AUTOMATIC & UNMANNED STOCK REPLENISHMENT

PROCESS USING SCALES FOR MONITORING

Ari Happonen and Erno Salmela

Department of Industrial Engineering and Management, Lappeenranta University of Technology, Skinnarilankatu 34

Lappeenranta, Finland

Keywords: Unmanned stock replenishment, automatic-replenishment, vendor managed inventory.

Abstract: This paper presents a process description for a fully integrated, unmanned warehouse inventory

replenishment system. This description serves as a springboard for discussion on the TEMO project

(TEMO-Project, 2006), where the Vendor Managed Inventory (VMI) model’s current situation and future

possibilities in the Finnish engineering industry are being studied. In this paper we concentrate exclusively

on C-class product management challenges. Because of the exceptional characteristics of these products, C-

class product monitoring cannot be directly adapted to the VMI models used by the food industry, which are

usually based on Point of Sale (POS) information. By presenting the “ideal” model, as well as an adapted

version which can be carried out using currently available technology, the researchers would like to open

scientific discussion on the subject, especially considering challenges in stock monitoring. The main idea is

to offer a higher level of service to customers as well as new business opportunities to service providers.

This kind of technology implementation would bring new challenges to supply chain management but also

new areas for research, such as information management, communications engineering, as well as e-

commerce and added value services, to the forefront.

1 INTRODUCTION

The Finnish engineering industry uses a wide scale

of different products classified as C-class products.

These products are often small in size, lightweight

and take up little space (e.g. nuts and bolts). On the

other hand, heavy industry uses some mid-sized and

heavy products in its processes (e.g. heavy copper

cable and axels). Furthermore, heavy industry also

makes use of large but lightweight products, such as

foam insulation, in their assembly. The selection of

C-class products used in the engineering industry is

heterogeneous and the category of products is

noticeably large.

Monitoring of C-class products is a particularly

challenging financial and administrative exercise.

Monitoring these as A- or B-products are monitored

isn’t financially feasible because amount of human

labour and the items. To the companies involved in

our study, as well as for their customers, balances of

these products’ are almost without exception not

known. Companies are classified as large and

medium-sized on the Finnish industry scale by

(Federation of Finnish Enterprises).

A similar management problem was present in

the previously done SERVIISI project (SERVIISI-

Project, 2005), where the product cost varied from

cent to tens of euros and unit amounts varied

between handfuls to tens of thousands. Researchers

asked reasons why the product balances were not

kept on IT systems and why companies didn’t use

technology to manage the whole life-cycle of the

products. The most common reason given was the

unit level monitoring of products would be such an

administrative burden that it simply isn’t feasible

task to do (products’ heterogeneous nature and

number of items was cited time and time again).

From a logistics point of view, these C-class

products are a challenge for the Finnish engineering

industry. Monitoring C-class products is a necessary

evil; it takes up a lot of time and it’s not the core

function of the companies. Product monitoring is

also challenging transportation-wise. In Finland,

distances are long and the land itself is divided by

lake districts along the North-South axis. Also, the

load sizes required to make road transportation

financially viable is a challenge itself. In addition,

labour costs are rather high in our country when

157

Happonen A. and Salmela E. (2007).

AUTOMATIC & UNMANNED STOCK REPLENISHMENT PROCESS USING SCALES FOR MONITORING.

In Proceedings of the Third International Conference on Web Information Systems and Technologies - Society, e-Business and e-Government /

e-Learning, pages 157-162

DOI: 10.5220/0001282801570162

 SciTePress

compared to other countries. Because of the reasons

mentioned above, companies are outsourcing C-

class product inventory replenishment process to

companies which offer VMI model-based services.

According to recent studies, metal and

engineering industry future trends are looking good,

which means that logistical challenges will continue

to grow in number (Statistics Finland, Teollisuuden

toimialakatsaus I/2006). At the same time, costs

connected to transportation operations have raised

faster than turnover (Statistics Finland,

Palvelualojen toimialakatsaus IV/2005). In practice,

this has led the engineering industry in Finland to

outsource their C-class products’ acquisition,

management, delivery and stock replenishment

activities to VMI operators. Based on our interviews,

we believe that this is one of the reasons why

logistics operators have expanded their operations

by 10 to 30% yearly over the past few years.

Current VMI stock replenishment model is

accomplished mainly through manual labour. This

project’s one goal is to find a technological solution,

through which operations efficiency is enhanced,

service can be offered to a larger customer base and

current customer base service level is upgraded.

