simulation of production lines in the …mper.org/mper/images/archiwum/2014/nr4/7-rostkowska.pdfand...

Management and Production Engineering Review

Volume 5 • Number 4 • December 2014 • pp. 53–65DOI: 10.2478/mper-2014-0036

SIMULATION OF PRODUCTION LINES IN THE EDUCATION

OF ENGINEERS: HOW TO CHOOSE THE RIGHT SOFTWARE?

Marta Rostkowska

Poznań University of Technology, Institute of Control and Information Engineering, Poland

Corresponding author:

Marta Rostkowska

Poznań University of Technology

Institute of Control and Information Engineering

Piotrowo 3A, 60-965 Poznań, Poland

phone: (+48) 609-623-065

e-mail: [email protected]

Received: 3 September 2014 Abstract

Accepted: 3 October 2014 The article discusses the problems of modeling and simulation in the design of productionlines, mainly from an educator’s perspective. Nowadays, there is a wide range of computerprograms that can be used to design production lines and to simulate various aspects oftheir operation. However, the programs being available vary considerably as to their func-tionality, the approach to production system design, and the visualization tools. Thereforewe demonstrate and evaluate in this paper four simulation programs, focusing on the eas-iness of system design, area of the possible applications in education of engineers, and thelimitations imposed by versions dedicated for students. We evaluate three of programs fordigital factory simulation on a common, simple assembly task, then demonstrate that theseprograms may be also used for more specialized simulations in various areas of production,and compare with a specialized program for simulation of robotised production lines andwork cells.

Keywords

simulation, factory automation, production line, simulation approaches, simulation appli-cations, 3D animation.

Introduction

In order to meet the demand for engineers ed-ucated in modern factory automation, universitiesprovide the students with courses on flexible manu-facturing and production automation. However, re-al production lines are often very complicated sys-tems of classic and/or numerically controlled ma-chine tools, assembly stations, various types of ro-bots, conveyors, and other means of transportation.Designing properly such a system requires knowledgein several areas, and some experience, which is hardto obtain for a student outside of a real productionplant.The industry wants to improve flexibility, com-

petitiveness and reduce costs by cutting down theneed of producing prototypes and test series [1].Production is planned in details with regard

to the efficient management of resources. Planningprocess is implemented by designating short-term

and long-term tasks. Simulation allows to check thebehavior of the system after the implementation ofthe proposed solution and the removal of all existingerrors. The use of 3D visualization allows to moreaccurately understand the process and explains theambiguities. Computer simulation of production lineenables testing of the designed solution without for-mer physical implementation. Digital factory allowsthe integration of devices, people and machines thatallows efficient use of factory resources and processlines [2, 3]. Simulations addressed the effective man-agement of resources [4], including the following is-sues:

• determining the effect of changes in parametersand properties in the course of the simulation,

• flexibility and capacity of the system,• bottlenecks,• number of elements of the production system andthe types of machines,

• the number of staff needed,

53


• performance of work,• delays in delivery, quality problems, equipmentfailures,

• reduction of losses and downtime with maximumsynchronization and coordination of all processesrelated to movement of materials and productionprocesses.The increasing usage of simulation in the process

of modernization and planning of production linesin factories caused a change in the education in thisfield.At the university, the students may gain this ex-

perience in two ways. One solution involves the useof scaled-down, small production lines mounted atteaching sites. Examples of such lines are those pro-duced by Koester [5] and Micro [6]. Although ap-pealing to the students, this solution is very expen-sive and allows testing of a limited number of con-figurations and component types. The other solutionis to use software simulators. With the simulations,various configurations of the equipment on the pro-duction line can be tested easily, production statis-tics can be carried out (costs and benefits) and thesolution optimal with respect to the structure andcomponents of the considered production line can bechosen.On the software market, a large number of com-

mercial programs for simulation of various aspectsof the operation of production plants and/or linesis available. However, it is unclear how to choose aright program for the students that want to gain theknowledge how to design production lines.The level of popularity of simulation software in

the industry has been presented in [7]. Although thiscomparison allows to familiarize with the basic fea-tures and capabilities of many programs, it does notcover the recent versions of the software, and doesnot take into account the educational aspects of sim-ulations. An interesting method for choosing a pro-duction line simulation program by using analyticalmethods has been presented in [8]. The Analytic Hi-erarchy Process employed to select the most appro-priate simulation program seems to be particularlyuseful for industrial users, that may have well de-fined priorities as to the aspects of the simulationthat are most important in their domain. However,formulation of such weights for programs used foreducation is less obvious, as the students should fa-miliarize with various aspects of the simulations.There are some criteria for making the informed

choice of a simulation program that stem from theexperience of the educators:• the program should allow to simulate varioustypes of production lines: assembly lines, continu-ous production lines (e.g. in the food industry), ro-

botized lines, lines involving human workers, etc.,• the program should allow the user to set the spa-tial layout of the machines and other components,

