A KNOWLEDGE-BASED PERFORMANCE MEASUREMENT

SYSTEM FOR IMPROVING RESOURCE UTILIZATION

Annie C.Y. Lam, S.K. Kwok and W.B. Lee

Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hung Hom, Hong Kong

Keywords: Performance measurement, Resource utilization, Knowledge-based system.

Abstract: In current manufacturing industry, there are various challenges including short product life cycle, process

automation and global competition. It is critical for the manufacturing companies to ensure effective

utilization of production assets for overall business success. In order to focus scarce resources on areas that

have the greatest impact on productivity, performance evaluation on the resource allocation is necessary to

assist companies improving the resource utilization and accomplishing their objectives. In this paper, a

knowledge-based performance measurement system (KPMS) is designed to evaluate the resource allocation

decisions and provide recommendations to improve the performance and physical asset utilization. The

framework of the proposed system, which is constructed by rule-based reasoning, case-based reasoning and

a mathematical model, is introduced. By integrating the mathematical model with knowledge rules,

performance indicators that are associated with the achievement of company objectives can be determined

to quantify the performance of the resource allocation. Moreover, case-based reasoning technique is adopted

to evaluate the performance and reuse the experience in past cases to provide recommendations for

improvement.

1 INTRODUCTION

In recent years, manufacturing companies are

confronted with challenges of shortening product

life cycles, growing emphasis on process

automation, global competition, and increasingly

mobile work force. These challenges require robust

methods for providing real-time monitoring and

control, understanding asset utilization and capacity,

and retaining valuable operation knowledge within

company, in order to achieve reliable and economic

performance (Wang et al, 2006). Moreover, it is

necessary to ensure effective utilization of

production assets for overall business success. The

importance of asset management is growing.

To achieve the purpose of asset management,

maintenance strategies have been widely adopted to

reduce downtime and increase both quality and

productivity. Many researchers have attempted to

develop various preventive and predictive

maintenance tools to assist companies monitoring

the health degradation of machines and scheduling

maintenance activities so as to avoid machine

downtime. Although maintenance is significant in

maintaining the physical assets in specified

operating condition, it is still incapable to guarantee

that companies can gain greatest utilization and

effectiveness from their physical assets. To tackle

this problem, it is necessary to measure the

performance of resource utilization in order to

reflect the productivity of the companies.

From the review of current studies, performance

criteria were chosen to evaluate the performance

pertaining to a specific goal of the company only.

However, in the turbulent and competitive market,

manufacturing companies often have to employ

different strategies in order to react to changes

responsively. This would affect the relative

importance of manufacturing criteria for

measurement. Therefore, performance measurement

should be a dynamic process, in which the adoption

of performance criteria and their relative importance

are dependent of the strategies and objectives of the

company.

In this paper, a knowledge-based performance

measurement system (KPMS) is introduced. The

purposes of the system are:

(i) Establish appropriate performance indicators in

accordance with the company objectives.

(ii) Evaluate the performance of resource allocation

187

C. Y. Lam A., K. Kwok S. and B. Lee W. (2008).

A KNOWLEDGE-BASED PERFORMANCE MEASUREMENT SYSTEM FOR IMPROVING RESOURCE UTILIZATION.

In Proceedings of the Tenth International Conference on Enterprise Information Systems - AIDSS, pages 187-192

DOI: 10.5220/0001684001870192

 SciTePress

and provide recommendations to improve

physical asset utilization.

(iii) Retain the knowledge of resource allocation in

the company to support continuous

improvement.

To achieve the above purposes, the framework of the

proposed KPMS consisting of seven main

components, which are constructed by rule-based

reasoning, case-based reasoning and a mathematical

model, is developed. The remainder of the paper is

organized as follows. Section 2 presents the reviews

of related literature. The system architecture of the

proposed system is explained in Section 3. In

Section 4, the functions and benefits of the system

are discussed. Finally, a conclusion is given in

Section 5.

2 LITERATURE REVIEW

2.1 Background Study

In today’s global and competitive manufacturing

industry, companies often need to react quickly and

cost-effectively to contingent changes such as

demand fluctuations and changes in customer

requirements. Their largest challenge is to increase

operational effectiveness, profitability and customer

satisfaction and reduce operating costs. As physical

assets form the basic infrastructure of all businesses,

it is essential to plan and monitor assets throughout

their entire life cycle to ensure effective utilization

of production assets.

Asset management is a strategic and integrated

set of comprehensive processes, which include

financial, management, engineering, operating and

maintenance, to gain greatest lifetime effectiveness,

utilization and return from physical assets (Mitchell

and Carlson, 2001). Its overall goal is to optimize

productivity in long term. Most researchers have

emphasized the importance of maintenance in asset

management. Preventive and predictive maintenance

approaches have been adopted to avoid machine

failure (Lee et al., 2006; Wang et al., 2006).

