Modeling and Simulation of Manufacturing Line Improvement

ISSN (e): 2250 – 3005 || Volume, 06 || Issue, 10|| October – 2016 ||
International Journal of Computational Engineering Research (IJCER)
www.ijceronline.com Open Access Journal Page 26
Modeling and Simulation of Manufacturing Line Improvement
Grzegorz Gołda1
, Adrian Kampa1
, Iwona Paprocka1
1
Silesian University of Technology ul. Konarskiego 18A, 44-100 Gliwice, Poland,
I. Introduction
Nowadays, one can notice increasing use of automation and robotization, which replace human labor. Industrial
robots are used especially for repetitive, high precision tasks (e.g. welding), monotonous activities and actions
demanding physical exertion. Industrial robots have mobility similar to human arm, and can perform various
complex actions like a human. In addition, they do not get tired and bored. It is estimated that using
robotization, many companies obtained the reduction of the production cost by 50%, the increase of productivity
by 30% and the increase of utilization by more than 85% [4]. However, the introduction of robotization is
related with high costs, thus is profitable only in certain circumstances, including, a high level of production,
work with repetitive and precision tasks with ensuring the safety and health at work [12]. Such conditions occur
in the automotive industry [10]. The problem is how to determine a real difference in work efficiency between
human and robot. The aim of the study is to develop a methodology, which allows to clearly define the
throughput growth associated with the replacement of human labor by industrial robots. In order to assess the
effectiveness of the robot application, we compare production uptime of humans and robots with the use of the
OEE indicator (Overall Equipment Effectiveness) to calculate work efficiency and we use Discrete Event
Simulation (DES) for verification. Production systems can be very complex and difficult for analysis. Therefore,
the DES method is widely used for design of manufacturing systems [15], for solving scheduling problems [9]
and for efficiency [8] and stability analysis [3] of production systems. There are some DES software tools
dedicated for production processes simulation, including ARENA, Enterprise Dynamics, FlexSim, Plant
Simulation and others [5, 11, 13].
II. Manufacturing line example
Analyzed example of manufacturing line consists of five presses and is used for production of car body parts.
Presses are often used in various production processes e.g. pressing, sheet metal forming etc. The schema of
press line is presented in Figure 1. The line consist of five serially linked machines, input and output stations
and operators or robots for machine tending.
Figure 1. Schema of the manufacturing line
ABSTRACT
In the paper an example of manufacturing line form automotive industry is analyzed. The line
consists of five presses, that can be operated by humans or industrial robots. Since human factors
affect destabilization of the production processes, robots are preferred to apply. The problem is how
to determine the real difference in the work efficiency between human and robot. Because of problem
complexity, Discrete Events Simulation (DES) method have been used. Three models have been
developed including manufacturing line before and after robotization and taking into account
stochastic parameters of availability and reliability of the machines, operators and robots. Analysis of
the production efficiency of the press shop lines operated by human operators or industrial robots are
presented. We apply OEE (Overall Equipment Effectiveness) indicator to present how the availability
and reliability parameters influence over the performance of the workstations, in particular in the
short and long time period.
Keywords: DES (Discrete Event Simulation), human factors, industrial robots, OEE – overall
equipment efficiency, machine reliability

