement allows replacement in a close

period, where the new components installed after

replacement of the damage must be replaced again at

the time of preventive replacement (tp).

(1)

Where (1):

D(t

) = unity time downtime

= preventive replacement time interval

H(t

) = expected amount of damage at the interval

(0, t)

= downtime that occurs due to replacement

damage

= downtime due to preventive replacement

2 METHOD

The method of calculation for forecasting and

inventory of raw materials is carried out through the

following steps:

1. From the data on the use of raw materials for the

last three years, then forecasting the raw material

needs for the next twelve periods is carried out

using the help of software, namely CB Predictor.

2. From several forecasting results for each type of

raw material, a forecasting method is chosen with

the smallest error percentage value (MAPE 

Mean Absolute Percentage Error (Salais-Fierro et

all, 2013).

3. After the prediction result is determined according

to the method with the smallest error rate, the

Variability Coefficient (VC) calculation is

performed. The calculation of VC is the

calculation of the coefficient from the results of

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

1318

forecasting calculations to determine the

deterministic method for calculating the quantity

of raw material orders, where if the VC is less than

0.25 then the monthly raw material requirements

are stable, and to determine the quantity of raw

material orders, the EOQ (Economic Order

Quantity) method is sufficient. Meanwhile, if the

VC is greater than 0.25, the raw material needs are

unstable each month, sometimes it increases and

sometimes it decreases, so to determine the

quantity of the order, a comparison of several

methods must be used, namely the Silver & Meal

(S&M), Wagner & Within (W&W) method. ), and

Part Period Balancing (PPB).

The system design method uses an object-based

information system development method using the

artifact Unified Model Language (UML). The activity

begins with an analysis of the current information

system. This analysis is to determine user information

requirements and user interaction with existing

system functions. System requirements are used to

propose system improvements related to information

availability and system functions

3 RESULT AND DISCUSSION

Current Business Processes

The company does not yet have a computer-based

raw material ordering system. Recording of raw

material inventory and ordering of raw materials is

done manually. The rich picture in Fig. 2 describes

the current business process for procuring materials.

Figure 2: Rich picture current business process.

For every month, the raw material warehouse

department sends a record of the amount of raw

material stock to the purchasing department. Based

on this raw material record, the purchasing

department then determines when to place an order

for raw materials and what is the order quantity. This

determination is not based on any calculation, but

only based on intuition. Then the purchasing

department makes a purchase note containing the

types of raw materials to be purchased and their

quantity. For ordering local raw materials, the

purchase receipt that has been made by the

purchasing manager is sent directly to the supplier via

a fax machine. As for imported raw materials, a

purchase note is given to the director, then the

director orders directly to the supplier. In addition to

purchase notes, the purchasing manager creates a

transaction report for the purchase of raw materials

and sends it to the accounting department.

If the raw materials have arrived at the warehouse,

then the raw material warehouse section will match

the purchase note with the travel letter from the

supplier whether the raw materials sent by the

supplier match the raw materials ordered. Raw

materials that have been checked and are in

accordance with the order are then entered into the

raw material warehouse and the head of the raw

material warehouse records the raw material stock.

The Proposed Forecasting System for the

Procurement of Raw Materials

This raw material ordering system model consists of

one main menu where every user who has logged in

can enter the system according to their authority. The

Order menu consists of one order form. This form can

only be seen by the marketing department which has

the task of entering customer order data into the

system. The Raw Material Menu is a menu consisting

of five forms. This menu can be accessed by the raw

material warehouse and purchasing department. But

not all forms can be accessed, but only a few forms.

The forecasting calculation form can only be accessed

by the head of the raw material warehouse, where the

head of the warehouse has the task of forecasting raw

material needs for the next few periods based on

historical data on the use of material

The calculation menu consists of two forms:

a) Forecast Calculation Form

The Forecast Calculation form in Fig. 3, can be

accessed by the Head of raw material Warehouse,

where the task of forecasting raw material

requirements for the next few periods based on

historical data on the use of raw materials.

The Applications of Ordering Materials using Time Series Forecasting with CB-Predictor

1319

Figure 3: Forecast Calculation Form.

The purchasing department also plays an important

role where a purchasing manager must calculate the

correct quantity of raw material orders. An incorrect,

excessive or insufficient order can cause losses to the

company. Of course, every company always wants

multiple profits. Therefore, the determination of the

order quantity should be calculated using several

methods, and the method that produces the lowest

total cost should be chosen.

b) Order Calculation Form

This Order Calculation form in Fig. 4 can accessed by

Head of Department of Purchasing, where the is

tasked with calculating the most economical quantity

of raw material orders, by comparing several

calculation methods.

Figure 4: Order Calculation Form.

4 CONCLUSION

The system is designed to have a purpose and

function for ordering raw materials, starting to record

the incoming or outgoing raw materials, calculating

the forecast, calculating the order quantity, then

making a purchase note for each type of raw material

with the quantity according to the calculation.

The warehouse section enters incoming and

outgoing data, calculates the use of raw materials

each month, calculates forecasting raw material needs

for the future period, and changes the status of the

purchase note if the ordered raw materials have been

sent by the supplier. With the raw material ordering

system, it will increase the accuracy of raw material

stock data.

The system facilitates the work of the raw

material warehouse section, where when the order is

received the system has already carried out a

breakdown of the needs for each type of raw material

for the order, so that when the warehouse department

clicks the order number, the details of the raw

materials are immediately displayed, and the raw

materials can be immediately prepared. Likewise,

when the raw materials are received, the warehouse

department only needs to look at the purchase notes

ordered whether they match the raw materials

received. Finally, the system will have a positive

impact in terms of controlling raw materials,

maintaining raw material stocks so that they can

always be monitored and controlled.

