APLYING THE KoFI METHODOLOGY TO IMPROVE

KNOWLEDGE FLOWS IN A MANUFACTURING PROCESS

Oscar M. Rodríguez-Elias

, Alberto L. Morán

, Jaqueline I. Lavandera

Aurora Vizcaíno

and Juan Pablo Soto

Mathematics Department, Universidad de Sonora, Blvd. Luis Encinas y Rosales, Hermosillo, Son., Mexico

Facultad de Ciencias, Universidad Autónoma de Baja California, Ensenada, B. C., Mexico

FAMOSA-Ensenada, Ensenada, B.C., Mexico

Alarcos Research Group, University of Castilla-La Mancha, Paseo de la Universidad No. 4, Ciudad Real, Spain

Keywords: Knowledge Management, Manufacturing Process, Process Analysis, Knowledge Flow Analysis.

Abstract: Development of methods to integrate Knowledge Management (KM) in organizational processes is an open

issue. KM should facilitate the flow of knowledge from where it is created or stored, to where it is needed to

be applied. Therefore, an initial step towards the integration of KM in organizational processes should be

the analysis of the way in which knowledge is actually flowing in these processes, and then, to propose

alternatives to improve that flow. This paper presents the use of the Knowledge Flow Identification (KoFI)

methodology as a means to improve a manufacturing process knowledge flow. Since KoFI was initially

developed to analyze software processes, in this paper we illustrate how it can also be used in a

manufacturing domain. The results of the application of KoFI are also presented, which include the design

of a knowledge portal and an initial evaluation from its potential users.

1 INTRODUCTION

To assist organizations to manage their knowledge,

different strategies and systems (Knowledge

Management Systems, KMS) have been designed.

However, developing them is a difficult task; since

knowledge per se is intensively domain dependent

whereas KMS often are context specific

applications. The lack of appropriate methodologies

or theories for the extraction of reusable knowledge

and reusable knowledge patterns has proven to be

extremely costly, time consuming and error prone

(Gkotsis, Evangelou et al., 2006). Additionally, an

actual concern is that KM approaches should be well

integrated to the knowledge needs of knowledge

workers, and to the work processes of organizations

(Scholl et al., 2004). Before developing a KM

strategy it is advisable to understand how knowledge

transfer is carried out by people in the different

processes where the strategy will be applied.

This paper presents the use of the KoFI

methodology developed to identify and analyze

knowledge flows in work processes, to improve a

manufacturing process. The goal of this paper is to

illustrate how this methodology can help to detect

knowledge deficiencies in a process, and can also

help to design strategies to solve them; in this case a

knowledge portal was designed. Hence, in the next

section the manufacturing process where the

methodology was used is described, after that in

Section three we illustrate the different stages

followed to improve that process. Then, in Section

Four a knowledge portal, designed as a result from

the findings obtained after applying the

methodology, is described. Section Five depicts the

results of a preliminary evaluation of this portal;

finally conclusions are outlined in Section 6.

2 THE MANUFACTURING

PROCESS

To test the KoFI methodology it was used in an

industrial company dedicated to the manufacturing

of cans. We focused our work on a department

where eight processes are carried out. It was decided

to centre on one of the most important process: the

one in charge of transforming the aluminum rolls

into the first versions of the cans (known as

309

M. Rodríguez-Elias O., L. Morán A., I. Lavandera J., Vizcaíno A. and Pablo Soto J. (2008).

APLYING THE KoFI METHODOLOGY TO IMPROVE KNOWLEDGE FLOWS IN A MANUFACTURING PROCESS.

In Proceedings of the Third International Conference on Software and Data Technologies - PL/DPS/KE, pages 309-314

DOI: 10.5220/0001879603090314

 SciTePress

“Formation area”). In this test 41 people were

involved, including the department manager, the

responsible of each area of the department, and the

operating personnel, which were integrated by leader

mechanics, productive processes mechanics, and

machine operators.

Nineteen employees were interviewed by using

the long interview technique. The duration of the

interviews ranged from 30 minutes to 2 hours,

depending on the level of responsibility of those

interviewed. Additionally, a total of 119 documents

and systems were also analyzed.

3 APPLYING KoFI TO THE

PROCESS

The KoFI methodology is divided in three phases

(Rodriguez-Elias et al., 2007a): a) the process

modeling phase, consisting of the definition and

modeling of the process, using a process modeling

language which provides elements to represent the

knowledge involved in the process; b) the process

analysis phase, which involves the identification and

analysis of knowledge sources, topics, and flows, as

well as the problems affecting the flow of

knowledge; and c) the knowledge flow support tools

analysis phase, consisting of the analysis of the tools

that might be useful knowledge flow enablers.

