IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Inventory management concepts and techniques

To cite this article: G Priniotakis and P Argyropoulos 2018 IOP Conf. Ser.: Mater. Sci. Eng. 459 012060

View the article online for updates and enhancements.

This content was downloaded from IP address 178.171.60.241 on 07/12/2018 at 17:12

Aegean International Textile and Advanced Engineering Conference (AITAE 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 459 (2019) 012060 doi:10.1088/1757-899X/459/1/012060

Inventory management concepts and techniques

G Priniotakis1 and P Argyropoulos2
1
Department of Industrial Design and Production Engineering,
School of Engineering - University of West Attica, Athens, Greece
2
Prosent S.A.,Technical Fabrics, Production Planning and Control (PPC)Department,
84 Archimidous str., 19400, Athens, Greece

[email protected]

Abstract: Inventory management has become one of the key elements of the supply chain
management and can greatly affect the performance of a business. The textile industry is no
exception. Traditional approaches in decision making based on manager instincts and hunches
are no longer enough in the today’s increasingly competitive environment. Small to medium
sized family owned textile businesses are usually prone to this way of thinking.
This paper discusses some basic concepts and techniques for classifying inventory, controlling
inventory levels, avoiding stock outs and increasing customer satisfaction. It also discusses the
importance of forecasting demand and uses the Root Mean Square Error (RMSE) as an
effective measure of the forecast error, which later becomes a basic driver for inventory
management. It addresses the Service Level (SL) as a performance metric and emphasizes on
the importance of Safety Stock (SS). Finally, it discusses the use of the Reorder Point (ROP) as
an efficient indicator for triggering production replenishment and proposes a simple technique
for prioritizing production orders.

1. Introduction
Inventory management is the process of monitoring and controlling inventory level and ensuring
adequate replenishment in order to meet customer demand. Determining the appropriate inventory
level is crucial since inventory ties up money and affects performance. Having too much inventory
reduces the working capital and impacts the company’s liquidity. On the contrary, having too little
inventory leads to stock outs and missed sales which leads to less profit. It becomes clear that
management attention should be focused on keeping inventory level somewhere in between, striving
for increased customer satisfaction and minimum stock outs while keeping inventory costs as low as
possible.

2. Service Level
The Service Level (SL) is an important performance indicator which in a simplified manner, measures
a company’s ability to service customer demand and is expressed as a percentage. In inventory
management, service level is the probability that the customer demand is met or that the customer
demand does not exceed the inventory. A service level of 95% means that there is 95% probability that
demand will be met and customer orders will be fulfilled on time, while the probability that a stock out
will occur, resulting in missed sales, is 5%.The higher the Service Level, the higher the customer
satisfaction but also the higher the inventory level. Since the future demand is uncertain, achieving a
100% Service Level would require an infinite amount of inventory which is clearly unachievable.
Management should understand the trade-off between the cost of inventory and the cost of stock-outs
and position against inventory levels based on specific criteria.

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
Aegean International Textile and Advanced Engineering Conference (AITAE 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 459 (2019) 012060 doi:10.1088/1757-899X/459/1/012060

3. Inventory Classification
ABC inventory classification is a very popular inventory control technique that follows the Pareto
Principle which states that, for many events, roughly 80% of the effects come from 20% of the causes.
In a case of a business, it could be stated that roughly the 20% of the end products generates the 80%
of the income. In ABC analysis, a company reviews its inventory and sorts all items into three
categories, called "A" items, "B" items and "C" items. A typical breakdown would possibly describe
“A” items as those that produce 70% of income, “B” items as those that produce 25% of income and
“C” items, as those that produce 5% of income. This classification might be different from company to
company but managers should be able to find the pattern that suits best their needs.
Clearly, “A” items require closer attention and should be handled differently. Assigning higher
Service Levels for those items is a wise choice. The higher service level will lead to higher inventory
but will also decrease the probability of a stock out. A 5% probability of a stock out for an “A” item
will result to much higher losses than of a 5% probability for a “C” item. A 99% Service Level for an
“A” item could have about the same impact as an 85% Service Level for a “C” item and managers
should position against each category accordingly.

4. Forecast Error and Safety Stock

It has been already stated, that customer demand is uncertain. Managers should try to predict future
demand based on statistical data and taking into account multiple criteria. It is highly desirable to try
to predict the future demand and get properly prepared even with a certain degree of uncertainty than
having no expectation of what is about to happen. The methods and tools used for forecasting are not
under the scope of this paper. What is really important, is to find a way to calculate how closely the
prediction of the demand meets the actual demand, thus how accurate a forecast is. The difference
between the actual and the forecasted data is the Forecast Error.
Since the Forecast Error can be calculated and not just predicted, it can be quite a safe driver for
inventory management. The goal is to keep just as much inventory as it is really needed and by
knowing that there is a fixed error in our estimation, we can safely take it into account and add up a bit
of extra stock to our inventory to compensate for this misalignment. This is called Safety Stock and its
purpose of existence is to absorb the error of the estimation and to protect the company against an
unexpected and unwanted stock out. In a simplified scenario, a 30% variation of the forecasted
demand should result in a 30% increase of the inventory level.

