Reducing the duration of maintenance tasks by using SMED techniques

(1)

Reducing the duration of maintenance

tasks by using SMED techniques

R.M. Bresser

S2361663

Version: 1.0

Date: 22-06-2014

University of Groningen

Faculty of Economics and Business

MSc. Technology and Operations Management

Supervisor:

dr. N.D. van Foreest

(2)

Abstract

The aim of this study is to explore whether techniques of the SMED concept are applicable to reduce the duration of maintenance tasks. This is done for a continuous production environment because for maintenance in this environment, the system has to shut down which results in production losses. A first theoretical selection is made of SMED techniques which are promising applicable to reduce the duration of maintenance tasks. This selection is made in consultation with FrieslandCampina Bedum and based on the differences between setup and maintenance tasks. The selected SMED techniques are practically tested within a case study at the FrieslandCampina Bedum factory, which is a continuous production environment. In this case study the duration of maintenance tasks of FrieslandCampina Bedum are analysed and with the selected SMED techniques various solutions are proposed that could lead to a reduction in time of the maintenance tasks. Finally the results of the case study are reviewed and the applicability of the selected SMED techniques is reviewed. This study confirmed that eight out of twelve SMED techniques are applicable to reduce the duration of maintenance tasks and these eight techniques can lead to a 32.5 hours reduction of downtime per year at FrieslandCampina Bedum. This research is an exploratory study because there are no studies so far that tried to apply the SMED concept to reduce the duration of maintenance tasks. In this study a theory is build and this theory is empirically tested.

(3)

Preface

From February to June 2014 I started my final thesis at FrieslandCampina Bedum to finalise my master degree Technology and Operations Management at the University of Groningen. Now after months of hard work I am satisfied that I can present the final version of my master thesis. Writing my thesis was for me an interesting, insightful and great experience. I had the opportunity to interview and observe a variety of people inside FrieslandCampina Bedum. FrieslandCampina gave me always a warm welcome and provided me with the right information for my thesis. I would like to thank all the people for their interest and time that helped me to get new insights and information. I would like to thank my supervisor dr. N.D. van Foreest for his constructive feedback during the personal and group meetings. The provided feedback and advice gave me new insights and helped me to successfully complete my master thesis.

Also I would like to thank my supervisor of FrieslandCampina Bedum Edwin Kreder. He was always available if I had any questions and he provided me with the right data that I needed for my thesis. Thanks for all the hours spent on my thesis to bring it to a successful end.

And finally I would like to thank my fellow students who also did their theses at FrieslandCampina Bedum. The discussions we had and the reviews we did of our theses helped me to complete my master thesis.

(5)

1 Introduction

In a continuous production environment it is desirable to have as little breakdowns of the machines as possible. This is because the plant cannot produce when it is down and this results in production losses which are very costly. When such a breakdown occurs corrective maintenance need to be executed to repair the system to an operating state. During this corrective maintenance the system is out of order. A method to reduce the amount of breakdowns is executing preventive maintenance. The downside of preventive maintenance is that in a continuous operation environment typically the line has to shut down in order to perform preventive maintenance because this results in production losses (Laggoune, Chateauneuf, & Aissani, 2009). So to minimize the downtime of the plant, it is important to keep the duration of corrective and preventive maintenance as short as possible. The aim of this study is to find a way to reduce the duration of maintenance tasks in a continuous operation environment. Various researches have been done about preventive and corrective maintenance but none of these studies aim at reducing the duration of the corrective and preventive maintenance tasks. A solution is to combine preventive and corrective maintenance tasks. In literature this is known as opportunistic maintenance, where by combining corrective and preventive maintenance the availability of the system increases (Rao & Bhadury, 2000). Literature about opportunistic maintenance does not take into account how the duration of preventive and corrective maintenance tasks can be reduced (Bäckert & Rippin, 1985; R Dekker & Smeitink, 1994; Rommert Dekker, Wildeman, & Duyn Schouten, 1997). But it is important to keep the duration of preventive maintenance tasks short, so more preventive maintenance tasks can be done during corrective maintenance. And shorter corrective maintenance tasks are desirable because this will reduce the overall maintenance duration.