In this paper we will focus on the process

analysis phase. Information about how to perform

the other two phases can be found in (Rodriguez-

Elias et al., 2007b) for the process analysis phase,

and in (Rodriguez-Elias et al., 2007c) for the

knowledge flow support tools analysis phase.

To identify

knowledge flow

problems

To identify

knowledge flows

To identify

knowledge topics

To identify

knowledge sources

Figure 1: The four steps of the process analysis phase of

the KoFI methodology.

The analysis phase of KoFI is composed of four

steps, as shown in Figure 1, which are performed in

an iterative way, since each step might provide

information useful for the others preceding it.

The first step is to identify the knowledge

sources involved in the process. This includes the

identification of all those sources of information or

knowledge that could be being used or could be

useful for performing the different activities

composing the processes. Those sources could

include the people consulted by the personnel in

charge of the process, the information systems

supporting the process, or documents.

The second step focuses on the identification of

the main knowledge topics or areas related to the

activities performed in the process. For instance,

knowledge required to perform the activities, or

created from them. The knowledge related to the

sources found in the preceding step should be

identified and classified. An important result of this

step might be the identification of important

knowledge topics not stored anywhere, or that might

be stored in sources not used or difficult to find.

These two initial steps also include the

classification of the sources and topics found, which

can be made through the definition of a taxonomy or

an ontology of knowledge sources; which are

considered an important initial activity towards the

development of KM systems (Rao, 2005). It should

be possible to relate the different sources to the

knowledge that can be obtained from them, and vice

versa, i.e. relate the knowledge to the sources from

where it can be obtained, or where it is stored.

The third step focuses on identifying the manner

in which knowledge is flowing through the process.

To accomplish this, it is required to analyze the

relationships between the knowledge sources and

topics, to the activities of the process. This includes

the identification of the activities where the topics

and sources of knowledge are being generated,

modified, or used. It is important to identify

knowledge dependencies, such as knowledge topics

generated in an activity and required in other; and

knowledge transfers mechanisms, such as

knowledge transferred from one activity to another

through a document, or through an interaction

between different roles or persons.

Finally, the fourth step of the analysis consists of

identifying and classifying the main types of

problems detected and which affect the knowledge

flow. KoFI proposes to do this by defining problem

scenarios (Rodriguez-Elias et al., 2007a), a

technique based on explaining a problem in the form

of a story describing a common situation. Once

described the problem, one or more alternative

scenarios are also proposed to illustrate the manner

in which such a problem could be addressed. Those

alternative scenarios are finally used to extract the

main requirements to propose the KM strategy to

follow, or the KM system to develop. The following

subsections describe how these steps were carried

out in the manufacturing company.

3.1 Identifying Knowledge Sources

In the first step of the analysis, the identified sources

were very diverse. To facilitate its management, and

ICSOFT 2008 - International Conference on Software and Data Technologies

310

following the recommendations of the KoFI

methodology, once the different sources were

identified, we proceeded to classify them. To do this

a taxonomy of knowledge sources was defined; it

included four categories of sources:

1) Documents, groups of all those sources which

consist of physical or electronic documents. It

includes three subcategories: a) process’s

documents, b) technical documents, and c)

organizational documents.

2) Information Systems, refers to the sources

consisting of information systems used in the

company. This category includes two

subcategories: a) query systems, and b)

transactional systems.

3) People, groups all the different types of people

involved in the process. It has been divided in

four subcategories: a) staff, b) specialists, c)

external clients, and 4) internal clients.

4) Others, groups those sources not included in

the preceding categories. Particularly it includes

two subcategories: a) problem analysis tools,

and b) simulation tools.

Each source was described by assigning it a unique

identifier, a name, a description, its type and

category, its location, its format, and the main

knowledge topics which could be obtained from it.

3.2 Identifying Knowledge Topics

The identified knowledge topics were also very

diverse, ranging from organizational behavior to

special machine maintenance. The topics identified

were classified in three categories, according to their

utility in the activities of the process.

1) Product Line Activities which includes

knowledge about the operation of machines,

about processes, and about quality of the

processes and products. It is divided in four

subcategories: a) product quality, b) machine

maintenance, c) operation, and d) information

technology (IT) application.

2) Organizational Culture, is all that knowledge

that employees must have about the company,

its internal organization and norms, etc. It

includes only one subcategory which is

knowledge of the company.

3) General Knowledge groups all those topics and

areas of knowledge that the employees might

have, and which is not directly related to the

process operation. It is subdivided in four

subcategories: a) resource management, b) IT

management, c) personnel management, and d)

other individual knowledge.