• if possible, the program should simulate both thediscrete-event systems and continuous-time dy-namic systems,

• a library of pre-defined components (machines, ro-bots, conveyors, etc.) has to be available, as well asa possibility to define custom components, includ-ing their logic programmed as a script an externalfunction,

• comprehensive reports should be generated re-garding production statistics, machine usage, de-fective products, etc.,

• 2D as well as 3D graphics should be available todemonstrate the layout of the designed system, ifpossible supported by animation,

• the program should be available to the individualstudents also for the homework, so there shouldbe a free or educational version provided.A broad comparison of the programs for simula-

tion of various processes can be found in [9]. Unfor-tunately, this comparison does not include the recentversions of the software available on the market, anddoes not take into account the educational aspects,which are central to our study.Simulation programs typically used in the robot-

ics and automation education, e.g. the very popu-lar Matlab/Simulink [10] do not fulfill these require-ments, usually focusing on the simulation of dynamicsystems. The software developed specifically to sim-ulate industrial robots and their co-operating ma-chinery shares some properties with the simulatorsof production lines, but usually focus on the simula-tion of physical interactions between the machines,and on off line programming of the robots (i.e. de-veloping programs that can be transferred to realrobots). Nevertheless, such programs may be usefulfor education in the design of robotised productionlines, and should be considered as a supplement tothe general purpose simulators.Due to the limited space of this article, it was de-

cided to focus on a case study of three programs thatwe consider representative for the current softwaremarket. While selecting these programs, mainly thepossibility of using them in the academic training hasbeen taken into consideration. The key factor influ-encing the choice was the availability of free studentor trial versions, with the functionality as similar tothe full version as possible, and an easy access todocumentation and instructional materials. Anotherimportant factor was the intuitive and user-friendlyinterface. The most important functionality was theability to design the logical schemes and the 3D simu-lation/animation in one program. We assumed, that

54 Volume 5 • Number 4 • December 2014


Table 1Comparison of programs for simulation of industrial processes.

Program property Program name

Production line simulation/Virtual factory Robotics-specificsimulation software

Arena FlexSim Tecnomatix PlantSimulation

Robot Studio

Trial version + + + +

Student version + + + −

Materials for students + + + +

Restriction applied in thetrial version

Model with only 20elements

No limits on the model.Limited functionality of theprogram such as:– no data tree view– no command console– no optimalizer

Model with only 80elements

30 days trial,full funcionality

Logical diagram + − − −

Simulation of manufactur-ing processes

+ + + +

Programming robots offline − − − +

Simulation of robot motion + + + +

Simulates people at work + + + −

2D Simulation/Animation + − + −

3D Simulation/Animation Postprocesor + + +

Access to the parameters ofobject

+ + + +

Import of CAD 3D models + + + +

Using custom 3D models − − − +

Script programming lan-guages

C/C++ C++, FlexScript. SimTalk RAPID

Creation of custom libraries + + + +

Production statistics + + + −

Diagnostic tools (break-points, watch, etc.)

+ + + +

File format of the report spreadsheet (csv),txt, pdf, html, xml

spreadsheet (csv),htm, html, png

html, htm, txt −

the programs are MS Windows (XP/7) aplications,as this operational environment is most popular andcommonly available to the students, also for thehomework. Finally, the chosen programs are: Tecno-matix Plant Simulation (version 11.0.3), Arena Sim-ulation Software (version 17.70.00000), and FlexSim(version 7.1.4). Additionally, the RobotStudio (ver-sion 5.61.01) simulation software is considered in thispaper, to demonstrate both the similarities and theimportant differences between the programs devel-oped for design and simulation of production plants,and the robotics-specific software.

A qualitative comparison of their key features ispresented in Table 1.

Programs

Simulation is a method involving replacement ofreal objects by digital models to obtain the neces-

sary information about how the system works at setparameters and how it reacts when input parame-ters change. Simulation overcomes the limitations ofanalytical modeling due to the huge computationalpower of modern computers. A detailed descriptionof the designing methodology for the technologicallines using simulation programs is presented in [11].