Although maintenance is critical in failure

prevention to assure the availability and reliability of

machines, asset management should encompass a

broader range of activities. Performance

measurement is necessary to evaluate the asset

utilization and improve the decision making quality

of resource allocation.

2.2 Performance Measurement in

Manufacturing

According to Ahmad and Dhafr (2002), a suitable

measurement methodology enables companies to

focus scarce resources on areas that have the greatest

impact on productivity. Accordingly, additional

production capacity can be achieved without

investing new machinery. Performance measurement

can quantify the efficiency and effectiveness of

action that leads to performance. It is used to

evaluate how well the activities within a process, or

the outputs of a process, achieve specified goals

(Chen, 2008). Key performance indicator (KPI) can

be used to compare with internal or external target

and identify any performance gaps.

In measuring manufacturing performance, most

manufacturers use criteria such as quality,

productivity, speed, customer satisfaction, diversity

of product line and flexibility. Ahmad and Dhafr

(2002) assessed manufacturing performance in the

areas of quality, delivery reliability, cost and

delivery lead time as they are important aspects of

manufacturing performance areas and easy to

measure.

2.3 Research Issues

The strategy adopted by a manufacturing company

can directly affect the relative importance of

performance criteria. However, most researchers

chose performance measures according to a

particular objective such as profitability (Yurdakul,

2002), financial health (Wen et al., 2008) and plant

utilization (Ahmad and Dhafr, 2002). It seems that

these measures are not dynamic in nature and cannot

adapt to the changing manufacturing environment.

Since companies may revise their objectives over

time, performance measurement should be a

dynamic process that the performance criteria and

their importance can be adjusted accordingly to

reflect the changes in companies’ objectives.

Moreover, Chen (2008) stated that existing

integrated performance measurement systems do not

have an explicit feedback loop that supports

improvement. It implies a need to include feedback

mechanisms in performance measurement for

companies to achieve continuous improvement.

To address these issues, this paper presents a

knowledge-based performance measurement system

which helps to establish appropriate performance

indicators to deal with changes of company’s

objective in a dynamic environment. The proposed

system is used to evaluate the performance of

ICEIS 2008 - International Conference on Enterprise Information Systems

188

physical asset under different resource allocation

decisions. Moreover, suggestions can be provided to

improve the decision making quality of resource

allocation, so that utilization of physical asset can be

significantly improved.

3 THE ARCHITECTURE OF THE

KNOWLEDGE-BASED

PERFORMANCE

MEASUREMENT SYSTEM

The proposed system is a knowledge-based system

aiming at assisting manufacturing companies to

evaluate the effectiveness of resource allocation in

accordance with their company objectives. The

proposed system not only assesses the effectiveness

of the physical assets in performing the job orders

and attaining the company objectives, but also

provides recommendations to improve productivity

performance. As illustrated in Figure 1, the

framework of the KPMS is composed of seven main

components: web-based platform, performance

indicator selector, performance indicator analyzer,

overall performance scoring model, performance

evaluation, database and knowledge repository.

3.1 Web-Based Platform

The web-based platform is the user interface of the

system that enables users to access its functions over

the Internet with a standard web browser. It consists

of web pages which are constructed by HyperText

Makeup Language (HTML). Moreover, Active

Server Page (ASP) and JavaScript are embedded to

make the web pages dynamic and interactive and

perform data validation and checking respectively. It

also allows the staff to input their requests of

performance measurement such as company

objective criteria and resource allocation decision.

Examples of the objective criteria for performance

assessment include profitability, productivity,

product quality and customer satisfaction. On the

other hand, details of the resource allocation such as

product type, machine category and manufacture

process have to be entered for the proposed system

to evaluate current performance.

The system displays the performance level of the

assessed resource allocation decision and

recommendations for improvement through the web-

based platform. In addition, the rules stored in the

knowledge repository for selecting appropriate

performance indicators and assigning performance

scores can be checked and updated through this

platform in order to cope with a fast changing

environment.

Figure 1: System architecture of KPMS.

3.2 Performance Indicator Selector

The performance indicator selector adopts rule based

reasoning method to select the appropriate

performance indicators for measurement. The

knowledge of selecting suitable performance

indicators is presented in the form of “if <antecedent

clauses> then <consequent clauses>” statements. If

the antecedent clauses are true, then the consequent

clauses are true. Based on the knowledge rules, the

selector chooses the relevant performance criteria

and their involved key performance indicators

according to the company objectives. Many

companies tend to pursue certain objectives such as

profitability, market share, customer focus, product

quality and productivity. As shown in Figure 2, they

are related to different performance criteria

including cost, quality, customer satisfaction and

efficiency, which consist of different key

performance indicators. Examples of rules involved

in the performance indicator selector are presented

below:

A KNOWLEDGE-BASED PERFORMANCE MEASUREMENT SYSTEM FOR IMPROVING RESOURCE

UTILIZATION

189

Figure 2: Hierarchical structure of performance indicator selection.