A possible method to try to reduce the duration of maintenance tasks is by using the Single minute exchange of die (SMED) method. SMED is a method found in the literature that is designed to reduce setup times. SMED is a process-based concept that involves the separation and conversion of internal setup operations into external ones (Benjamin, Murugaiah, & Marathamuthu, 2013; Shingo, 1985), where internal setup operations can only be done if the system is shut down and external setup operations can be done while the system is still operating. SMED uses twelve techniques for separation and conversion of internal to external activities. Originally this SMED method is designed for faster setup times, but it can also serve other losses in production processes such as the reduction of the duration of maintenance tasks (Benjamin et al., 2013). Although researchers stated that SMED is applicable to reduce the duration of maintenance tasks, this has not yet been truly researched.

(6)

Chapter 2 will address the theoretical background of the SMED method. Chapter 3 explains the methodology. In chapter 4 a selection of SMED techniques is chosen. In chapter 5 the data analysis is done and the solutions are proposed. Chapter 6 and 7 are the discussion and conclusion respectively.

1.1 Research Question

The main research question of this research is:

‘How can techniques of SMED be used to reduce the duration of maintenance tasks in a continuous production environment?’

Sub-questions:

 What are the current SMED techniques used in reducing setup times?

 Which of these SMED techniques are usable to reduce the duration of maintenance tasks?  How can these techniques be applied to FrieslandCampina Bedum?

 What are solutions for FrieslandCampina Bedum when implementing the SMED techniques?  Does the case study prove that the SMED techniques indeed are applicable to reduce the

(7)

2 Theoretical background

The aim of this study is to come up with techniques of the SMED concept that are implementable to reduce the duration of maintenance tasks. In this chapter 2 the current literature about this topic is reviewed. First in section 2.1 maintenance is discussed and what the different is between preventive and corrective maintenance. SMED is explained in section 2.2 on how it is currently used for setup time reduction.

2.1 Maintenance

The throughput of a continuous production system depends on the breakdown-free equipment and processes. The throughput will increase when these breakdowns can be reduced in a cost effective way (Pandey, Kulkarni, & Vrat, 2012). Preventive maintenance can be a good method to reduce the amount of breakdowns. Preventive maintenance is scheduled maintenance, usually periodical, in which (e.g. inspection, replacement, cleaning, adjustment and repairing) tasks are performed. These tasks reduce the probability of system failure and also increase product quality (Laggoune et al., 2009). Most preventive maintenance tasks can only be done if the system is shut down which is a problem for continuous production systems (R Dekker & Smeitink, 1994). Reducing the duration of these preventive maintenance tasks will increase availability of the system and thus can decrease loss of throughput (Waeyenbergh & Pintelon, 2004).

If there is a breakdown the system must be repaired. This type of maintenance is called corrective maintenance, which involves the activities to repair the system after a breakdown to a new state (Ashayeri, Teelen, & Selenj, 1996). Because it is not known when a breakdown occurs, it is hard to know in advance when corrective maintenance needs to be performed.

2.2 Single Minute Exchange of Die (SMED)

(8)

To illustrate the SMED applicability to reduce the maintenance task duration, first the concept SMED will be further explained.

The SMED concept uses 3 stages to reduce losses (Shingo, 1985). These stages are shown in Figure 3-1 (Shingo 1985:92) and are as follow:

Stage 1. Separating internal and external setup.

In this step a distinguishing is made between internal activities and external activities. External activities are activities that can be done before the system shuts down. Internal activities are activities that can only be performed if the system is shut down.

Stage 2. Converting internal to external setup.

Re-examine all the internal activities to look for activities that are wrongly assumed to be internal, or internal activities that can be converted so that they will become external activities.

Stage 3. Streamlining all aspects of the setup operation.

Internal time can be reduced with stage 1 and 2. But with a detailed analysis of all the aspects of all the activities done for a setup often a further reduction can be made. In this stage, both the internal and external setups are examined and considered in what way they could be shortened.Step 3 does not have to continue after step 2. This means that these two steps do not have to be performed sequentially but can be performed in parallel.

Figure 2-1 SMED conceptual stages and techniques (Shingo 1985:92)

(9)

Stage 1: 1. Using a checklist

2. Performing function checks

3. Improving transportation of dies and other parts

Stage 2: 4. Preparing operating conditions in advance 5. Function standardization

6. Using intermediary jigs

Stage 3: 7. Improving storage and transportation of blades, dies jigs, gauges, etc.

Overall 8. Implementing parallel operations 9. Using functional clamps

10. Eliminating adjustments 11. Least common multiple system 12. Mechanization

Table 2-1 SMED techniques

1. Using a checklist (stage 1). With this technique a checklist is made of all parts and steps required

in an operation including names, specifications, number of items, settings and numeric values for all measurements and dimensions.