According to [9], the simulation process involvesthe following steps:

1. Analysis of description of the system, whichshould be designed and tested.

2. Collection and processing of information: the in-put data, model parameters, timing of individualcomponents of the system and the efficiency of themachines.

3. Creation of a model in the simulation program.4. Verification of the model.5. Development of simulation experiments.

Volume 5 • Number 4 • December 2014 55


6. Conducting simulations to answer the questionsposed in the analysis phase.

7. Performance tests and simulations by changingthe parameters of machines and their properties.

8. Evaluation of experiments and final report.In the remainder of this section we present the

three plant and production line simulation softwarepackages considered in our study, focusing on theirusability for students (availability, user interface),and the functionality when designing and simulatingthe technological processes. We also briefly mentionthe limitations of the student/trial versions used inour experiments.

Tecnomatix Plant Simulation

Tecnomatix is a collection of programs for theoptimization and design of technological lines, devel-oped by Siemens [12]. Tecnomatix platform is com-posed of: Tecnomatix Jack, Intosite, Robcad andPlant Simulation. Plant Simulation is a tool for thesimulation of discrete events, which allows creationof digital models of logistics, so that properties oftechnological lines can be examined and their per-formance can be optimized.In this article, Plant Simulation details will be

presented. It allows to model the production linesin 2D and 3D, space, optimize the material flow atall levels of planning, starting with creation of thefactory, to the individual production lines. It alsohas advanced tool to analyze bottlenecks, generatevarious kinds of statistics, graphs and evaluations ofdifferent scenarios of production.The platform is available on a proprietary license.

However, student version can be used. Some limi-tations are applied – only 80 material flow objectscan be put. Siemens allows using of full version forgraduated work. In this case, the finished projectshould be forwarded to Siemens. The program hasvery good documentation, instructional videos andmanual which significantly facilitates the start ofwork.

Arena Simulation Software

Another programwhich enables the design of pro-duction lines is the Arena from Rockwell Automa-tion [13]. In this program, the user builds a model byplacing modules on the scene, reflecting the process-es.With the combination of process simulation and

optimization, Arena helps demonstrating, predictingand measuring the system performance.The program allows simulations in 2D and 3D.

Using the 2D mode, the logical model of the wholeproduction process can be created. It is adapted

to the duration and the speed of production. Tomove around the worksheet containing the model,the navigation-key, which will allow a full view ofthe selected element, can be used. 3D animations canbe created inside Arena by the module called ArenaVisual Designer. Creating the 3D models is incon-venient. If queuing, handling resources, stations orother functionalities are desired to be simulated, itis necessary to create objects in 2D scheme and thenupdate their properties to 3D environment.

Three-dimensional animation shows the arrange-ment of machinery, production line workers and illus-trates the whole process, starting with the deliveryof materials and ending with taking away of finishedproducts. For each carried simulation a report is gen-erated.

The full version is not free. However, student ver-sion can be used. It has limitations that significantlyhamper the work. Models composed of only 20 prop-erties can be used. As a result, it is possible to de-sign only a very simple process. After purchasing theacademic license, training materials in the form ofpresentations and exercises with solutions that canbe used when conducting classes with students areaccessible. The program has a very good technicalsupport. There are several books describing the pro-gram’s features. On the manufacturer’s website, in-structional videos can be downloaded and watched.Also, the online support is available.

FlexSim

FlexSim is a program for creating discrete eventssimulations developed by FlexSim Software Prod-ucts Inc [14]. Family of FlexSim programs current-ly consists of basic simulation software FlexSimand FlexSim Healthcare Simulation (FlexSim HC).The main program uses the OpenGL environment torender 3D images in real time. It is possible to sim-ulate not only the work of machines and conveyorsbut also working men, robots or forklifts.

The software is proprietary but students can usetrial version for free. Unfortunately, it has limita-tions. Certain features cannot be used, such as: thetree view, debugging test scripts. The program hasa very good technical support. Several books describ-ing the workflow in the program are available on themarket [15].