Through interviews and fieldwork, the researchers

have studied Finnish VMI operations and have

found out the basic problems present in current

operation models. The researchers also considered

the possibilities of mobile technology. As a result an

idealistic technological solution’s framework is

presented as well as an adapted concept of this ideal

model, which could be implemented using current

technology. The researchers would like to open the

discussion on the automated management of C-class

products in the scientific world to bring a strong

theoretical viewpoint as a part of this research.

The research area is limited to the VMI models

customer-supplier relationship. During the project, a

practical prototype based on our concept will be

built in order to find the practical limitations which

usually go unnoticed when models are studied only

theoretically. In the next section, the aims of the

study, the research process and the methods are

covered, as well as the criteria for selecting the

technology connected to the model’s practical

implementation. Section three discusses the current

situation of VMI within the Finnish engineering

industry. Section four handles the fully integrated

technical solution’s process description and in

section five, a potential way to practically

implement the model is described. Practical

limitations and the potential uses for the model are

reflected upon in section six. Finally, the topics and

ideas for further research are presented.

2 METHODOLOGY

The study was carried out as a quantitative and

exploratory situation study. Research methods

included partially structured, face-to-face interviews

as well as workshop-style brainstorming in groups.

Methods were chosen to gain an in-depth knowledge

about the research subjects and also to make sure

that all of the interviewees and researchers

understood the concepts being handled in the same

way. The groups were made up of the people

representing the companies in this project. Both the

heads of the business units as well as operative-level

employees were involved in our studies. Among the

companies were customers, suppliers, wholesalers as

well as two logistics companies.

During the first round of interviews the problems

were limited to those that could be easily improved

with automation. After this, one of the researchers

did a field study. He studied the replenishment

process at the shelf level in 3 companies VMI cases.

The fieldwork supported the findings which were

obtained from the first round of interviews and it

familiarized the researchers with practical operation.

Also during this field work the researcher found out

some limitations of the technology-based solution

because of manual work. During the second round of

workshop-style meetings, the observations based on

first interview round were tested and the problems

were investigated further. There were numerous

stock replenishment processes being implemented in

slightly different ways among the companies.

In this study, customer means companies which

produce different machines for end users in industry

on an assembly and project basis. The VMI suppliers

are mainly technical wholesalers. Therefore, this

research concentrates on the cooperation between

customer and the wholesaler (or other material

supplier) in the Finnish industrial environment.

The abovementioned companies will be used as

the basis for reflection upon the challenges

connected with bringing technical solutions to VMI

operations in Finland. In Finnish VMI operations,

the largest single cost comes from using manual

labour (SERVIISI-Project, 2005). In Finland,

manual labour is particularly expensive; the fixed

costs related with labour are extensive and each new

pair of hands means more costs, for both

administration and production. In this study, the

expense of manual labour was the impetus for

WEBIST 2007 - International Conference on Web Information Systems and Technologies

158

finding technical solutions that would make current

VMI model replenishment processes more efficient.

Through technology the researchers try to find long-

term benefits for many different parts of the process

at once. For example: minimalizing manual labour,

making follow-up and prediction more efficient and

improving the management of exception situations.

In practice the management of exceptions leads to

manual labour reduction because each exception

needs the work of at least two people (one on the

customer’s end and one on the supplier’s end).

The goal is to create a fully integrated VMI

operational model including communication,

warehouse management and supply delivery

improvements trough automation. From this fully

integrated model the researchers will design a model

which can be implemented using technology that is

currently available. In essence, this is cross-road of

two objectives – the first is the streamlining of the

process and the second is new operational models

which are built on the efficient combination of

information handling and network technology.

The study began by the researchers familiarizing

themselves with the previously done SERVIISI

project (SERVIISI-Project, 2005). SERVIISI was a

study on VMI models on supplier-customer

relationships. Next the researchers continued by

going through the food industry’s VMI models,

where technology was used to minimalize stock

replenishment control and the management burden

(Pohlen and Goldsby, 2003), (Katz et al., 2000).

The researchers realized that in practice the solutions

used in the food industry won’t directly fit into the

engineering industry as is. Storage times are from

days to months in the food industry, but they are

from months to years in engineering. The food

industry’s solutions are based on strict inventory

accounting and Point of Sale (POS) information

(Elvander, 2005). In engineering, the materials have

an unknown balance. The materials are located on

shelves in workshops, where anyone can come and

take whatever they need whenever they need it

without book keeping. The differences are so large

in practice that the models used in the food industry

are not easily transformable to the VMI environment

present in engineering industry.