2. Performing function checks (stage 1). A checklist is needed to see if all parts are where they

should be. A function check checks if all parts are working properly and are in working order.

3. Improving transportation of dies and other parts (stage 1). Parts and tools need to be moved

from storage to the machines, and back when the setup task is finished. This needs to be done as an external activity, in which the operator moves the parts himself while the machine is running, or another worker need to be assigned to transport the dies and parts.

4. Preparing operating conditions in advance (stage 2). Preparing operating conditions in advance

can shorten the internal activities. To give an example, if for a setup a special part is needed and this part consist of several other parts, the subassembly can already be made as an external activity before the system shut down and the actual setup is performed.

5. Function standardization (stage 2). These techniques suggest to design the setup tasks in such a

way that they can only be performed in one way and that they are performed in the same best way. This prevents that the setup tasks require more time because they are not performed in the best possible way.

6. Using intermediary jigs (stage 2). A jig is a custom made tool to fix or control the location of

another tool and helps to keep it in place. For a setup two standardized jig plates of the appropriate size and shape can be made. While one jig plate is attached and being processed, the other jig plate can be modified as an external setup so this jig plate is ready to attach once the setup need to be performed. Hereby no adjustments have to be made to the jig as an internal activity.

7. Improving storage and transportation of blades (stage 3). If technique 3 is used to convert the

transportation of parts from internal to external there is still some optimization possible. The transportation of parts in the external setup can also be reduced if the parts can be located closer to the setup location or the parts are on a location where they are better visible and they can be found more easily.

8. Implementation of parallel operations (all stages). Some setup tasks including adjustments on

large machines involving work at the front and back of the machine. If a single person perform this tasks he is constant moving from the front to the back and time is being wasted. Parallel operations involving more than one worker can be very helpful in speeding up this kind of tasks.

9. Using functional clamps (all stages). A functional clamp is a device that holds an object in place

(10)

are used to fasten the bolt. The other treads in between the first and last thread can be seen as waste.

10. Elimination of adjustments (all stages). Reducing test runs and adjustments can save a lot of

time. If more attention is paid during the first stages of the setup task less test runs and adjustments are needed because everything is already in perfect place. Continuous attention is needed during a setup to ensure test runs and adjustments consume less time.

11. Least common multiple system (all stages). The person who does the setup tasks should only

perform the functions required for a given operation. The mechanics need to be designed so that during a setup task only the function should be changed and the mechanism can stay the same. Doing as less as possible can enhance the speed of setup operations.

12. Mechanization (all stages). Mechanization helps the person who is doing the setup so that the

(11)

3 Methodology

The aim of this study is how techniques of the SMED concept can be applied to reduce the duration of maintenance tasks. A case study is used within the cheese factory of FrieslandCampina Bedum to test if these proposed techniques are implementable in practice. FrieslandCampina Bedum performs preventive and corrective maintenance tasks on a continuous operation system. For most preventive maintenance tasks at FrieslandCampina Bedum, the system has to shut down in order to perform these tasks. Because FrieslandCampina Bedum produces cheese which is a perishable good, it is also important that maintenance tasks are executed as fast as possible. So FrieslandCampina Bedum is a suitable environment in which this study can be tested in practice.

For this study a case study is used because a case study is a good method to answer a “how” question (Yin 2003:5)(Karlsson 2009:164). The build theory about the SMED concept applied to maintenance tasks has not yet been tested and the framework is not yet well understood. This makes this study an exploratory study were the theoretical framework is built and empirically tested (Karlsson, 2009). Because FrieslandCampina Bedum is a continuous operation environment the system is relatively hard to adjust without loss of production and thus an experimental study does not suit this study. No control of the behavioural events can be done, but observations can be made about the current way of use of their preventive and corrective maintenance tasks. This suits a case study because for a case study no control of these events is needed (Yin 2003:5). The data needed can be obtained from both contemporary events and current events which suits a case study as well (Yin 2003:5).

During this study of six months a number of data collection techniques were used to obtain the necessary data. Mechanics are followed to observe the current way in which they perform maintenance tasks. Semi-structured interviews are hold with people with different job functions to obtain information from different perspective. Open questions are asked to get the opinion of the employees. Closed questions are asked to get feedback about the proposed solutions. The following 7 steps are executed during this study to answer the research question of section 1.1.