Once identified, the main knowledge topics were

described assigning them a unique identifier, a

name, a description, its classification, and

information to know where such topic could be

useful, and why and how knowing it could benefit

the organization or the person who knows about it.

With the knowledge topics descriptions, a

knowledge dictionary was developed for the process.

3.3 Identifying Knowledge Flows

In this step we modeled the knowledge required in

each activity of the process, the knowledge that each

role needs to perform these activities, and the

knowledge sources consulted or generated in each

activity, following an adaptation of the Rich Picture

technique (Monk and Howard, 1998). Figure 2

presents an example of this type of diagrams, in

which there are represented the knowledge required

in the “Lift trucks operation and management”

process carried out in the company studied. The

figure shows the role in charge of such activity, the

experience, skills and knowledge it provides to the

activity, and the main source of knowledge used in

the activity, which is an application for managing

security rules and regulation of the company.

Mechanics/

Operator

Secure System

Lift trucks operation

and management

•Operation procedures

•Security norms

•Transported material

•Security rules and

regulation

•Process work's experience

•Tools management

experience

•IT management experience

Knowledge required

for the activity

Activity

Knowledge

source

Knowledge obtained

from a source

Role / Actor

Knowledge provided

by Role / Actor

Mechanics/

Operator

Mechanics/

Operator

Secure System

Lift trucks operation

and management

•Operation procedures

•Security norms

•Transported material

•Security rules and

regulation

•Process work's experience

•Tools management

experience

•IT management experience

Knowledge required

for the activity

Activity

Knowledge

source

Knowledge obtained

from a source

Role / Actor

Knowledge provided

by Role / Actor

Figure 2: Example of an adapted rich picture to analyze

knowledge flows.

This type of models helped us to identify the

relationships between the knowledge sources and

topics, and the activities of the process. The above

allowed us to create a knowledge meta-model

(described in Section 4), which was used as the

structure for developing a Knowledge Map useful to

identify the knowledge that might be obtained from

each source, and the activities in which the sources

or the knowledge were being used or generated. This

map was used in the construction of a Knowledge

Portal (described also in Section 4) proposed to

solve some of the main knowledge flow problems

observed, as it is described next.

APLYING THE KoFI METHODOLOGY TO IMPROVE KNOWLEDGE FLOWS IN A MANUFACTURING PROCESS

311

3.4 Identifying Knowledge Flows

Problems

The final step of KoFI proposes to identify and

classify the main problems affecting the knowledge

flow in order to propose alternatives to minimize or

avoid them. In our study, it was observed that some

areas of the process were not well supported with

documentation. An additional problem was the

identification of important knowledge sources that

were not being used. Some reasons for the last were

the difficulty for consulting some of those sources,

either because they were unknown, or because they

were difficult to find by employees.

To address this problem, it was decided to

develop a Knowledge Portal to facilitate the access

to all the available sources, according to the areas,

processes, or activities for which they are useful.

Additionally, the portal would provide ways for

pointing out to all those knowledge areas for which

no sources exist. The last should be useful to

identify all those areas for which knowledge sources

should be created. Additionally, it was also decided

that the portal should provide access not only to

documents, but also to other types of sources, such

as information systems, or support tools, in order to

promote the use of all the available types of

knowledge sources of the company.

4 DESIGN OF THE KNOWLEDGE

PORTAL

In this section we describe a meta-model developed

for structuring the knowledge map used into the

portal, the structure of such portal, and the design of

its user interface.

4.1 Meta-Model

The meta-model comprises the knowledge types and

sources involved in the knowledge generation and

acquisition process (Figure 3). In it, the knowledge

concepts are integrated with the knowledge topics

and sources. The knowledge concepts are required,

generated or modified by the activities, which are

described as work definitions. The work definitions

can be processes, activities or decisions. Each

knowledge concept/source association contains

information about the knowledge level it requires.

The available format and location for consulting

each source are specified.

Knowledge

source

Location

Source

types

Location

types

Source

categories

Format

Knowledge

concept

Required level

Knowledge

topics

Categories Areas

Work

definitions

Goal

Processes Activities

Decisions

Located in Of type

Knows about

Has

launches…

Implies

1..*

Has

1..*

Has

Figure 3: Meta-model of knowledge types and sources.

4.2 Knowledge Portal Structure

The meta-model was used as a base to design the

structure of the knowledge portal. Figure 4 shows

the resulting general structure of the portal. This

structure comprises a first level in which initial

interfaces (pages) are accessible (e.g. home and

registration pages). The second and third levels are

pages which correspond to the manufacturing areas

and sub-areas of the organization, respectively. The

fourth level corresponds to pages on the processes

that integrate each of the sub-areas identified from

the involved knowledge flows. Finally, the fifth

level presents all the identified knowledge sources

for the specific process of the sub-area.