5. Root Mean Square Error (RMSE)

In order to calculate the errorof a forecast the RMSE method is preferred among others, as it calculates
the standard deviation of the residuals between the actual and the forecasted data. As the name
suggests, the difference between the forecasted demand and the actual demand is squared and then it is
expressed as the square root of the average squared residuals. In the table below, real data are
presented.

Table 1. Calculation of the RMSE of the Forecasted Demand for SKU X

Total Average
Actual Forecasted Squared
Month Squared Squared RMSE
Demand Demand Error
Error Error
1 1.508 1.533 625 1.658.065 138.172 372
2 1.884 1.867 289
3 1.024 1.234 44.100
4 1.458 1.523 4.225
5 2.433 2.763 108.900
6 3.523 2.707 665.856
7 2.322 1.998 104.976

2
Aegean International Textile and Advanced Engineering Conference (AITAE 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 459 (2019) 012060 doi:10.1088/1757-899X/459/1/012060

8 818 444 139.876

9 1.753 1.015 544.644
10 1.889 1.767 14.884
11 1.438 1.509 5.041
12 1.283 1.440 24.649

The reader will quickly realize that the forecasted data are not highly accurate. Although forecast
accuracy is important, it will soon become clear that calculating the forecast error and taking itinto
account when positioning against inventory levels is more significant than the accuracy of the forecast
itself, when our primary concern is to protect against stock outs.

6. Mixing Service Level, Safety Stock and RMSE

When the desired Service Level for a stock keeping unit (SKU) is selected and the RMSE of the
forecast has been defined then the Safety Stock can be accurately calculated. The SKU in table 1 was
classified as an “A” item and the Service Level was set to 97%. Under the assumption that our data are
following a normal distribution curve, then in 97% of the cases, the actual demand will be less than
1,88 standard deviations above the mean demand (see appendix A).
Any sample value will lie within μ±σ×Z. Since we are looking only for extra stock we actually
calculate only for μ+σ× Z. In addition, we know our estimation’s variance but we need to calculate the
variance during the lead time. The safety stock needed during the manufacturing lead time can be
calculated using the following formula:

= × √ ×

Where:
Z =Z-score for the desired Service Level (see appendix A)
LT = The Lead Time using in the same time period as in the forecast
RMSE = The Root Mean Square Error between the actual and the forecasted demand

Table 2. Parameters for SKU X

RMSE 372
Service Level 97%
Working Days per Month 22
Manufacturing Lead Time (Days) 5
Z Statistic 1,88

Given the data from table 2, the Safety Stock for SKU X can be calculated as below:

5
= 1,88 × × 372 = 333
22

Thus, keeping 333 units of extra stock for SKU X, will prevent a stock out for 97% of the time.

One could note that in the textile industry, the manufacturing lead time is also been subjected to a
certain variance. This can be due to the irregular machine cycle time or due to an unexpected
downtime of the machinery. For instance, the setup time of a stenter machine is greatly affected by the
process it previously underwent resulting in extra time for heating it up or cooling it down. It is wise to

3
Aegean International Textile and Advanced Engineering Conference (AITAE 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 459 (2019) 012060 doi:10.1088/1757-899X/459/1/012060

take this variance under consideration and adjust the lead time in the equation accordingly. If the
variance of the demand and the variance of the lead time are independent of each other,then the safety
stock is Z-score times the square root of the sum of the squares of the individual variabilities, as in the
following equation:

= × ( × ) +( × )

But if both the variance of the demand and the variance of the lead time are dependent of each
other then safety stock can be computed as below:

= × √ × + × ×

Where:
Z = Z-score for the desired Service Level (see appendix A)
LT = The Lead Time using in the same time period as in the forecast
RMSED = The Root Mean Square Error between the actual and the forecasted demand
RMSELT=The Root Mean Square Error between the actual and the standardlead time
= The average forecasted demand

Table 3.Parameters for SKU X

RMSED 372
Service Level 97%
Working Days per Month 22
Manufacturing Lead Time (Days) 5
Z Statistic 1,88
RMSELT (Days) 0,25 (6 hrs)
Average Forecasted Demand (monthly) 1650

Given the data from table 3, the safety stock can be calculated. It was decided that the variance of
the lead time and the variance of the demand was dependent of each other. This is quite typical in
textile factories since the life expectancy of the industrial machinery is very high (many decades) and
as a result, new machinery with higher output rate is added before old machinery is regarded obsolete
and its use is been discontinued. Thus, when demand rises and extra capacity is needed, managers
decide to add shifts and make use of old and new machines simultaneously which produce the same
products at different output rates resulting in higher variance in the manufacturing lead time. Taking
this into account, the Safety Stock is calculated below:

5 0,25
= 1,88 × × 372 + 1,88 × × 1650 = 369
22 22

Thus, keeping 369 units of extra stock for SKU X, will prevent a stock out for 97% of the time.