3.1 Step 1: Design of a theoretical framework about SMED

Information about SMED techniques must be collected in order to define if these techniques could be used in maintenance tasks. A further research in the current literature is done to define SMED techniques obtained by other researchers and these techniques will be critically reviewed and explained. This theoretical framework consists of twelve current existing SMED techniques and is used as the basis of the rest of this research. This theoretical framework is necessary to be known before the case study at FrieslandCampina Bedum will be executed, in order to analyse the current way in which FrieslandCampina Bedum performs their maintenance tasks

3.2 Step 2: First round of observations and interviews

(12)

manager thinks this machine is the best machine because this machine has the most maintenance tasks per year compared to the other machines at FrieslandCampina Bedum.

3.3 Step 3: Selection of SMED techniques

After the maintenance tasks are observed in the first round of observations and interviews and it is known how FrieslandCampina Bedum executes their maintenance tasks, this data will be analysed. Based on the difference between setup and maintenance tasks and the first round interviews with FrieslandCampina Bedum, SMED techniques are chosen that are possible applicable to reduce the duration of maintenance tasks. This research will focus on these chosen SMED techniques.

3.4 Step 4: Second round of observations and interviews

First semi-structured interviews were done using the Kipling method technique to get information about what the employees think what aspect of the maintenance tasks could be optimized. The Kipling method uses the five W and H questions (who, what, where, when, why and how). All these questions cannot be answered with a simple yes or no answer, so the respondents give a richer information response. The Kipling method is used to find out what they really think and what in their eyes are the best solution to their own proposed problems (Kanawaty, 1992:100). This method is chosen so that the employees are not yet biased by the researcher his ideas of improvement. After the opinions of the employees were obtained, examples were given of what the literature would suggest that would be good solutions to reduce the duration of maintenance tasks. The employees of FrieslandCampina Bedum were asked if these solutions indeed would lead to a reduction in the duration of their maintenance tasks and how much time actually can be saved with these solutions. To validate the outcome of the interviews the researcher followed mechanics. With a stopwatch the time of several tasks of maintenance where measured to see if these correspond with the answers of the interviews.

3.5 Step 5: Designing a list of possible solutions

After the second round of observations and interviews a list is designed with solutions which most likely lead to a reduction in time of the maintenance tasks. This list contains solutions that are obtained by implementing SMED techniques. This list is shown to the FrieslandCampina Bedum employees and they were asked to validate this list if the durations mentioned in the list are realistic.

3.6 Step 6: Review of the solutions

The different solutions are compared with each other to see which solution has the most potential to reduce the duration of maintenance tasks. The expected amount of time that can be saved is provided along with the fact if the solution is easy to implement in the FrieslandCampina Bedum factory.

3.7 Step 7: Critical review of proposed SMED techniques

(13)

4 Selection of SMED techniques

The SMED techniques previously discussed are the twelve SMED techniques originally developed for reducing the duration of setup times. Because setup tasks and maintenance tasks are two different types of tasks it is possible that not all twelve techniques are also applicable to maintenance tasks. Therefore the differences between setup tasks and maintenance tasks are discussed in section 4.1 and in particularly the maintenance tasks at FrieslandCampina Bedum in section 4.2. After these differences are known a selection out of these twelve SMED techniques is made in section 4.3 of which this research will focus on.

4.1 Difference between setup tasks and maintenance tasks

Setup tasks occur far more frequently compared to preventive and corrective maintenance tasks. Also there are only a couple of different setup tasks that are repeated over and over. Thus the complete system can be adjusted so that the environment is best suited for these specific setup tasks. For preventive maintenance and corrective maintenance it is hard to adjust the machine so that the preventive maintenance and corrective maintenance tasks can be done easier. The diversity of preventive maintenance and corrective maintenance tasks is bigger compared to setup tasks, and preventive maintenance and corrective maintenance tasks occur far more infrequently.

Another difference is the scope of machinery. Maintenance overlaps a wider part of the complete system. Where for setup tasks the tasks are a repeating operation to a specific part of the machine, only that part that need to be changed in order to perform another run. For every maintenance tasks a different part of the machine needs maintenance. The diversity of maintenance tasks is therefore bigger compared to setup tasks. Things like the use of the same size bolts etcetera are hard to implement because these things are dependent on the type of parts used for the machines that are operating. Parts and machines are bought from a supplier and are already assembled. The companies might not have that many choices to make. So the setups can be reduced by such implementation but for preventive maintenance and corrective maintenance tasks it is difficult to implement these things and make the system best suited for maintenance. If a preventive or corrective maintenance task occur frequent in time then there is a problem with the machine. A solution will be found to eliminate this frequent occurring maintenance task. This confirms that preventive and corrective maintenance do not occur frequently in time.