Our research tries to determine whether or not

technical management is viable solution. If it is,

considering the practical limitations, will the added

value provided by technology be enough to justify

studying new operational models in the future?

3 THE VMI MODEL’S CURRENT

SITUATION IN THE FINNISH

ENGINEERING INDUSTRY

This section will outline the typical VMI model used

in the Finnish customer-supplier relationship. The

description of the operational model focuses on

normal functions and will not be concerned with, for

example, the ”learning curve” that occurs at the

beginning of implementation. This description will

begin with the initial visit to the shelf by the stock

replenisher and will conclude to the next visit.

The stock replenisher arrives at the shelf with the

supplies from the previous order. He shelves the

products and starts the replenishment order process.

The inspection process is based on visual inspection.

Each product is assigned with a refill limit and order

point which is in direct relationship with the speed

of consumption and the time between replenishment

rounds. Ordering of individual products happens

seldom in relation to the time between replenishment

rounds. The stock replenisher checks which products

are below the minimal stock limit. Most typical

replenishment order process methods are the

following: off-line or on-line handheld computer for

“instant” order, or pen & paper method with later on

updates to order system. If the frequency of

replenishment round is faster than the supply

delivery round, it is possible that the items have

numerous overlapping orders. Usually the stock

replenisher has thousands of products to worry about

so there is no way that he can remember the unfilled

orders outright. In this case, human error can easily

result in overstocking or non-filling of the order

because the stock replenisher remembers that he has

already made an order. There are many ways to

resolve this problem but typically solutions involve

the management person who checks all orders before

the refill order activation in the order system.

After order information is fed into the order

system, new supply collection process can begin at

the facilities of the supplier. An order-based

collection list is printed out for the collection person.

The collected items then end up in shipping, where

the items are inspected, signed off and cleared for

delivery. The products end up with the stock

replenisher on the customer’s shelves and the whole

process begins again. Billing information is usually

generated during the shipping stage. The supplier

provides a cost report as well as an itemized invoice.

At this point it is important to note that no one is

aware of the real consumption, just the delivery

AUTOMATIC & UNMANNED STOCK REPLENISHMENT PROCESS USING SCALES FOR MONITORING

159

rates. This means that an in-depth analysis on the

item level consumption is an impossible task to do.

4 ICT-BASED SOLUTION FOR A

FULLY-INTEGRATED VMI

MODEL

The idea behind this system is automated

monitoring, order parameter setting, automated order

generation based on additional parameters as well as

manpowered order collection and shelving.

A fully integrated stock replenishment process is

based on exploiting the maximum benefits from

technology and minimizing the involvement of

human labour. The main problem with the current

nowadays VMI model is that shelf rounds have to be

made according to a timetable. It is possible for the

stock replenisher to go through several shelves, or

even through several customers, without making a

single refill order. The stock replenisher has to make

his rounds at appropriate intervals so that an

understocking situation does not occur. For this

reason, the supplier is forced to play it safe, both

with stocks on the customer’s shelves and with the

time between rounds (a noticeably tighter timetable

than what might actually be necessary on average).

For this problem we propose usage of dynamically

changed parameters. The proposed dynamic setting

of parameters in IT based solution could be achieved

in the way described in paper (Disney and Towill,

2002). Disney and Towill present a scheduling

algorithm Automatic Pipeline, Inventory and Order

Based Production Control System (APIOBPCS) to

be used within VMI operations supply chain.

In this model, all of the customer’s products are

monitored trough computer system. Monitoring

system can be based on for example, micro switch,

Radio Frequency Identification (RFID), video

surveillance, measuring technology, etc. which

provides measuring precision that is equal to or

better than visual inspection. Technical solution will

allow item checks without doing it manually on site.

Product’s inventory is then continuously monitored,

the information is analysed, cleaned up and used to

make refill decisions fully automatically. The refill

decision is based on the customer’s confirmed order

base, predicted sales, production plans, product

make-up information, etc., as well as real

consumption information received through

monitoring. By combining all of this information,

the supplier can make an optimal refill order. In this

case, optimal means the kind of refill that minimizes

the amount of manual labour but at no time risks the

provided service level. The main principle behind

this is that when ordering, the order is as large as

possible at once and the period between orders is

maximized, in a ”reasonable” manner.

When the order is placed on system, it is transfers

automatically to the supplier’s system, which then

prepares to print out a collection list. Before this list

is printed, system checks for changes at the

monitored sites and updates the list accordingly.

The supplier collects the products, checks the order

and clears the shipment for delivery.