7. The Re-Order Point

The Re-Order Point (ROP) technique is a process of triggering inventory replenishment based on
inventory level. The ROP is set at a level which is calculated based on the forecasted demand during
the replenishment time plus a safety stock, as in the following equation:

4
Aegean International Textile and Advanced Engineering Conference (AITAE 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 459 (2019) 012060 doi:10.1088/1757-899X/459/1/012060

= × +
Where:
= The average forecasted demand
= The average Lead Time
SS = The Safety Stock

Keeping the same data as before, the ROP for SKU X is calculated below:

1650
= × 5 + 369 = 744
22

Thus, the inventory level of SKU X should be closely monitored and as soon as it becomes equal or
less than 744 units, then production should be reinitiated. There should be just enough time to
reproduce the item before its inventory level reaches the safety stock level but even if it does there will
be 97% chance that the safety stock will be enough to protect against an unexpected customer demand
or a production delay and only 3% chance that the inventory level will fall to zero and a sales order
will not be satisfied on time.
There will also be times when the accumulated pending orders will exceed the inventory level even
before the ROP is reached. Thus, the future fulfilment of the pending orders will probably lead to a
stock out. Clearly, the Projected Inventory level is of concern and it is calculated according to the
formula bellow:

= + −

Where:
Stock On Hand = The available inventory
WIP = The inventory that is already in process but not yet finished (Work in Process)
Pending Orders = The pending orders that are due today or already past due

Based on the above approach, production should be initiated whenever the Projected Inventory
level falls below the ROP.

8. Prioritizing Production Orders

In real life, the inventory or the projected inventory level of multiple stock keeping units will
simultaneously reach or fall under their ROP. There are also capacity constraints or other constraints
that force managers to perform some kind of prioritization of the production. When avoiding a stock
out is the primary concern, then some of the key elements for deciding how to prioritize production
are: the ABC classification, the Inventory Turnover Ratio and the Projected Inventory Level.
The Inventory Turnover ratio is metric that calculates the number of times that inventory cycles
(turns) during a given period. It is calculated as the Cost Of Goods sold over the average inventory but
can also be expressed as the number of units sold over the average number of units. The importance of
the ABC classification and the Projected Inventory Level has already been discussed.
The following excel sheet depicts a real life situation:

Table 4.Excel Inventory Report

Pending ABC Inv. Projected
Code WIP Stock SS ROP Status
Orders Rating Turnover Stock
SKU2463 260 102 194 756 C 4 <0 
SKU2794 296 88 167 32 C 4 >ROP 
SKU2854 300 75 143 280 C 2 <=SS 

5
Aegean International Textile and Advanced Engineering Conference (AITAE 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 459 (2019) 012060 doi:10.1088/1757-899X/459/1/012060

Pending ABC Inv. Projected

Code WIP Stock SS ROP Status
Orders Rating Turnover Stock
SKU1282 910 370 681 51 B 6 >ROP 
SKU1362 600 800 502 871 900 A 7 <=SS 
SKU2341 778 88 167 350 C 2 >ROP 
SKU2361 1.001 292 537 450 B 5 >ROP 
SKU2461 600 1.082 419 725 960 A 8 <=ROP 
SKU2791 1.618 923 1.600 253 A 12 <=ROP 
SKU2851 780 641 1.111 858 A 7 <0 

The Status becomes green and signifies the need for production when the projected inventory level
falls below the ROP. SKU2851 & 2463 have not enough stock to fulfil the pending orders thus their
production should be prioritized. SKU1362 & 2854 will soon fall under their safety stock level thus
they should be produced right after. SKU2461 & 2791 will soon fall under their ROP but SKU2791
cycles a bit faster than 2461 thus it would be wiser to produce it first.
The idea behind it is to prioritize those items that their projected stock falls below zero. If more
than one item meets this rule then prioritize based on their ABC classification. If again more than one
item meets this rule then prioritize based on their turnover ratio. Then check for all the items that their
projected inventory will fall under their safety stock level prioritizing them the same way and finally
check for those items that their projected inventory level that will fall under their ROP.

9. Conclusions
This paper emphasizes on the importance of inventory management. The methods and tools that are
used can add great business value. Holding the right amount of inventory can increase business
performance by reducing the response time to customer demand which results to higher customer
satisfaction. The textile managers and business owners should weigh and balance the trade-offs when
deciding how much inventory to hold and strategically decide based on the proven concepts and
techniques that are described.

6
Aegean International Textile and Advanced Engineering Conference (AITAE 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 459 (2019) 012060 doi:10.1088/1757-899X/459/1/012060

Appendix A

7
Aegean International Textile and Advanced Engineering Conference (AITAE 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 459 (2019) 012060 doi:10.1088/1757-899X/459/1/012060

References
[1] Plossl G, 1994, Orlicky's Material Requirements Planning
[2] King L P (2011): Crack the code: Understanding safety stock and mastering its equations,
APICS Magazine, July/August 2011, p. 33-36
[3] Chockalingam M, 2009, Forecast Accuracy and Inventory Strategies
[4] Curwin J and Slater R, 2002, Quantitative Methods for Business Decisions
[5] Slack N, Chambers S and Johnston R, 2004, Operations Management