4.2 Maintenance tasks at FrieslandCampina Bedum

(14)

4.3 SMED techniques for maintenance tasks

In this research the focus lies on the SMED techniques which are applicable to maintenance tasks. Because no other literature is found on this topic this research will make an assessment of the techniques that are possibly useful for maintenance tasks. With the mentioned differences between maintenance tasks and setup tasks, the information about FrieslandCampina Bedum obtained from interviews and observations, the current knowledge of maintenance and the literature about SMED, the following SMED techniques are chosen as techniques that are possible applicable to reduce the duration of maintenance tasks.

1. Using a checklist (stage 1). Using a checklist can be used for maintenance tasks. A checklist is

made of all the parts, tools and steps that are required to perform the maintenance tasks. This list includes names, specifications, numbers of tools and equipment and settings. By making a checklist a good overview is given what is included in the performing of the maintenance task.

2. Performing function checks (stage 1). Performing a function check can be used for maintenance

tasks. It can be used for preventive maintenance and corrective maintenance to check if the machines are working in a way as they should be. Function checks can make sure in maintenance tasks if the machines are functionally properly.

3. Improving die transportation (stage 1). Improving die transportation can be used for maintenance

tasks. Although there are no dies in maintenance tasks this technique can be used for other parts and tools. Parts and tools need to be moved from storage to the machines, and back when the maintenance task is finished. This need to be done as an external activity. This is especially applicable to preventive maintenance tasks because these are planned tasks. Because corrective maintenance is unplanned it is difficult to move parts from storage to the machines in advance. An example of this SMED technique could be to move the spare parts closer to the machines that need maintenance.

4. Preparing operating conditions in advance (stage 2). Preparing operating conditions in advance

can be used for maintenance tasks and can shorten the internal activities. If for example a machine part has to be replaced and this part consist of several other parts, the subassembly can already be made as an external activity before the system shuts down and the actual maintenance is performed. This is easier for preventive maintenance tasks because they are known beforehand. This can also be done for corrective maintenance tasks because it could be possible that there are spare parts available that include the complete subassembly of parts. So if a part breaks down it can be convenient to just replace the complete subassembly with a new one and later repair the broken sub assembly.

5. Function standardization (stage 2). Function standardization cannot be used for maintenance

tasks, because there is a lot of diversity in the maintenance tasks. It is not possible to standardize these tasks. Preventive maintenance and corrective maintenance tasks are many different tasks that repeat far less frequently over time compared to setup tasks. This makes it difficult and cost inefficient to standardize the maintenance tasks.

6. Using intermediary jigs (stage 2). Using intermediary jigs cannot be used for maintenance tasks. A

jig is not used for maintenance tasks. This is a specific part used for setup tasks.

7. Improving storage and transportation of blades (stage 2). Improving storage and transportation

(15)

place better to store these spare parts and tools. Not only the transportation is important but also the way the spare parts and tools are stored. Can the parts be found easily, is the used system congruent and are parts often lost, are questions that relate to this technique.

8. Implementation of parallel operations (stage 3). Maintenance tasks can be so complex that if they

are done by a single person it can take a long time to perform. But it could be possible with assistance of a co-worker, that the duration of the maintenance task can be reduced drastically. It may be possible that the co-worker can be an unskilled worker that assists the maintenance person and does not need any information or skill set for this assistant task.For example a task, done by a single person, can take 12 minutes. But with two persons it will not take 6 minutes but 4 for example. A second person is not only good to assist the first person physical but the second person can also give his opinion about what he thinks is the best way to perform the specific maintenance tasks.

9. Using functional clamps (stage 3). Using functional clamps cannot be used for maintenance tasks.

Although it will save time this technique cannot be applied because this is too much work. Maintenance tasks overlap a wider part of the machines, so if functional clamps are used they need to be designed for all the machines, which is a time-consuming activity.

10. Elimination of adjustments. Elimination of adjustments can be used for maintenance tasks. It is

partly the same as for setup tasks where if attention is paid in the first stages of the maintenance tasks this results in less test runs and adjustments because everything is already in perfect place. For maintenance tasks there could be an addition to this technique where if a part breaks down it can be time saving to just replace the whole part instead of repairing the part (which is an internal activity) during the downtime. If a spare part is in stock this part can be replaced so the machine can go operating as soon as possible and the part that is broken can be repaired as an external activity in the workshop.