One digital line

and three simulation programs

In order to compare the properties of simulationprograms, a model of technological process was de-



signed using three programs: Tecnomatix Plant Sim-ulation, Arena and FlexSim.The chosen technological process is decorating

by engraving a kit consisting of ornamented pack-aging and two components: a USB memory and apen. A customer selects a model and defines a con-tent of inscriptions and decorative logo to be placedon objects. The logic diagram is in Fig. 1. It waschosen because it is a simple process, which can besimulated with students versions of all the consid-ered programs. Moreover, this process contains ele-ments which are necessary in the simulation of pro-duction lines. In this simple process, we can demon-strate the potential and difference of all programs.In this task, logic diagram, statistics, identifying ofbottleneck and capability of the production line willbe presented. The bottleneck is a quality control ofengraved pens and USB. The process is divided intosteps:• Every 1 minute: providing a new set into the pro-duction slot.

• Split a set pen and USB memory (30 seconds).

• Engraving: pen (3 seconds), USB memory (5 sec-onds). The engraving of both elements is simulta-neous.

• Quality control (1 minute) – rejecting those items,that do not comply assumptions of quality, suchas: incorrect engraving or hidden defects of theproducts. The amount of rejected pens is 3% ofthe entire batch, and 1%. for USB memories.

• Completing the kit and packing in the box (30 sec-onds).The simulation of process was performed for

8 hours. The modeled process can be observed andaffected by changing the sample’s size, speed of en-graving and the percentage of recoil. In all three pro-grams, it is possible to change the speed and dura-tion of the simulation. This process is implementedin a small company. The students involved in thecourse taught by the author had a realistic produc-tion process to simulate, and some of them had eventhe opportunity to check the solution on the actualproduction line. The decorating kit, which is a resultof students’ work, is shown in Fig. 2.

Fig. 1. Logic diagram of the decorating by engraving process.

Fig. 2. Example of a finished product.



Fig. 3. Model and statistics of the decorating process realized in Arena.

Arena

Model and statistics realized using the Arena ispresented in Fig. 3. All elements of the model areplaced on the sheet by dragging and dropping theselected blocks. To insert again the same block, dragthe object onto the sheet. A similar situation occurswhen combining individual elements. After clickingon the icon Connect, only one pair of blocks can beconnected. In order to connect the second pair, Con-nect function must be used again.It is possible to simulate the operating conditions

of the machine. Figure 3 presents two states: workingand waiting for the product. There is also the possi-bility of changing the graphical representation of theobject to be processed at different stages of produc-tion, e.g.: the whole set is marked with a blue sphere,but after the set is splitted, red sphere represents thepens and USBs are represented by green ball. Thefinished product was presented again as a blue ball.Unfortunately, the blocks simulating the process donot have the properties connected with efficiency. Forthis purpose, there are conditional blocks that acceptthe conditions as percentage or logical value. Deci-sive blocks cannot set the duration of this process. Inorder to simulate the duration of the quality controlprocess, appropriate blocks should be used.The program contains examples of models that

shall facilitate the work. These models can be foundin the Smart Files. They show how to solve commonproblems such as: animation of resources, logic of thedecision, how to create a queue, how to work with ex-ternal files or how to simulate the state of machinesat work. In Arena there is no separate block usedto carry out the process queuing and buffering. Thequeuing takes place using certain properties in someblocks e.g. the Process.The effect of the simulation can be observed by

the number of elements on the output of each module

displayed on the model blocks. After each simulation,a report is generated. In addition, user can use theobjects’ Variable value to display the selected val-ue anywhere. During the simulation, breakpoint andwatch functions are available.

In Arena, creating a model of the process is veryeasy, but the user must keep in mind, that not allunits have the ability to act in such time period asthey should in reality. Blocks used to establish mod-els are easy and simple to use. However, this simplic-ity implies their limited functionality.


The model and statistics implemented usingPlant Simulation are presented in Fig. 4 and ma-chines’ states distribution during the process areshown in Fig. 5. All the elements of the model areplaced on the sheet using the “drag & drop” method.One of the drawbacks of the program is a methodof connecting and inserting elements. After clickingthe icon, only one element can be added. To insertthe next element, the operation must be repeated.An analogous situation occurs during the process ofcombining elements of the model. The program al-lows to change and create own graphical visualiza-tions, even at different stages of production.

Built-in models contain a lot of configuration pa-rameters such as: time which the product spendsin the machine, the number of products beingprocessed, the conditions of entry and exit from themachine and its efficiency. The program allows to dis-play any values that are the attributes of the block.In Fig. 4, it can be seen that not only the num-ber of items leaving the individual blocks are shown,but also, in some cases, the number of input ele-ments. Transporting elements such as conveyor beltsor employees can also be simulated. The program hasbroad capabilities in terms of analysis: cost and ener-



Fig. 4. Model and statistics of the decorating process realized in Plant Simulation.