Usually, operator is required for loading and unloading a machine and for transferring a product from one
machine to a next production stage. Robots can make that work faster and more regular then human operators,
but what will be the difference in production throughput?
The reliability of each component play important role, because failure of one element causes production
stopping on the whole line. The machine failures are mostly random and are difficult to predict; therefore, we
have used computer simulation for the research. We have developed models including human resources and
industrial robots. Because elaborated model is a simplified image of the real system, the next problem is which
system parameters are important for the simulation.
III. Work efficiency and OEE
Work efficiency and the use of the production means can be expressed by using the OEE metric that depend on
three factors: availability, performance and quality [6].
OEE=(Availability)x(Performance)x(Quality) (1)
Availability is the ratio of the time spent on the realization of a task to the scheduled time. Availability is
reduced by disruptions at work and machine failures.
(2)
Performance is the ratio of the time to complete a task under ideal conditions compared to the realization in real
conditions or the ratio of the products obtained in reality to the number of possible products to obtain under
ideal conditions. Performance is reduced (loss of working speed) by the occurrence of any disturbances e.g.
human errors.
(3)
Quality is expressed by the ratio of the number of good products and the total number of products.
(4)
The number of good quality products is a random variable, which can be described by a normal distribution with
standard deviation sigma. Quality levels are determined for ranges of the standard deviation sigma. In traditional
production systems, the level of 3 sigmas is considered to be sufficient. However, in the modern automated and
robotic systems the level of 5-6 sigmas is possible to achieve [2].
In order to make these models more realistic we decided to include availability, performance and quality
parameters into account.
3.1. Availability and failures
The term of availability contains planned work time and unplanned events e.g. the disturbances at work and
random machine failures. Any unplanned event causes that machines are unavailable and work efficiency
decreases. The reliability of objects such as machines or robots is defined as the probability that they will work
correctly for a given time under defined conditions of work. The most popular method for estimating reliability
parameters uses theory of probability to forecast a value of failure-free time and repair time parameters, under
the condition that a trend based on historical value of the parameter is possible to notice. The examples of using
normal, exponential, triangular distributions to describe both failure and repair times are described in [4]. In
practice, for description of reliability, in most cases the parameter MTTF (mean time to failure) is used, which is
the expected value of exponentially distributed random variable with failure rate λ [14].
(5)
In the case of repairable objects the parameters MTBF (mean time between failures), and the MTTR (mean time
to repair) are used.
(6)
For complex systems, consisting of n serially linked objects, the resultant failure rate λs of the system is the sum
of the failure rates of each element λi:

(7)
Alternatively, the system MTBFs is the sum of the MTBFi inverse:
(8)
For the example of robotic line, presented in figure 1, we can use formula 8 with different failure parameters for
machines and for robots :
(9)
Machinery failures affect the availability of means of production and may cause severe disturbances in
production processes. The average availability can be calculated with formula 10.
(10)
3.2. Human factors
Humans are one of the most unreliable parts of manufacturing systems. In the case of a manually operated
systems, a number of human factors (human errors) can lead to destabilization of the manufacturing process [7].
Workers also require a break for rest. They are unpredictable and can make errors and accidents at work can
occur. Sick and absent workers can cause the failure of a production plans.
In the computer software, used for production processes simulation, the human factor is not sufficiently
modelled. People are treated as quasi-technical elements of production system and they should operate in the
same way as a machine. In practice, the human behaviour is unpredictable, thus it might help to explain why
simulation models do not respond to the reality as it would be expected [1].
IV. Modelling of manufacturing line
In order to analyze the presented problem the mechanical press line from automotive enterprise, has been taken
into account. We used the Enterprise Dynamics software, which allows computer-modeling and simulation of
discrete production processes with the use of human resources as well as robots. Parameters of typical
manufacturing process were taken into account including constant machine cycle time 5s, constant robot speed
180°/s, stochastic parameters of operator time, which was described by normal distribution with 10s mean time
and 2s standard deviation. Availability parameters include planned setup time 15 minutes per shift, and break
for rest for workers 15 minutes per shift. Failure parameters were taken into account including short time human
errors rate and long time machine and robot MTBF. Typical reliability parameters include MTBFm=500 hours
and MTTRm=4 hours for machines and MTBFr=1000 hours and MTTRr=4 hours for robots. Also normal
quality distribution of good and bad product has been included into the model.
The first model presented in the Figure 1, is a simply reference model. It consist of five machines, input and
output station. It represents production in ideal conditions and can be used for calculation of maximal
production limit.
Figure 2. Reference model of manufacturing line