A site-specific route can be optimized so that the

stock replenisher drives a geographically optimized

route but still in a way which prioritizes the route

according to the calculated rate of product

consumption. The idea is to minimize the travelling

costs without jeopardizing service level. This way

the time used to travel (or total travelling cost

depending which is preferred) is minimized but there

should not be any shortages at the serviced site. The

stock replenisher’s product delivery times are

entered into the system, the current balance of items

is updated and the company is invoiced either on a

per order basis or later on by monthly invoice. The

system checks automatically that the customer’s

invoice corresponds to replenishments.

Customer gets delivery time, consumption rate

and exception reports automatically. In theory, this

type of a system would guarantee that the product

will never be out of stock because the supply system

continuously monitors product consumption.

Particularly large consumption can be dealt with on

an emergency basis, instead of having the situation

exposed only after the customer has informed the

supplier or discovering it too late during regular

replenishing rounds.

By breaking down the ideal operational model

into parts, one can very quickly see the

“impossibility” of it because of the limitations set by

the technology. A fully integrated model would not,

in practice, need a user interface for managing

information but an interface on the system is needed

for example to manage exceptions etc. In the TEMO

project, a prototype will be built, which is based on

practical implementation and which the researchers

will be able to use to further explore the problem

areas related to information management and

communication technology in VMI context.

WEBIST 2007 - International Conference on Web Information Systems and Technologies

160

5 A SUGGESTION FOR

IMPLEMENTATION USING

CURRENT TECHNOLOGY

The prototype will be built upon an implementation

model using current technology. Some of the

following technological choices, for monitoring,

were examined and considered: camera surveillance,

RFID, weight measurement, sliding potentiometers

and switches, etc. The characteristics of these

technologies were checked and suppliers and

customers were interviewed about the practical

limitations connected to these technologies.

After weighing up the needs of the suppliers and

customers, the researchers decided to base the

prototype solution on scale technology. Measuring

the items weight offered the best overall

compromise based on the following criteria:

− real time inventory balance information

(with enough precision)

− no manual labour for data entry required

− ease of historical data collection

− ability to analyse consumption trends, such

as top and bottom 100

− automatic order replenishment

− dynamic parameter setting (order limit, refill

limit)

− real time emergency situation alarms

− delivery time monitoring

− simple technology from a functional

standpoint

− solutions response to change

Next, the imaginary progress of the

replenishment process achieved through scale

technology is described. As before, this cycle will

begin on the initial replenishment visit and it will

end on the following visit.

The stock replenisher arrives at the shelves with

the previous order’s delivery and proceeds with the

shelving. At the same time he visually inspects

measuring apparatus for malfunctions. When the

replenisher finishes the replenishment process he

moves on to the next replenishment point. Refill

orders are made by the monitoring system. The

system inspects the inventory balance on the shelves

and sends an order as soon as one of the products

drops below the order point. Products are not

ordered individually. Actually, as soon as one

product drops below its order point, all of the

products near their order point are also ordered. This

process should decrease the number of orders.

The refill order is transferred as an order

proposal to the supplier’s system. The proposal is

analyzed and items are classified into product groups

such as even consumption rate, easily forecasted and

unsure. The person handling the order (order

manager) adjusts the order based on experience and

the prior information received from the customer,

and confirms the order. The system is aware of

unfilled orders and does not make overlapping

orders. At least the system informs the order

manager and he can then check the overlapping

orders in order to avoid overfilling the shelves.

As the refill order is done in this manner,

shelves have to bee visited only if there is

replenishment work to be done. The automated

system updates order point and replenishment

parameters automatically (taking space restrictions

into account) and in this way, reduces again need to

visit at the shelves. Built in this way, the system is

simple enough so that all of its users can understand

the principle of the system. In addition, the system

can easily adapt and be adapted to changes.

After confirming the order, the products are

gathered on the supplier’s end according to current

practice and the stock replenisher delivers the goods,

stocks them on the shelves and the process begins

from the beginning once again.

6 DISCUSSION AND

CONCLUSIONS

In discussing the presented implementation solution

during the second round of interviews, a very

positive interest in this project was raised but at the

same time, the practical problems were seen as a

large challenge. For example, how service provider

can handle exceptions like Internet communication

problems. There were some concerns about system

robustness against changes of the system state on

monitoring network down time. Depending on the

construction of the monitoring system it might

handle changes easily but the statistics collection

part of the system has to be able to figure out

differences between before and after situation and

the changes between, which might not be an easy

task at all. For this reason a prototype is under

construction through which the scale technology’s

practical limitations can be studied.