Fig. 5. Analyzing the machine states during the decorating process realized in Plant Simulation.

gy consumption, the percentage of working time andwaiting time of machines. In addition, a workersschedule, charts and graph of states of the machine atevery stage of the simulation can be created (Fig. 5).Based on this functionality, bottleneck and input-output relationship between machines can be iden-tified. When built-in properties are not sufficient,custom functionality can be created by using blockscalled methods.

Each object has a lot of properties and character-istics which significantly impedes the start of work.However, such number of features allows the simu-lation of very complex and sophisticated processes.In the exemplary process of engraving, efficiency of

the machine was taken into account, but only in theprocess of quality control following the diagnosis cor-rect and incorrect elements and their segregation tothe respective blocks. However, the main drawbackof the program is the lack of possibility to go back tothe previous version of the project after saving thechanges. Workaround is to make copies of the mod-el, so that at any moment user could go back to theprevious version. However, it is very annoying andsignificantly delays the progress of the work.

Each machine used in the program has an abilityto assign to the model documentation and detailedinformation. The program has also models of hierar-chy and inheritance, open architecture, genetic algo-



rithm optimization and enables the simulation andanalysis of energy consumption, automatic analysisof simulation and an ability to generate reports. Thevisualization allows to observe the product at variousstages of production from different points of view.

FlexSim

The FlexSim can only simulate and visualizein 3D. The simulation and analysis is presented inFig. 6.

Users can build a model by dragging and drop-ping predefined 3D objects from different classes.The parameters and functionality of the objects canbe changed in the Properties tab or by using pro-gramming languages such as C++ or FlexScript. Us-ing Tree View, the structure of the entire model in-cluding the characteristics of each element can beviewed, which helps in creating custom scripts.

The program allows to change graphical visual-ization elements at different stages of production,e.g.: in the examples process the whole kit is markedas a yellow square. After the engraving process, allthe correct products are green and the incorrect arered. The student version cannot use the console and

TreeView. The lack of this functionality makes it dif-ficult to create custom methods. The program allowsto generate a report in Excel format and analyze theoutput data using the table, graphs and by display-ing the selected variables. Sample analysis is shownin the inset Fig. 6. Program allows to send mes-sages between individual elements. For example: ifthe process did not end on one machine, a messagecan be sent from the other machine to stop the firstone.At the beginning of the creation of the program,

a time unit in which the simulation will take placewas defined. Unfortunately, the selected unit is verydifficult to change during further work with the simu-lator. This application, as the only one, has the func-tion of remembering which element was inserted in-to the worksheet and the function of remembering ifthe elements have been combined, which significantlysimplifies and accelerates the process of creating thesimulation. In this program, user can simulate thework of men, robots, machinery and transportationby conveyor belts, forklifts or workers. The processesduring the simulation allow observation of the prod-uct at various stages of production and in differentviews.

Fig. 6. Model and statistics of the decorating process realized in FlexSim.



Simulation of more complicated

processes with 3D visualization

The production line layout and control logic vi-sualized in 2D are enough to make assessment aboutthe correctness of the design and the basic proper-ties of the modeled process: bottlenecks and possibledeadlocks, production statistics and efficiency of themachines. However, there are some aspects of pro-duction line modeling where the 3D visualization andeven animations are much helpful. Among those as-pects there are modeling and assessment of the phys-ical interactions between machines (robots in partic-ular), interactions between the machines and the hu-man workers, and the ergonomics of work stands. Inorder to tell if the layout of a complicated produc-tion line is not only logically correct, but also if itleaves enough space to the workers, if the machinesdo not endanger the humans, etc. it is often neces-sary to see the whole design in 3D. Moreover, the 3Dsimulation and visualization may play crucial rolewhen presenting the new design to the customers,who not necessarily are familiar with the symbols ofthe production line design and the specific simula-tion program. In addition, some programs can buildthe actual model of the operating environment usingdata from laser scanners. Such example is describedin [16].

Therefore, in the second part of our analysis,we demonstrate how the considered simulation pro-grams can deal with more complicated design prob-lems and their 3D visualization. In this section, theRobotStudio has been included, although. this sim-ulation software belongs to a different category thanthe other programs. It allows not only to simulatethe operation of robots, but also to program themdirectly. On the other hand, RobotStudio does notallow to create an in-depth analysis of the simulatedproduction process in terms of production statistics,percentage of defective products, etc.

