0 views

Original Title: 1-s2.0-S1474667016332700-main

Uploaded by Đạt Thành Võ

- math 1010 project
- 10.11648.j.ajbes.20150102.12
- Assignment and Transportation
- Mb0048 Unit 02-Slm
- aubrey 1010 class project 2
- Redesign of Marine Structures by Perturbation
- adcampaignprojectcombined
- max advertising
- optimizing an ad campaign math 1010 newlinear programming project 3
- Sw August 2012 Article
- project 1
- 1010 optimization project f 16-2
- linear programming project optimizing an advertising campaign
- Multi-site Aggregate Production Planning With
- math1010project2014
- 1010 Optimization Project
- 1010 optimization project 1
- Second Sem Syllabus
- sunday
- project and reflection1

You are on page 1of 6

Bucharest, Romania, May 23-25, 2012

Lucia Pascale*, Marin Mainea*,

Paul Ciprian Patic*, Luminita Duta*

*Electrical Engineering Faculty, ValahiaUniversity of Targoviste, 24 Unirii Ave., 130065

ROMANIA (e-mail: lucia.p2005@gmail.com, marin_mainea@yahoo.com, patic@valahia.ro, duta@valahia.ro)

Abstract: The maintenance department has a key role so as to ensure the continuous production flow by

up-keeping the equipment at its normal functioning state. Total Productive Maintenance is a concept that

involves everyone in the company. Overall Equipment Effectiveness is the key measure of the tangible

benefits obtainable from the TPM strategy. This paper proposes a decision aid method based on a

mathematical model to improve the OEE indicator and the overall performance of the manufacturing line.

Keywords: Maintenance, line balancing, decision analysis, decision modeling

The goal of the TPM program is to increase production while

1. INTRODUCTION

achieving zero defects, zero breakdowns and zero accidents,

Total Productive Maintenance (TPM) was developed in Japan in the same time increasing job satisfaction of the worker.

at the beginning of the seventh decade of the last century by

The targets of TPM are to: a) maintain an accident free

the Japanese Institute of Plant Maintenance (JIPM, 2002).

environment, b) enhance the machine efficiency towards

This concept challenges the view that maintenance is no more

100%, c) reduce manufacturing wastes and costs, d) deliver

than a function that operates in the background and only

the goods in time to clients, and e) develop multi-skilled

appears when needed. The objective of TPM is to engender a

workers.

sense of joint responsibility between supervision, operators

and maintenance workers, not simply to keep machines The five principles of TPM are (McCarthy and Rich, 2004):

running smoothly, but also to extend and optimise their

1 Adopting improvement activities designed to increase the

global performance (Venkatesh, 2006).

overall equipment effectiveness;

TPM concept brings maintenance into focus as a necessary 2 Improving existing planned and predictive maintenance

part of the enterprise policy. TPM is no longer regarded as a systems;

non-profit activity. Systematically maintenance is scheduled 3 Establishing a level of self-maintenance and cleaning

as a part of the manufacturing day work and is an integral carried out by highly trained operators;

part of the manufacturing process. The goal is to hold 4 Increasing the skills and motivation of operators and

emergency and unscheduled maintenance to a minimum engineers by individual and group development;

level. 5 Applying early management techniques to create reliable

and safe equipment and processes;

This paper is organized as it follows: first a brief presentation

of the TPM concept is done. Principles and characteristics of These principles are supported by seven corresponding

the TPM are presented. In the next section, calculus of some planned maintenance steps meant to guide the standardization

important TPM indicators is given, emphasizing the role of and simplification of maintenance activities (Tajiri and

managing their values during the manufacturing process. A Gotoh, 1999). This stepwise process has the effect of raising

mathematical decision model is developed in the third section the capability of production, maintenance and supervision

for a manufacturing line. By minimising the objective and of releasing specialist resources to focus on the next

function from the proposed model, new improved values of development stage. That is process optimization, a key part of

the TPM indicators are obtained. The results will show that a the active maintainer’s role once breakdowns are brought

well-balanced line provides a maximum value of the Overall under control. The designer of an industrial process has to

Equipment Effectiveness (OEE), one of the most important take into account in his activity the prerequisites demands of

indicators of the TPM strategy. A case study is presented in the maintenance steps (Patic and Pascale, 2010).

the last section.