Homepage

Area Area Area

Sub-area Sub-area Sub-area

Sub-process Sub-process Sub-process

Knowledge source Knowledge source Knowledge source

Homepage

Area Area Area

Sub-area Sub-area Sub-area

Sub-process Sub-process Sub-process

Knowledge source Knowledge source Knowledge source

Figure 4: General structure of the Knowledge Portal.

4.3 Knowledge Portal UI Design

The design of presentation and navigational features

of the user interfaces (pages) also emerged from

insights identified in the analysis and initial phases

of design. These include information about the

identified knowledge flows, the main sub-areas of

the organization, and the structure of the portal

previously identified, which resulted in the options

included in the menus and main layout sections of

the pages. These allow users to find the required

information by simply identifying the specific area

in which information is generated or required, and

following the resulting navigational structure (area

→ sub-area → process) to locate the specific

knowledge source, instead of just alphabetically (or

randomly) browsing through the information.

ICSOFT 2008 - International Conference on Software and Data Technologies

312

a) Area

b) Sub-areac) Process

d) Knowledge

sources

e) Contextual

f) Search

engine

a) Area

b) Sub-areac) Process

d) Knowledge

sources

e) Contextual

f) Search

engine

Figure 5: Example of the page contents and layout of the

Knowledge Portal.

Figure 5 depicts an example of the layout and

content of a page from the current prototype for the

“Formation” area.

The information provided includes the name of

the manufacturing area being consulted (5.a), the

name of the specific sub-area (5.b), the name of the

selected process within the sub-area (5.c), and most

importantly, links to knowledge sources (and types)

available for that process (5.d).

Additionally, the page includes a “contextual”

sub-area menu to facilitate navigation through the

information (5.e), which is always available while

the user stays in that particular sub-area of the

portal. Also, it includes a search engine (5.f) which

allows a search to be performed by simply

specifying a keyword on the required topic, and

optionally, the “places” in which the information

should be searched for.

The interface in Figure 5 represents the final

destination for users looking for a particular

knowledge source who, by following only three

links (area → sub-area → process), arrive at the

knowledge sources (either documents, systems or

people) required to perform their intended activities.

Finally, this design adheres to the organization’s

established standard guidelines for this kind of

applications.

5 EVALUATION OF THE

KNOWLEDGE PORTAL

We conducted a preliminary evaluation in one of the

production areas to determine the impact and

acceptance level of the users on the system, and to

provide support for the decision-making process

concerned with the continuation of the system’s

implementation in other areas of the organization.

The evaluation considered aspects concerning

perception of usefulness and ease of use (Davis,

1989). The evaluation consisted of 1) an induction

session, in which the system was presented to the

users, and its functionality demonstrated to them.

This included examples on how to search for and

retrieve knowledge sources by means of navigating

through areas, sub-areas and processes, as well as

through the search engine; and 2) the application of

a questionnaire containing 12 questions referring to

perception of usefulness (6) and ease of use (6).

Each evaluation session (induction and application

of the questionnaire) was done in about one hour.

The subjects of the study were 41 employees of

the “Formation” area for which the prototype was

developed, whose participation was voluntary. The

sample was divided into 4 groups according to the

natural operative processes (3 groups of ten people

and 1 of eleven). The application process of the

evaluation was completed in three days.

5.1 Analysis and Discussion of

Evaluation Results

The subjects had positive appreciations with regard

to the knowledge portal, as is reflected in their

answers in the questionnaire. Figure 6 shows the

answers to the questions about the perception of

usefulness of the tool. The users perceived that the

portal would allow them to increase their

productivity and to perform their tasks more easily

(82.93% “Agree” in both cases), although some of

them had doubts regarding the fact that this would

increase their productivity (24.39% “Have Doubts”).

Only one person (2.44%) “Disagreed” that the tool

would help him/her to complete his/her tasks faster.

Perception of Usefulness

78.05

82.93

75.61

80.49

82.93

87.8

19.51

17.07

24.39

19.51

17.07

12.2

2.44

0% 20% 40% 60% 80% 100%

Complete the task

faster

Increase task

performance

Inc rease

productivity

Improve efficiency

Ease the task

Is useful

Agree (%)

Have Doubts (%)

Disagree (%)

Figure 6: Perception of Usefulness.