The second model shown in the Figure 3, includes five human operators, six buffers, and output for good and
bad quality products. Work time schedule is used for defining planned work pauses and additional MTBF
objects are used for simulation of random failures.
Figure 3. Model of manually operated line after 8 hours of simulation
Irregular work of operators and human errors cause disturbances in the production process. Assuming human
unreliability on the basis of HEART (Human Error Assessment and Reduction Technique) for “routine and
highly practiced rapid tasks involving relatively low level of skill”, the nominal value of human error equals to
0.01 [16]. Therefore, human errors rate can be described by parameters: MTBFh=8 hours and MTTRh=5
minutes.
We assume that other employee can replace sick and absent worker, but it is impossible to replace broken
machines and robots, and they require repairing.
The third model presented in the Figure 4, includes industrial robots. Because of the constant speed and very
regular work of robots, buffers are not needed.
Figure 4. Model of robotized line after 8 hours of simulation
We can see a great improvement of production throughput and machine utilization for robotic line, however
each simulation run, can give little different results because of random failures. Therefore, we have made series
of experiments, containing different number of simulations runs and simulation time from 8 hours to 6000
hours.
V. Experiment results
The production value P obtained from a single simulation run is a random variable that consists of several
parameters. The random nature of failures causes a significant dispersion of obtained values and relatively large
standard deviation. The average production value Pavg of simulation experiments are summarized in Table 1.
Each experiment consists of fifty samples (simulation runs). The value MaxLimit determines the maximum
possible production volume in a given period of time under ideal working conditions and for a machine cycle
time (Tm=5 seconds). Since the model was build based on the OEE components, and contain parameters of
availability, performance and quality, the production value from simulation can be directly used to calculate the
OEE indicator.
(11)
The standard deviation shows the differences between the average value of production and the value of
production achieved in each simulation run. For the robotic tended line, the values of standard deviation are
greater because of a much greater production volume and possibility of robots failures. This phenomenon can be
explained that sick humans can be replaced but robots not.

Table 1. Simulation results for manually operated and robotic lines (average production value Pavg in [PCs.] for
50 runs of simulation, α=0,95, MTBFm=500h, MTBFr=1000h, MTTR=4h)
Human Robots Human Robots Human Robots Human Robots
Operators Operators Operators Operators
Time 8h 24h 2000h 6000h
Max Limit
[PCs.]
5760 17280 1440000 4320000
Average
Production
Pavg [PCs]
1681 4404 5111 13300 427179 1094949 1281345 3279888
Standard
deviation
[PCs]
50.72 619.3 85.28 1432 714.96 19331 1207 27496
OEE 0,2918 0,7646 0,2958 0,7697 0,2967 0,7604 0,2966 0,7592
Production throughput of the robotic line has increased about 2.6 times comparing to the line before
robotization.The OEE related performance of the production line operated by robots has improved by 46%
comparing to the manually operated line. The OEE indicator equals to OOEh=29.18÷29,67% for humans and
OEEr=75,92÷76,97% for robots, and correspond with the values assigned by the theory presented in Table 2.
Values calculated by the theory are: availability of the whole robotic system A=0,912; performance P=0,8333;
quality Q=0,99999. That gives OEEr=76,0%.
Table 2. Values of the OEE indicator
Manufacturing line Human operated line OOEh Robotic line OEEr
Availability 0,893 0,912
Performance 0,333 0,833
Quality 0,9973 0,99999
OEE 0,297 0,760
This shows that reliability parameters have significant influence on the productivity of the production system
and reliability improvement can change the OEE score. Comparing the OEE factors achieved for human
operator and robot the greatest improvement is in the performance.
VI. Conclusion
As it was predicted, the experiments confirm the advantage of the application of robotic operated production
lines comparing to manually operated lines. The computer simulation of the simplified model of the production
line with machines, operators and robots with stochastic (short-time and long-time) reliability parameters allows
for better representation and understanding of a real production process. This is particularly to see in the case of
work in three shifts a day for a long period of time. The work organization and robots synchronization play
important role and therefore the efficiency of the production line operated by robots has improved the OEE
indicator by 46% comparing to a manually operated line. Because of irregular work of human operators, the
buffers (queue) are needed for equalization of production flow and therefore loading (unloading) products from
buffers result in low performance of human operators. Also breaks for rest result in lower OEE value. This is
one of the best examples of robotic improvement in manufacturing. However, in other cases of machine tending,
the difference between human operator and robot is not so clearly to see even for long time simulations.
The use of OEE factors allows to compare results from other manufacturing systems. The reality is that most
manufacturing companies have OEE scores closer to 60%, but there are many companies with OEE scores
lower than 40%, and small number of world-class companies that have OEE scores higher than 80%.
There are some place for improvement of availability, performance and quality. Availability depends on planned
and unplanned breaks at work. Performance score depends on machine cycle time and high robot speed. The
quality depends on stability of manufacturing process parameters.
Results obtained with presented methodology can be used for detailed designing of a robotic system and for
economic analysis, regarding labor costs and costs associated with the investments in robotization.
References
[1] T. Baines, S. Mason, P. O. Siebers, J. Ladbrook. Humans: the missing link in manufacturing simulation? Simulation Modelling
Practice and Theory, No 12, 2004, pp. 515–526.
[2] M. Barney, T. McCarty. The new Six Sigma. Prentice Hall Professional, New York, 2001.
[3] A. Burduk. Stability Analysis of the Production System Using Simulation Models”, Process Simulation and Optimization in
Sustainable Logistics and Manufacturing, Springer, 2014, pp. 69-83.
[4] A Glaser. Industrial robots, Industrial Press, New York, 2009.
[5] G. Gołda, A. Kampa, I. Paprocka. Simulation model of robotic manufacturing line. Position Papers of the 2016 Federated
Conference on Computer Science and Information Systems, M. Ganzha, L. Maciaszek, M. Paprzycki (eds). ACSIS, Vol. 9, pages
107–113 (2016), http://dx.doi.org/10.15439/2016F307