TPM recognizes the importance of operator involvement

2. TOTAL PRODUCTIVE MAINTENANCE developing a close relationship between him and his machine.

This relation gave birth to the autonomous maintenance

Total Productive Maintenance is a maintenance strategy that which provides the solution for raising equipment condition

involves people in all levels of organisation. It can be to a level where zero breakdowns is possible. TPM tools

considered as the medical science for machine functionality. improve the effectiveness of the transformation process.

INCOM 2012, May 23-25, 2012

Bucharest, Romania

The method most commonly used in Lean Manufacturing to 4. MATHEMATICAL DECISION MODEL

measure the productivity of machinery is the calculus of OEE

Mathematical models with decision variables are often use in

performance indicator.

problems of tasks assignment. The idea is to achieve an

optimal assignment of task on stations so as to minimize the

3. OVERALL EQUIPMENT EFFECTIVENESS

real cycle time of the line.

Overall Equipment Effectiveness (OEE) is used as an

In the next section the following notations are used:

indicator of how well equipment is used in batch production.

This parameter is obtained in relation to losses that can m number of workstations

impede the equipment efficiency. OEE is a quantitative

indicator. It is the product of the availability, performance n number of tasks

and quality of the equipment (Venkatesh, 2006). j index of a task

i index of a workstation

OEE% Availability% Performance% Quality%

tcy cycle time

Availability measures the percentage of time the machine

runs for production. It could also include a comparison of F the objective function

available size. If a machine is available for production, but it X tasks possible assignment matrix

is stopped for an unscheduled maintenance repair, OEE is Y tasks effective assignment matrix

reduced. The machine is available to run, but it can’t until the

T[n] operational times array

maintenance operations are finished.

Wi workstation i

In fact, availability indicates the performance of the

Maintenance Department. Its value is can be calculated using X is the assignment matrix and defines the possible off-line

equation (1). assignment of the tasks to workstations as it follows:

actual run time X (i, j )

Availability% 100 0 elsewhere

planned run time (1)

1 i m

Where 1 j n

actual run time=planned run time -idle time One could note that X(i,j) is a decision variable.

Performance measures the production rate compared to rated In equation (3) given above, if the real cycle time is

speed. Performance indicates the operator’s performances minimized, then the speed operating rate increases, so the

and the way the equipment is used: value of the OEE goes towards a maximal value. The

augmentation of the speed operating rate value is possible

Performance% speed operating rate net operating rate 100 only if the cycle time of the line is minimized, otherwise this

(2) aim can lead to bottlenecks. Hence, the idea is to balance the

line and to minimize the real cycle time so as to maximize the

Where OEE indicator and the efficiency of the line.

The objective function of present optimization problem is:

standard cycle time

speed operating rate 100 (3)

real cycle time F (t ) min (real cycle time)

(6)

number of products real cycle time In (Duta, 2006a; Duta et al, 2008), the cycle time is the

net operating rate 100 (4) maximal value of the sum of operational times:

actual run time

Quality is given by the percentage of good product produced.

Quality indicates the capability of the process:

tcy max

Wi

j( tasks on Wi )

tij (7)

quantity produced right function becomes:

Quality % 100 (5)

X (i, j) t

quantity produced

F (t ) min(max ij

) (8)

Wi

j( tasks on Wi ) i

935

INCOM 2012, May 23-25, 2012

Bucharest, Romania

Some terms in brackets before could be 0 depending on the The line is supposed to work in continuous flow, operational

value of the decision variable X(i,j). times are deterministic and production is of flow shop type.

Another assumption is that the line is never starving: there

The minimum value of this function is obtained when the

are always products to manufacture on the line.

manufacturing line is balanced (Duta et al, 2006b). The

effective assignment matrix Y is derived from matrix X and is

obtained after simulation running.