Rule 1:

If Company_Objective = Profitability then

Performance_Criteria = “Profit, Cost,

Dependability, Quality”

Rule 2:

If Company_Objective = Productivity then

Performance_Criteria = “Cost, Quality, Time,

Efficiency”

Rule 3:

If Performance_Criteria = Cost then

Key_Performance_Indicator = “Unit_Cost,

Inventory_Cost, Running_Cost”

Rule 4:

If Performance_Criteria = Quality then

Key_Performance_Indicator =

“First_Pass_Yield, Defect_Ratio, Scrap_rate”

After knowledge reasoning in the rules, performance

indicators that are associated with defined company

objectives are selected and passed to performance

indicator analyzer for analysis.

3.3 Performance Indicator Analyzer

The performance indicator analyzer is responsible

for analyzing operational data retrieved from the

enterprise database and assigning score to the key

performance indicators determined by the

performance indicator selector. The values of the

performance indicators can be retrieved from the

database directly or determined by equations. For

instance, the key performance indicator of defect

ratio can be found in Equation (1).

productionTotal

Defects

ratioDefect =

(1)

The number of defect units and total production

units are retrieved from the database and calculated

to get the defect ratio. After that, the value of the

performance indicator is analyzed together with

decision rules retrieved from the rule base so as to

assign a performance score to each indicator. An

example of knowledge rules to assign a score to the

performance indicator of defect ratio is given below:

Rule 1:

If Defect_Ratio < 5% then Performance =

“Very satisfied” and Defect_Ratio_KPI = 4

Rule 2:

If Defect_Ratio > 5% and Defect_Ratio <

10% then Performance = “Satisfied” and

Defect_Ratio_KPI = 3

Rule 3:

If Defect_Ratio > 10% and Defect_Ratio <

20% then Performance = “Normal” and

Defect_Ratio_KPI = 2

Rule 4:

If Defect_Ratio > 20% and Defect_Ratio <

40% then Performance = “Unsatisfied” and

Defect_Ratio_KPI = 1

Rule 5:

If Defect_Ratio > 40% then Performance =

“Very unsatisfied” and Defect_Ratio_KPI = 0

3.4 Overall Performance Scoring

Model

The overall performance scoring model is used to

integrate all key performance indicators into an

overall performance score according to appropriate

weighting. The parameters used in the model are

shown below.

ICEIS 2008 - International Conference on Enterprise Information Systems

190

Parameters

Overall performance score of a resource

allocation decision,

Performance score obtained related to

the perspective of objective

Weighting of the objective

i to evaluate

the overall performance,

Number of objectives to be considered

in performance measurement,

KPI

Score of key performance indicator

Weighting of key performance

indicator

to determine the

performance score related to a particular

objective,

Number of key performance indicators

that are required to determine the

performance score related to a particular

objective.

The overall performance scoring model:

∑

xfS

(2)

where

∑

jji

wKPIf

ni L,2,1=

(3)

10 ≤≤

ni L,2,1

(4)

∑

(5)

10 ≤≤

aj L,2,1

(6)

∑

(7)

The goal of Equation (2) is to determine the overall

performance score of a resource allocation

decision,

S , which is calculated by multiplying the

score obtained in each objective perspective,

f ,

with the weighting of that objective on overall

performance,

x .

x represents the relative

importance of the objective

in evaluating the

overall performance and it is determined by the

managers when they input the requests of

performance measurement. As shown in Equation

(3),

f is found by multiplying the score of key

performance indicator,

KPI

, which is determined by

the performance indicator analyzer, with its

weighting to determine the performance score

related to an objective,

w , which is retrieved from

the knowledge base. Moreover, constraints are

included in the model. Constraints (4) and (6)

specify that the weighting of objective and the

weighting of key performance indicator should be a

numeric value between 0 and 1 respectively.

Constraints (5) and (7) require that the summation of

all weighting of objectives and weighting of key

performance indicators should be equal to 1

respectively.

3.5 Performance Evaluation

The performance evaluation module is used to

evaluate the overall performance score of the

resource allocation decision and present the results

of the performance along with suggestions for

improvements. It adopts case based reasoning

method to retrieve past cases for performance

evaluation. The objectives for performance

measurement and the specifications of resource

allocation are case attributes that are used to browse

and retrieve relevant cases from the case library.

After generating a list of cases based on the degree

of similarity, their overall performance scores can be

compared with the new case to assess whether

current resource allocation decision is correct and

identify any new approach to improve the

performance level. If the performance of the current

case is poorer than that of past cases, the resource

allocation decisions in those past cases can serve as

useful suggestions for the staff to reallocate the

resource effectively. Subsequently, a performance

evaluation report can be generated to show the

results and recommendations of the performance

while the new case is retained in the case library.