[6] R. C. Hansen. Overall Equipment Effectiveness, Industrial Press, 2005.
[7] C. E. Harriott, J. A. Adams. Modeling Human Performance for Human–Robot Systems, Reviews of Human Factors and
Ergonomics, No 9, 94, 2013, http://rev.sagepub.com/content/9/1/94
[8] A. Ingemansson, and G.S. Bolmsjo. Improved efficiency with production disturbance reduction in manufacturing systems based on
discrete-event simulation’, Journal of Manufacturing Technology Management, Vol. 15, No. 3. 2004, pp. 267–279.
[9] A. Kampa. Planning and scheduling of work in robotic manufacturing systems with flexible production. Ed. B. Skołud: Hueristic
Methods of Project and Production Scheduling. Journal of Machine Engineering, Vol. 12, No. 3, 2012, pp. 34-44.
[10] R. Kulanek. Press Excelence Center COMAU, i.e. robotics press line at the world level”, available at www.robotyka.com 2014, (in
polish, accessed 15.12.2015).
[11] B. Santhosh Kumar , V. Mahesh, B. Satish Kumar. Modeling and Analysis of Flexible Manufacturing System with FlexSim.
International Journal of Computational Engineering Research (IJCER), Volume 05, Issue 10, October – 2015, pp. 1-6.
[12] S. Y. Nof. Handbook of industrial robotics. John Wiley & Sons, New York, 1999.
[13] I. Paprocka, W. Kempa, K. Kalinowski, C. Grabowik: “Estimation of overall equipment effectiveness using simulation
programme”. Modern technologies in industrial engineering (ModTech2015), 17-20 June 2015, Mamaia, Romania. Materials
Science and Engineering ; vol. 95 1757-8981. Institute of Physics Publishing, Bristol, 2015, s. 1-6, doi:10.1088/1757-
899X/95/1/012155
[14] D.J. Smith. Reliability, Maintainability and Risk. Practical methods for engineers, Elsevier, Oxford, 2005.
[15] C. Xie, T. T. Allen. Simulation and experimental design methods for job shop scheduling with material handling: a survey. The
International Journal of Advanced Manufacturing Technology, 2015, p. 233-243.
[16] D.D. Woods. Modeling and predicting human error”, In J. Elkind, S. Card, J. Hochberg, and B. Huey (Eds.), Human performance
models for computer-aided engineering Academic Press, 1990, pp. 248-274.

Modeling and Simulation of Manufacturing Line Improvement

More Related Content

What's hot (6)

Viewers also liked (19)

Similar to Modeling and Simulation of Manufacturing Line Improvement (20)

Recently uploaded (20)

Modeling and Simulation of Manufacturing Line Improvement