The optimisation problem consists in calculating the minimal

cycle time from the objective function F (equation (8)) taking

into account the four linear constraints below. The input size

of the problem is the length of a binary data representation.

a The non-divisibility constraint : Every task is performed on

a single workstation. A task cannot be divided between

workstations.

m (9)

Y (i, j) 1

i 1

j 1..n

Fig. 1 Input Excel file

In our study m=5, n=10.

c. Time constraint:cycle time is an upper bound on the The possible assignment matrix is taken from (Mainea et al,

workload assigned to each workstation. 2010) and it fulfills the precedence constraint:

11 0 1 0 0 11 0 0

n (11)

(t ij Y (i, j )) t cy i 1..m 0 1111 0 0 1 0 0

j 1 X 1 0 11 0 1 0 11 0

d. Real assignment constraint

0 0 111 0 1111

0 0 0 1 0 11 0 11

Y (i, j ) {0,1} i 1..m, j 1..n (12) Operational times (in minutes) are given bellow:

Task 1 2 3 4 5 6 7 8 9 10

T[n] 38 10 8 26 13 18 25 9 15 30

5. CASE STUDY

Minimum standard cycle time is also given and is equal to

The proposed method was implemented at the Romanian

38.4 min.

enterprise ARCTIC Gaesti, a domestic appliances

manufacturer. 1 The proposed solving method is to apply integer

programming on a linear function (8), with linear constraints

Here, the TPM concept was successfully implemented since

(9), (10), (11), (12). To run simulations, the FICO Xpress

2004. This aim was accomplished with the help of a Japanese

Optimization Suite 7 Software was utilized (FICO Xpress,

consultant. The system’s concept is “zero losses” and it

2009). This is a mixed integer programming optimizer which

correlates the augmentation of the production capacity with

has been programmed to handle a broad range of

the costing diminution (Mainea, 2006). The values of the

optimization problems. His quadratic module allows the

necessary parameters to calculate the OEE are registered on

optimization of a quadratic function. The main advantage of

line in Excel files as the one presented in figure 1. Parameters

this software product is that the user works in the Console

are considered input data for our optimization method. The

Mode and can modify the code of the program to suit data of

line has five workstations performing ten different

the problem.

operations, whose operational times are given below.

The XPRESS optimizer uses Branch and Bound technique to

1

Acknowledgements to Romanian enterprise for information solve mixed integer programming problems.

availability

936

INCOM 2012, May 23-25, 2012

Bucharest, Romania

The relaxed problem is a linear programming problem and Table 3 Tasks assignment after balancing

can be solved by exploring the tree of solutions using the cut-

off value method. When a better value of the solution is found Wi 1 2 3 4 5

in a solution node, this can act as a cut-off for outstanding

nodes (Dash, 2007). Simulations performed with FICO Task 1 2, 4 3, 6, 8 5, 10 7, 9

XPRESS optimizer give the number of iterations and cuts

made in the solution space.

This effective assignment fulfills constraints from (9) to (12).

After simulations, the minimum value of the real cycle time

After balancing the line, new values for the workstation’s

is obtained: t cy

43 min. operational times are obtained (Table 4):

The obtained effective assignment matrix is:

Table 4 Operational times after balancing the line

1 0 0 0 0 0 0 0 0 0 Wi 1 2 3 4 5

0 1 0 1 0 0 0 0 0 0 38 36 35 43 40

Y 0 0 1 0 0 1 0 1 0 0

0 0 0 0 1 0 0 0 0 1 One can notice that the line is better balanced and the cycle

0 0 0 0 0 0 1 0 1 0

time is minimized.

Simulations were performed on an Intel Core 2Duo T7500 Replacing the values of standard cycle time and real cycle

2.2 GHz Processor with a RAM of 2 GB. time in equations (1) to (4), one can calculate the new values

of availability and performance for each workstation (Table 5

6. RESULTS and Table 6).

In Table 1data before simulation is presented. Table 5 Workstations availabilities after balancing

are taken from the Excel file (fig. 1)

A[%] 86,84 80,55 77,14 100 92,5

Performance

Availability

Quality

OEE

Wi 1 2 3 4 5

19,37 69,22 100,00 13,41

80,72 86,43 92,92 64,82 Further, these results were used to calculate the new values of

the OEE indicator. Final results are presented in Table 7.