3.6 Database

The database is built to manage and store different

enterprise information, such as production

requirements, resource conditions, customer order

specification and asset information. Operational data

required in calculating the KPIs is included in the

database and stored in relational table format. In the

proposed KPMS, Microsoft Access is adopted to

create the database and an ODBC driver is used to

access the data stored in the database.

3.7 Knowledge Repository

The knowledge repository contains a rule base and a

case library for storing knowledge of performance

indicator analysis and resource allocation

respectively. The rule base is used for storing the

knowledge of selecting, weighting and analyzing the

key performance indicators. It is presented in “If-

A KNOWLEDGE-BASED PERFORMANCE MEASUREMENT SYSTEM FOR IMPROVING RESOURCE

UTILIZATION

191

Then” rule structure to reason and analyze the key

performance indicators. The case library is used to

record all past cases of resource allocation decisions.

In those cases, the basic job requirements, resource

adoption, performance scores and their objectives

for assessment are organized and saved. They can be

retrieved in the performance evaluation module for

assessing the performance.

4 DISCUSSION

The proposed KPMS enhances the performance of

manufacturing companies by facilitating them to

acquire the greatest utilization and effectiveness

from their production assets, which is the ultimate

purpose of asset management. Performance

measurement enables managers to understand the

effect of their resource allocation decisions through

an overall performance score. As it is a dynamic

system that can select appropriate performance

indicators according to the company objectives, it is

suitable for current competitive environment in

which companies occasionally need to revise their

strategies and objectives for enhancing the

competitiveness. The importance of performance

criteria would be adjusted by the system to react to

the changes in company objectives.

Moreover, the system encourages companies to

achieve continuous improvement on their resource

utilization. After determining the overall

performance score, the assessed resource allocation

decision, as a new case, is evaluated against some

past similar cases of resource allocation retrieved

from the case library in order to realize current

performance and identify any area for improvement.

Consequently, it can suggest ways for improving the

performance by reusing the experience gained in

past cases while the new case of resource allocation

can be retained in the knowledge repository.

This paper mainly provides a framework of the

proposed KPMS. Implementation of the system is

being conducted in a local polyfoam manufacturer.

Once results are obtained, they will be presented in

further publication.

5 CONCLUSIONS

A good resource allocation planning is important for

manufacturing companies to utilize their physical

assets effectively. Performance measurement on the

resource allocation is necessary to assist companies

improving the resource utilization and

accomplishing their objectives. However, selection

of suitable performance indicators in accordance

with company objectives is a complex task

especially in a dynamic manufacturing environment.

The staff usually judge the resource allocation

decisions by their own knowledge and experience.

However, it cannot guarantee that the physical assets

are utilized in the most effective way. In this paper,

the knowledge-based performance measurement

system establishes an approach for evaluating the

performance of resource allocation decisions. It

helps manufacturing companies to improve their

operations in order to gain maximum utilization and

effectiveness from the physical assets.

ACKNOWLEDGEMENTS

The authors wish to thank the Research Committee

of the Hong Kong Polytechnic University for

supporting the project.

REFERENCES

Ahmad, M.M., Dhafr, N., 2002. Establishing and

improving manufacturing performance measures,

Robotics and Computer Integrated Manufacturing,

Vol. 18, pp. 171-176.

Chen, C.C., 2008. An objective-oriented and product-line-

based manufacturing performance measurement,

International Journal of Production Economics, Vol.

112, pp. 380-390.

Lee, J., Ni, J., Djurdjanovic, D., Qiu, H., Liao, H.T., 2006.

Intelligent prognostics tools and e-maintenance,

Computers in Industry, Vol. 57, pp. 476-489.

Mitchell, J.S., Carlson, J., 2001. Equipment asset

management – what are the real requirements?,

Reliability Magazine, October, pp. 4-14.

Wang, W.M., Cheung, C.F., Lee, W.B., Kwok, S.K., 2006.

A knowledge-based infotronics supervisory system for

intelligent service automation, Proceedings of the

Institution of Mechanical Engineers - Part B -

Engineering Manufacture, Vol. 220 No. 5, pp. 745-

758.

Wen, W., Chen, Y.H., Chen, I.C., 2008. A knowledge-

based decision support system for measuring

enterprise performance, Knowledge-based systems,

Vol. 21, pp. 148-163.

Yurdakul, M., 2002. Measuring a manufacturing system’s

performance using Saaty’s system with feedback

approach, Integrated manufacturing systems, Vol. 13

No. 1, pp. 25-34.

ICEIS 2008 - International Conference on Enterprise Information Systems

192