This time the designs do not reflect actual pro-duction lines existing in some companies. As larg-er companies are quite reluctant to unveil details oftheir technology, we conceived some examples thatspan the spectrum of typical discrete productionfrom packaging of pharmaceutics to the automotiveindustry. Although being only examples, the produc-tion lines are based on the description and/or CADmodels of real components and machines that arefreely available from catalogues and technical notesof various manufactures, and the general knowledgeas to the respective technology. As these exampleshave been entirely designed by the author and thestudents working under her supervision they do not

infringe any intellectual or commercial rights of realmanufacturers, and thus may be used in educationwithout any limitations.

Arena

In Arena, simulation of technological line is doneusing a separate tool: Arena Visual Designer. Thistool utilizes information about objects used in themodel. Unfortunately, this information is not used bythe application to put elements into the environment.User puts all the necessary equipment and sets at-tributes. However, this process can be simplified us-ing the information imported from the logical model.During the simulation, placement of the camera canbe changed. There is also the possibility to share thevisualization window to track the simulation fromdifferent angles of view. During the simulation, diag-nostic and analytical tools can be used.As an example simulation in 3D space, the line

producing drugs in capsules has been used. Figure 7presents the designed production line.The production line consists of: the initial stor-

age, encapsulating machine, conveyor belt, blistersealing machine and packaging machine. During themanufacturing process, five stages of production canbe distinguished. First one is the transport of sub-strates from the warehouse that are needed for thenext production stage – dispensing and encapsulat-ing. In this step, workers bring materials using spe-cially adapted carts. In the machine, capsules areopened, a suitable amount of powder is dispensedand the capsules are closed. Prepared capsules hitthe conveyor. Every eighth capsule is subjected toquality control by inspection of weight. The thirdstep is the delivery of the capsules to blister seal-ing machine. The machine uses rolls of foil deliveredby workers. Next, the conveyor belt collects cap-sules. The next machine imprints lot number andexpiry date. Each blister is checked for alignmentof capsules within and if there are any holes thatcould be created during the sealing of the foil. Allthe work carried out by the machine is controlledby man. The fourth step is to pack blisters to car-tons. The machine gets from the tray folded car-ton, leaflet and the right amount of blisters. Opensthe carton and inserts blisters and leaflet, closingthe carton and printing validity and batch number.Ready boxes hit the conveyor. The final step is pack-ing boxes of drugs in cartons. Cartons are collect-ed by staff and placed in a cardboard box. Whiledoing so, the packaging is inspected. Closed box-es are placed on a pallet and transported to thewarehouse, where they are handed forth to the cus-tomers.



Fig. 7. 3D simulation of packed medicine technological process in Arena.


Tecnomatix Plant Simulation designing in 3Dspace is achieved by changing the view of the dis-play. 3D stage represents objects, that were used tocreate the logical model. User can design and loadown graphical representation to the simulation. Dur-ing the simulation, diagnostic and analytical toolscan be used. The simulation of pouring beverage bot-tles and packaging of finished products into cartons(Fig. 8) was prepared in this program.

Fig. 8. 3D simulation of beverage technological processin Tecnomatix Plant Simulation.

The line consists of machines for unloading pal-lets, bottles unpacked from cartons, pouring, spiral-ing and packing nut. Employee brings palette fromthe storage and sets it to a designated place. Whatfollows next, is sliding of layers of boxes from thepallet directly to the buffer standing in front of themachine. On the basis of the buffer fill level, the em-ployee determines the frequency of delivery of pallets.After unloading all cartons, empty pallet is trans-ferred to a separate conveyor belt leading to a bufferin front of the machine stacking cartons on palletsof ready products. Empty boxes will be used againduring the packaging of filled bottles. Due to that,

they are deposited on a separate conveyor leading toa buffer of packaging machine. When a worker placesthe barrel in a designated place, runs a machine thatgrabs the barrel opens it and tilts. Then the bever-age is poured into the bottles that are closed withcaps. The first inspection step is carried out in themachine scanning bottles. It checks the degree of fill-ing of bottles and correctness of spins. The productis screened when passing through the machine. Bot-tles not meeting the requirements are marked. Thenthe sensor located at the end of the machine findsdefected bottles and changes their trajectory. Next,the correct bottles are labeled. After that, bottlesare checked by two quality controllers during trans-port to the packing machine. They selectively raisethem from conveyor and examine. They are lookingfor disqualifying defects in the products, such as ab-normal sticking labels, damaged bottles, chips on thebottles or bent nuts. Employees’ rating is the mostimportant element of quality control. Bottle packingmachine for cardboard is the last stage of production.