78,86 74,92 92,61 54,71

97,49 90,00 97,53 85,57 Table 7 New improved values of the TPM indicators [%]

31,57 65,00 100,00 20,52

Performance

Availability

Quality

OEE

Table 2 Initial operational times for each station (min)

Wi 1 2 3 4 5

86,84 88,37 100,00 76,74

38 44 38 33 39

80,55 83,72 92,92 62,66

In the previous paragraph, the value of the real cycle time 77,14 81,39 92,61 58,14

was obtained after minimizing the value of the function F

100 95,44 97,53 93,08

from the equation (8). The effective assignment of the tasks

after simulations is given by matrix Y and also in Table 3. 92,5 93 100,00 86,03

937

INCOM 2012, May 23-25, 2012

Bucharest, Romania

One could notice that, after applying the proposed method, Figure 4 is a print screen of TPM indicators displayed in real

OEE and machine performances are maximized and are time by Doruk system2. Different OEE values are shown on

almost equally distributed on the line (Fig. 2 and 3). That the screen. They are displayed in the same time on

means that workstations are equally loaded, thus increasing computer’s display and on the monitoring panels from the

the efficiency of the line. workshop walls.

7. CONCLUSIONS

TPM is a new manufacturing approach implemented in

Romania and the results in increasing the productivity are

over expectations.

An optimization method for the Overall Equipment

Effectiveness maximization, the most important indicator of

the Total Productive Maintenance, is presented in this paper.

The method is based on the line balancing accomplishment

after minimizing the real cycle time using a linear

programming model. The originality of the approach is the

presence of decision variables in the objective function as

well as in model constraints. After performing simulations on

the mathematical model, TPM indicators values are

Fig. 3 TPM indicators after balancing the line

significantly improved.

Results of the above optimization methodology are visualized

The results were implemented at ARCTIC Gaesti enterprise

on the monitoring system and presented on screens like that

which is in present in the third stage of the TPM

from figure 4.

implementation concept.

A computerized system monitors the work on line and

Further research will take into account the multicriterial

displays the values of OEE for each workstation in real time.

character of TPM indicators optimization.

Integrating our method in the informational system gives the

possibility to control tasks assignment on workstations and to

increase OEE values to almost 100% which is the maximal

performance of the equipment.

2

Doruk Automation Company, Turkey is an industrial

equipment supplier

938

INCOM 2012, May 23-25, 2012

Bucharest, Romania

REFERENCES

Optimization Ltd., Canada

Duta, L. (2006a) Introduction dans l’étude des systèmes de

désassemblage – Programmation et contrôle, Expert

Publishing House Bucharest (in French language)

Duta L, Filip F G., Henrioud J.M. (2006b), Applying Equal

Piles Approach to Disassembly Line Balancing Problem,

ELSEVIER, Vol 16, pp. 152-157

Duta L, Filip F. G., Caciula I. (2008): Real Time Balancing

of Complex Disassembly Lines, Proceedings of the 17th

IFAC World Congress, ELSEVIER, Volume# 17 | Part#

1, pp 913-91

JIPM (2002), TPM Encyclopaedia, Japan Institute of Plant

Maintenance,USA

Mainea M., (2006) Quality Management. Design,

implementation and development of the quality system in

industrial enterprises, PhD Thesis, Polytechnic

University of Bucharest

Mainea M., Duta L, Patic P. C., Caciula I., (2010) A Method

to Optimize the Overall Equipment Effectiveness,

Manufacturing Control, Production and Logistics

Conference MCPL 2010, Coimbra, Portugal

McCarthy D, Rich N, (2004), Lean TPM – a Blueprint for

Change, Elsevier Butterworth-Heinemann

Patic P.C., Pascale L., (2010) CAD in Industrial Application,

Bibliotheca Publishing House, Romania

Rich, N. (2002) Turning Japanese, PhD Thesis, Cardiff

University

Tajiri M., Gotoh F. (1999), Autonomous Maintenance in

seven steps, Productivity Inc., USA

Venkatesh J (2006)A Introduction to Total Productive

Maintenance (TPM),Plant Maintenance Resource

Center.