The savings that need to be achieved through

the minimizing of manual labour and the costs of

implementation should be noted and the question

should be asked whether or not benefits justify the

system being developed into an operational model in

AUTOMATIC & UNMANNED STOCK REPLENISHMENT PROCESS USING SCALES FOR MONITORING

161

industry. The researchers hope to get an answer to

this problem through building the prototype.

It is the researchers’ opinion that the system

offers operational cost savings as well as new

business models. Operational cost savings are

achieved by optimizing refill amounts, by reducing

the frequency of visits on factory sites and by

improving the reaction to understock situations. In

addition deepened co-operative relationship between

customer and supplier results in synergy benefits.

Supplier expands supply to customer and service

improves for the customer. With the most efficient

stocking strategy, a supplier is able to deliver service

to customers which are far away from the delivery

centre (can’t be served using current operational

model). Furthermore, when competing for new

customers, the supplier has a competitive edge. The

clear advantage to the customer is the availability of

more detailed reports. For example, the analysis of

an understock situation is easier than before because

the supplier will be able to give detailed shelf

inventory balances and historical data to the

customer. From the supplier’s point of view, one can

be assured of maintaining the current customer base

and raising interest of new customers by offering a

better reporting system than the competition.

Will this implementation model be too

expensive compared to its benefits? It’s the

researcher’s opinion that this will only be clarified

through the prototype as well as through inspecting

constructing and maintaining tasks of the system in

practice. The businesses involved in the study also

felt that it would be important to find out customer

groups that are best suited to the proposed system.

Even though the proposed solution might be

financially non-viable in its entirety at this stage, it's

good to be aware of how the technology and costs

should develop in future in order to achieve the

necessary economic benefits which would make the

proposed solution viable.

7 FURTHER RESEARCH

The topic areas for further research will be price-

quality ratio analyses, the possibilities of

optimization of visit frequencies through simulations

and future cost predictions based on historical data

(basically mathematical analysis). Cost research will

concentrate on the total costs including servicing,

maintenance, installation, calibration, system

integrations and information engineering costs.

The researchers are also considering doing a

theoretical study about information exchange and

the problems associated with it. The Collaborative

Planning, Forecasting and Replenishment CPFR

framework will be studied as a development

direction for VMI operations and also the

exploitation and prediction of measuring data, to

assure that the ”service quality” of automated

ordering is not forgotten.

Because C-class products have an

indeterminable balance in the engineering industry,

their management models have been some what ad-

hoc in nature. Proposed solution would change the

nature of C-class products to that of B-class

products, so it would be natural to investigate B-

class management models in this context. It would

be particularly interesting to study how well current

Enterprise Resource Planning (ERP) systems could

be integrated into this type of “approximate”

inventory balance produced by this kind of

monitoring system. In other words, can current ERP

systems and their B-class management models be

directly adopted (for these C-class products) as is, or

does a completely new management theory have to

be developed for them?

REFERENCES

Disney, S.M., Towill, D.R., 2002. A Procedure for the

optimization of the dynamic response of a Vendor

Managed Inventory system. Computer & Industrial

Engineering, volume 43, issue 1-2, pages 27 – 58.

Federation of Finnish Enterprises. Retrieved 21.10.2006,

from http://www.yrittajat.fi/sy/home.nsf/www/english

Katz, M., Klaris A., Scorpio C., 2000. CPFR: Moving

Beyond VMI. Bobbin, Volume 41, Number 9, pages

90-94.

Pohlen, T.L., Goldsby, T.J., 2003. VMI and SMI

programs: How economic value added can help sell

the change. International Journal of Physical

Distribution & Logistics Management, volume 33,

number 7, pages 565 – 581.

SERVIISI-Project (In Finnish). Retrieved 17.10.2006,

from http://partnet.vtt.fi/serviisi/

Statistics Finland, Palvelualojen toimialakatsaus IV/2005

(in Finnish). Retrieved 20.10.2006, from

http://www.stat.fi/til/artikkelit/2006/art_2006-03-

28_004.html?s=0

Statistics Finland, Teollisuuden toimialakatsaus I/2006 (in

Finnish). Retrieved 20.10.2006, from

http://www.stat.fi/til/artikkelit/2006/art_2006-06-

28_001.html?s=0

TEMO-Project (In Finnish). Retrieved 10.10.2006, from

http://partnet.vtt.fi/temo/

Elvander, M.S., 2005. A theoretical mapping of the VMI

concept – A literature review. Report ISRN

LUTMDU/TMTD-4016--SE, Lund University. 40

pages.

WEBIST 2007 - International Conference on Web Information Systems and Technologies

162