FlexSim

The FlexSim allows to analyze the technologicalline only in 3D. Therefore, the analysis of the func-tionality and features of this scheme has been de-scribed in the previous paragraph. With the abilityto change the view of the scene, the simulation canbe watched from different angles. Simulation of anindustrial robot and the other machines takes placeby changing their kinematics. Creation of an anima-tion of changing the position of the element duringthe simulation can be also performed. Additionally,custom 3D models can be loaded and new librariescan be created. In order to coordinate and controlthe work, transporters and operators can use theDispatcher object. Task sequences are sent to theDispatcher from an object and the Dispatcher dele-gates them to the transports or operators that areconnected to its output ports.



Fig. 9. 3D simulation of cars painting in RobotStudio.

RobotStudio

RobotStudio is a software used for offline pro-gramming of robots and simulation of how they workin production line. The program is developed by theABB company [17]. Using this program, an auto-matic production line can be created. Simulation canonly be done in 3D. Design process of a productionline consists of placing all the required componentsand setting their configuration. The program has em-bedded ABB robots models. Using simulation andanalysis, the robots’ workspaces and coordination ofall paths can be determined without risk of colli-sions. In order to program the robot, the simulationmust install RobotWare that reflects the actual con-troller and allows the robot operation in a virtualenvironment. The program has an extensive data-base of machine models. It can be expanded withusers’ own models. Models published by other userson the manufacturer’s website can be used as well.RobotStudio is delivered as 30-day trial version. Theprogram has a large number of instructional materi-als, such as: films and production lines examples.

In this application it is not possible to use thetools performing statistics and analysis of flow ele-ments. Work of robots and machines is coordinatedby the controller and inputs and outputs of each el-ement of the model. The trajectories of the tool ofthe robot can be modeled by creating virtual pathsby points in space, using the virtual joystick or theRapid language. In order to give proper function-ality to machines and components, object namedSmartComponent is used. Sensors are placed andobjects are programmed to work as the completemachine. The application also allows the detectionof two alarm states: the state before the collisionand the collision state. By declaring the appropri-ate alarm conditions, information on the work of thewhole can be contained, such as: possible conflicts,lack of power line components, information about thespeeds of the moving parts and the position of therobots’ arms.

The simulation designed using RobotStudio in-volves cars painting (Fig. 9). Robots paint the hullwith mounted fenders, hood, doors and flaps. Afterentering the paint shop, next car is subjected to thepre-treatment. The body is rinsed with a special mix-ture of water with the appropriate chemicals thatremove impurities. After that, the body is drainedfrom the remaining liquid on tilting station. Next iscataphoresis – a process in which the body is im-mersed in a bath filled with water-borne paint. Thesurface is covered using the resultant electric fieldwhich contributes to very good coverage of paint.The sheet gets anti-corrosion properties. This is thefirst procedure that gives protection of the car. In thenext block, three layers of paint are applied: lacquerprimer, topcoat and clear. Paint is applied by elec-trostatic method which will allow it to reach hard-ly penetrable places. It is pulverized by four robots(Fig. 10). After application of the clearcoat, bodymoves to a dryer, where it is thoroughly dried.

Fig. 10. 3D simulation of robots’ motion in RobotStudio.Planned trajectories of the end effectors are visible.

Conclusions

Computer simulation of the processing line is animportant step in the process of its design. Imple-



mentation of this step by using traditional meth-ods, i.e. manual calculations would be extremelytime-consuming due to the huge amount of data.In the simulation programs analytical calculations,such as those of machine capacity and efficiency,are processed automatically. The 2D modeling al-lows for the necessary studies related to the logi-cal model of the process. It allows the user to testall decision-making processes, throughput and thenumber of batches. In contrast, the 3D simulationoptimizes the deployment plan and prepares equip-ment for ergonomic workplaces in the most effi-cient way. At the stage of the design and simula-tion the equipment that will be used in the targetproject is selected. Thanks to this, before the finalchoice, the designer can test several different ma-chines and select those that best meet the expec-tations.