Figure 7 shows the answers to the questions

about the perception of ease of use. As can be seen,

although most of the users perceived that it was easy

to learn to browse through the information (85.37%

“Agree”), some had doubts concerning the ease of

finding information (39.02% “Have Doubts”), and

APLYING THE KoFI METHODOLOGY TO IMPROVE KNOWLEDGE FLOWS IN A MANUFACTURING PROCESS

313

even more users had doubts concerning becoming

experts on the use of the tool (46.34% “Have

Doubts”). A possible explanation could be that a

little more than a third of the users had doubts

concerning the clarity of the presented interfaces, as

well as about the interaction flexibility that these

provide (34.15% in both cases).

Perception of Easy of USe

85.37

60.98

65.85

53.66

68.29

14.63

39.02

34.15

46.34

31.71

0% 20% 40% 60% 80% 100%

Learning to browse

Finding information

Clear user

interfaces

Flexible interaction

Becoming an

expert

Is easy to use

Agree (%)

Have Doubts (%)

Disagree (%)

Figure 7: Perception of Easy of Use.

In general, most of the users considered the

knowledge portal as a useful (87.80% “Agree” –

Figure 6) and easy to use tool (68.29% “Agree” –

Figure 7) for the accomplishment of their work.

6 CONCLUSIONS

In this paper we have illustrated the use of the KoFI

methodology to analyze a manufacturing process in

order to improve the flow of knowledge in it. The

KoFI methodology was initially developed to aid in

the design of KM approaches to improve software

processes. In this initial application domain, the

methodology was also useful to propose the design

of KM tools, and to structure and create knowledge

maps of the studied processes (Rodriguez-Elias et

al., 2007a, 2007b, 2007c).

In the present study the processes were much

more formally defined and documented than those of

the previous studies we made. Also, these processes

were already modeled with a common business

process modeling language, which has not explicit

representation of knowledge related issues. From the

models we made in the study, we were able to

identify knowledge requirements and sources, which

were not identified from the existent process models

of the company. This observation has gave us

insights to argue that independently of how well

defined and documented the process could be, if

there is not an explicit representation of the

knowledge and sources involved in the activities of

the process, important sources and knowledge

requirements could be lost or ignored during the

analysis.

Finally this study has provided us with the initial

evidence to argue that KoFI is open enough to aid in

the design and construction of different types of KM

approaches, and in different domains. However,

more case studies are required to continue evaluating

the benefits and limitations of KoFI in different

settings. This constitutes part of our ongoing and

future work.

ACKNOWLEDGEMENTS

This work is partially supported by UABC, project

0191 of the XI Convocatoria Interna de Proyectos,

and the MELISA project (grant PAC08-0142-3315)

financed by the Consejería de Educación y Ciencias

de la Junta de Comunidades de Castilla-La Mancha,

Spain. The authors also acknowledge the support

provided by FAMOSA-Ensenada.

REFERENCES

Davis, F. D. (1989). Perceived usefulness, perceived ease

of use, and user acceptance of information technology.

MIS Quarterly, 13(3), 319-340.

Gkotsis, G., C. Evangelou, N. Karacapilidis and M.

Tzagarakis (2006). Building Collaborative

Knowledge-based Systems: A Web Engineering

Approach. WWW/Internet, Iadis International

Conference, Murcia.

Monk, A., & Howard, S. (1998). The rich picture: A tool

for reasoning about work context. Interactions, 5(2),

21-30.

Rao, M. (Ed.). (2005). Knowledge management tools and

techniques: Practitioners and experts evaluate km

solutions. Amsterdam: Elsevier.

Rodríguez-Elias, O. M., Martínez García, A. I., Favela, J.,

Vizcaíno, A., & Soto, J. P. (2007a). Knowledge flow

analysis to identify knowledge needs for the design of

knowledge management systems and strategies: A

methodological approach. Proc. of the 9th ICEIS,

Funchal, Madeira - Portugal.

Rodríguez-Elias, O. M., Martínez-García, A. I., Vizcaíno,

A., Favela, J., & Piattini, M. (2007b). Organización de

conocimientos en procesos de ingeniería de software

por medio de modelado de procesos: Una adaptación

de SPEM. Proc. Of the VI JIISIC, Lima, Perú.

Rodríguez-Elias, O. M., Martínez-García, A. I., Vizcaíno,

A., Favela, J., & Piattini, M. (2007c). A framework to

analyze information systems as knowledge flow

facilitators. Information and Software Technology,

50(6):481-498.

Scholl, W., König, C., Meyer, B., & Heisig, P. (2004). The

future of knowledge management: An international

delphi study. Journal of Knowledge Management,

8(2), 19-35.

ICSOFT 2008 - International Conference on Software and Data Technologies

314