ACKNOWLEDGEMENTS

This work is supported by Research and Technology

Transfer Office, Bina Nusantara University as a part

of Bina Nusantara University’s International

Research Grant entitled Supply Chain Optimization

Using E-Commerce with contract number:

No.026/VR.RTT/IV/2020 and contract date: 6 April

2020

REFERENCES

Ainuri, M., and A. D. Guritno (2017), “Optimasi Model

Persediaan Bahan Baku (Raw Material Inventory

Model) Pada Industri Kecil Kelompok Pangan Produk

Unggulan di Daerah Istimewa Yogyakarta,” agriTECH,

vol. 16, no. 4, pp. 23–29, Feb. 2017, doi:

10.22146/AGRITECH.19318.

M. Ainuri and A. D. Guritno, “Optimasi Model Persediaan

Bahan Baku (Raw Material Inventory Model) Pada

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

1320

Industri Kecil Kelompok Pangan Produk Unggulan di

Daerah Istimewa Yogyakarta,” agriTECH, vol. 16, no.

4, pp. 23–29, Feb. 2017, doi: 10.22146/A

GRITECH.19318.

Bahtera, N.T. (2017), “The Indonesian Journal of Applied

Business,” Indones. J. Appl. Bus., vol. 1, no. 1, pp. 14–

29, 2017

Chen, X.F., L. H. Cen, and Y. G. Cen (2009), “A raw

material purchasing decision-making model of order

allocation with multiple suppliers and its optimization,”

in 2009 1st International Conference on Information

Science and Engineering, ICISE 2009, 2009, pp. 66–69,

doi: 10.1109/ICISE.2009.143.

Citra, P.T., A. Bosco, G. Sofiyanurriyanti, and M.

Syarifuddin (2020), “Analysis of Raw Material

Inventory Control for a Minimum Total Cost Method

with EOQ (Economy Order Quantity) in,” Int. J. Sci.

Eng. Inf. Technol., vol. 02, no. 02, pp. 72–77, Jul. 2018,

Accessed: Sep. 03, 2020. [Online]. Available:

https://journal.trunojoyo.ac.id/ijseit/article/view/6487.

Dekker, R,. and E. Smeitink (1991), “Opportunity-based

block replacement,” Eur. J. Oper. Res., vol. 53, no. 1,

pp. 46–63, Jul. 1991, doi: 10.1016/0377-2217

(91)90092-A.

Sukmawati, F., T. E. Putri, and I. Yuniar (2019), “Raw

Material Ordering Control Application Using

Economic Order Quantity (EOQ) and Reorder Point

(ROP) Methods for Shoe Company in Indonesia,” Glob.

Bus. Manag. Res. An Int. J., vol. 11, no. SI 1, p. 152, Jan.

2019, Accessed: Sep. 03, 2020. [Online]. Available:

https://www.questia.com/read/1G1-613798900/raw-mat

erial-ordering-control-application-using-economic

Glasser, G.J. (1969), “Planned Replacement: Some Theory

and its Application,” J. Qual. Technol., vol. 1, no. 2, pp.

110–119, Apr. 1969, doi: 10.1080/00224065.1969.11

980359.

Goldman, L.I. (2002), “Crystal ball professional

introductory tutorial,” in Winter Simulation Conference

Proceedings, 2002, vol. 2, pp. 1539–1545, doi:

10.1109/wsc.2002.1166430.

Irmayanti, H. (2019), “Analysis of Raw Material Ordering

with Economic Order Quantity Method,” in IOP

Conference Series: Materials Science and Engineering,

Nov. 2019, vol. 662, no. 3, doi: 10.1088/1757-899X/

662/3/032011.

Mart, T., S. Duran, and I. S. Bakal (2013), “Tactical

inventory and backorder decisions for systems with

predictable production yield,” in International Journal

of Production Economics, Jun. 2013, vol. 143, no. 2, pp.

294–303, doi: 10.1016/j.ijpe.2012.01.029.

Montaño Moreno, J.J., A. Palmer Pol, A. Sesé Abad, and B.

Cajal Blasco (2013), “El índice R-MAPE como medida

resistente del ajuste en la previsiońn,” Psicothema, vol.

25, no. 4, pp. 500–506, 2013, doi: 10.7334

/psicothema2013.23.

Nadella, P., A. Swaminathan, and S. V. Subramanian

(2020), “Forecasting efforts from prior epidemics and

COVID-19 predictions,” Eur. J. Epidemiol., vol. 1, p.

3, Jul. 2020, doi: 10.1007/s10654-020-00661-0.

Salais-Fierro, T.E., J. A. Saucedo-Martinez, R. Rodriguez-

Aguilar, and J. M. Vela-Haro (2020), “Demand

Prediction Using a Soft-Computing Approach: A Case

Study of Automotive Industry,” Appl. Sci., vol. 10, no.

3, p. 829, Jan. 2020, doi: 10.3390/app10030829.

Undersander, M.A., R. M. Kettler, and M. Stains(2017),

“Exploring the item order effect in a geoscience concept

inventory,” J. Geosci. Educ., vol. 65, no. 3, pp. 292–

303, 2017, doi: 10.5408/16-235.1.

Voulgaris, C.T., (2019), “Crystal Balls and Black Boxes:

What Makes a Good Forecast?,” J. Plan. Lit., vol. 34,

no. 3, pp. 286–299, Aug. 2019, doi: 10.1177/

0885412219838495.

The Applications of Ordering Materials using Time Series Forecasting with CB-Predictor

1321