FlexSim, Arena and Tecnomatix are dedicatedfor making analysis and logic diagrams. This is par-ticularly a drawback whenever the production lineinvolves robotic manipulators (industrial robots).In those programs we can simulate the work of ro-bots, but it is impossible to program them, and checkthe validity of their workspaces and trajectories. Ifonly the logic diagram has to be created, Arenafrom Rockwell offers the easiest and fastest tools todo that. Creating not only the production line logicdiagram, but also adding some properties to themcan be achieved more efficiently in Tecnomatix orFlexSim. All of these programs allow simulation in3D environment.

RobotStudio differs significantly from the men-tioned three programs. It is used mainly for off-lineprogramming of robots. This program allows to plantrajectories for the robots, check the collisions possi-bility, and develop programs for the robots, but lacksan ability to create logic diagrams nor does it offerany tools for production flow analysis.

Thus, it should be considered as a tool support-ing the production automation course in these areas,where the production plant/line simulators are notspecific enough.

Simulation programs combine the issues of pro-duction and process engineering. Thanks to such pro-grams, students learn how to design different produc-tion scenarios. Simulation programs can also serve asmock factories, where all the relevant information onproduction processes are located. Optimization of lo-gistic efficiency and optimal use of resources providesas the result the information needed to make rationaldecisions at a very early stage of planning.

Unfortunately clearly identifying a single perfectapplication is not possible. Program selection is done

individually, depending on the technological process-es and the aspects of simulation that are most im-portant to the particular project.

The involvement of Automation and Robotics

course students at the Poznan University of Tech-

nology in the evaluation of the presented programs is

gratefully acknowledged.

References

[1] Gregor M., Medvecky S., Application of digital en-geenering and simulation in the design of products

and production systems, Management and Produc-tion Engineering Review, 1, 1, 71–84, 2010.

[2] Bubarnik P., Plinta D., Modelowanie i symulacjaw sterowaniu systemem produkcyjny (in Polish), Ma-teriały konferencyjne “Inżynieria Produkcji 2000”,zeszyt numer 7, Zeszyty naukowe nr 57, PolitechnikaŁódzka Filia w Bielsku-Białej, Bielsko-Biała 2000,pp. 34–40.

[3] Palajova S., Figa S., Gregor M., Simulation of man-ufacturing and logistics systems for the 21-th cen-

tury, Simulation, emulation, metamodelling, cloudcomputing, 7, 2, 47–60, 2012.

[4] Krajcovic M., Plinta D., Comprehensive approachto the inventory control system improvement, Man-agement and Production Engineering Review, 3, 3,34-44, 2012.

[5] Koster Systemtechnik, http://www.koester-system-technik.de/, date of access: July 2014.

[6] MICRO, http://e-micro.pl/stanowiska-dydaktycz-ne-c-1.html, date of access: July 2014.

[7] Dias L., Pereiro G., Vilk P., Oliveira J., Discretesimulation tools ranking – a Commercial Soft-

ware Packages comparison based on popularity,http://repositorium.sdum.uminho.pt/bitstream/1822/15634/1/ISC 2011 Veneza 5 10.pdf, 2011, dateof access: July 2014.

[8] Zdanowicz R., Choice of software for productionprocess simulation and modeling [in Polish], PAR,1, 1, 10–17, 2006.

[9] Abu-Taieh E., Rahman el Sheikh A., Commercialsimulation packages: a comparative study, Interna-tional Journal of Simulation, 8, 2, 66-67, 2007.

[10] MathWorks, http://www.mathworks.com/index.html?s tid=gn logo, date of access: July 2014.

[11] Zdanowicz R.,Modeling and simulation of manufac-turing processes [in Polish], Wydawnictwo Politech-niki Śląskiej, Gliwice, 2007.



[12] Siemens, http://www.plm.automation.siemens.com/en us/products/tecnomatix/index.shtml, date ofaccess: July 2014.

[13] Rockwell Automation, http://www.arenasimula-tion.com/, date of access: July 2014.

[14] FlexSim, http://www.flexsim.com/, date of access:July 2014.

[15] Beaverstock M., Greenwood A., Lavery E., Nord-

gren W., Applied Simulation: Modeling and Analy-sis using FlexSim, FlexSim Software Products, Inc.,2010.

[16] Plinta D., Improving of production processes withthe usage of tools for computer visualization [in Pol-ish], PAR, 2, 165–170, 2009.

[17] RobotStudio, http://new.abb.com/products/robo-tics/robotstudio, date of access: July 2014.


simulation of production lines in the …mper.org/mper/images/archiwum/2014/nr4/7-rostkowska.pdfand...

Documents