FICO Xpress Optimization Suite 7, (2009)

939

- math 1010 projectUploaded byapi-240285476
- 10.11648.j.ajbes.20150102.12Uploaded byIsmat Naqvi
- Assignment and TransportationUploaded byrajkd1uw
- Mb0048 Unit 02-SlmUploaded byNeetu Ahmed
- aubrey 1010 class project 2Uploaded byapi-253606257
- Redesign of Marine Structures by PerturbationUploaded byvinaysa181810
- adcampaignprojectcombinedUploaded byapi-236056742
- max advertisingUploaded byapi-316772829
- optimizing an ad campaign math 1010 newlinear programming project 3Uploaded byapi-316640542
- Sw August 2012 ArticleUploaded byVarun Mehrotra
- project 1Uploaded byapi-233325571
- 1010 optimization project f 16-2Uploaded byapi-315753172
- linear programming project optimizing an advertising campaignUploaded byapi-242859609
- Multi-site Aggregate Production Planning WithUploaded byxandercage
- math1010project2014Uploaded byapi-241533045
- 1010 Optimization ProjectUploaded byLacey
- 1010 optimization project 1Uploaded byapi-356225918
- Second Sem SyllabusUploaded byShorabh Sharma
- sundayUploaded bydocumentary123
- project and reflection1Uploaded byapi-315204993
- math 1010 projectUploaded byapi-313254091
- 1010 optimization project repairedUploaded byapi-253464342
- project 1Uploaded byapi-285455289
- adam math 1010 projectUploaded byapi-301993273
- math 1010 linear programming projectUploaded byapi-241824441
- EnTDDUaliteUploaded bygabboudeh
- 1010 optimization projectUploaded byapi-325533351
- math 1010 projectUploaded byapi-242502418
- 1010 optimization projectUploaded byapi-325533351
- Quantatative Methods SylUploaded byNguyên Bùi

- Online Learning in Virtual Environments with SLOODLEUploaded bySLOODLE
- Subsea development systemUploaded byM.c. Tim Lim
- Six Gun ShootoutUploaded byremow
- Dynamic Systems FractalsUploaded byRafael Thomas
- Super Passage Trough Design-user ManualUploaded bypaaji
- Clinical Laboratory Management - OrganizationUploaded byMaryam Zainal
- Conducting a Needs Assessment FinalUploaded byBhavika Jain
- Performance Analysis of Drying of Green Olive in a Tray DryerUploaded byFiraz Rizaldy
- marketing strategy of hero hondaUploaded bykumarsah
- PrintStream (Java Platform SE 7 )Uploaded bypcdproyecto
- Sun Tzu’s Art of War for Project Delivery SuccessUploaded bysanazh
- Fact Sheet - Big BirdUploaded byNick Numlk
- OneDirect ProfileUploaded byishaan90
- Solution With SemaphoreUploaded bykrupanick24
- sn74ls240n_OctalBufferLineDriverUploaded bysas999333
- AN-C240Uploaded bykymk21
- 33A Syllabus and ScheduleUploaded bythatscribdacnt
- Enrollment System Policy- AtanozaIMRADUploaded byVe Leslie Atanoza
- Leopoldal2010_11ICRSproceedingsUploaded byMohd. Yunus
- Nikola TeslaUploaded byReșetnic Dragoș
- NeverstromUploaded byAnz Joy
- Fiber Optic PlantUploaded byashish20705
- ANALOGValue ProcessingUploaded byalisuse
- Perl Tutorial for BeginnersUploaded byAbhilash V Pillai
- Trx Oil Test PropertiesUploaded bybibhoo
- Owners Man HT R590 en WebUploaded bydarianhawke
- wasteland bestiary (pdf).pdfUploaded byFrancisco
- irf8010Uploaded bywilly190486
- PRV Test ProcedureUploaded byNatraji
- EBS Application Object LibraryUploaded byPreethi Kishore