11. Least common multiple system. Least common multiple system cannot be used for maintenance

tasks. These techniques focusses into much detail to a single aspect of a setup operation. This cannot be done for every maintenance task individually because there are many different maintenance tasks and it is difficult to predict which maintenance tasks will occur in the future.

12. Mechanization. Mechanization can be used for maintenance tasks. Mechanization makes it

(16)

In section 4.1 and 4.2 the difference between setup and maintenance are explained. These differences are shown in figure 4.1 along with the 12 techniques of SMED. In this figure 4.1 a selection is shown of the 12 techniques based on the differences proposed in section 4.1 and 4.2.

Figure 4-1 Selection criteria versus techniques

Because of the differences between setup tasks and maintenance tasks not all SMED techniques are applicable for maintenance tasks. To give a summary, eight out of the twelve SMED techniques are applicable to maintenance tasks. This can be seen in figure 4-1. With a case study, it is tested, which of these techniques indeed are applicable to maintenance tasks of FrieslandCampina Bedum. These eight techniques are:

 Using a checklist (stage 1)

 Performing function checks (stage 1)

 Improving transportation of tools and other parts (stage 1)  Preparing operating conditions in advance (stage 2)  Improving storage and transportation (stage 3)  Implementation of parallel operations (all stages)  Elimination of adjustments (all stages)

 Mechanization (all stages)

1. U sing a che ck lis t 2. P e rfor m ing func tion che ck s 3. Im pr ov ing tr ans por ta tion of di e s a nd ot he r pa rt s 4. P re pa ring ope ra ting condi tions in adv anc e 5. Func tion st anda rdi za tion 6. U sing int e rm e di ar y j ig s 7. Im pr ov ing st or ag e a nd tr ans por ta tion 8. Im pl e m e nt ing pa ra lle l ope ra tions 9. U sing func tiona l c la m ps 10. Eli m ina ting a dj us tm e nt s 11. Le as t c om m on m ul tipl e sy st e m 12. M e cha ni za tion

Corrective maintenance tasks are not known in advance when to happen. V V V V V - V V V V V V

PM and CM tasks overlap a wider part of the machines. V V V V X - V V X V X V

PM and CM tasks are many different tasks that repeat far less frequent over time. V V V V X - V V V V V V

PM and CM do not occur that frequently in time. V V V V V - V V V V V V

There is far more variation in PM and CM compared to setup tasks. V V V V X - V V V V X V

(17)

5 Data Analysis

Because maintenance tasks of FrieslandCampina Bedum are very diverse not all components of these tasks can be named. Also measuring how much time they take to execute is difficult to do. Maintenance tasks can take from several minutes to multiple hours. Maintenance tasks ranges from replacing a motor to re-screwing a bolt. Maintenance tasks often do not repeat over time because if a maintenance tasks repeats itself over time FrieslandCampina Bedum will adjust the system so that this task is in the future no longer necessary. Although there is a wide variety of maintenance tasks a lot of tasks have components which are common among almost all tasks. Stage one of the SMED concept and the corresponding three techniques of this stage are used to identify the internal and external activities of the maintenance tasks. After this was done four major subjects were found where possibly waste could be eliminated. These subjects are the type of tools and instruments available for maintenance, the current design of the warehouse for spare parts, the transportation of spare parts and the availability of mechanics. These problem subjects are explained in section 5.1. Section 5.2 gives possible solutions to the problems stated in section 5.1. Section 5.3 presents a review of the proposed solutions. And at the end the eight SMED techniques are reviewed in section 5.4

5.1 Problem analysis

Instruments and tools

For all maintenance tasks tools and instruments are needed to perform the maintenance tasks. This holds for preventive as well as corrective maintenance tasks. Recently FrieslandCampina Bedum implemented a new system for storing the tools and instruments. Previously every mechanic has his own toolbox containing the tools depending on what the mechanics think they need for the tasks they perform. Currently these toolboxes are replaced by bigger toolboxes on every department. Every department now has its own toolbox containing all the tools and instruments for this particular department. So the mechanics need to use the toolbox of the department of where they perform a maintenance tasks. The problem with these toolboxes is that they contain a lot of tools which will never be used. Because the toolbox contains a lot of tools the mechanics are searching longer for their tools then needed. During the interviews this is indicated as an annoyance. It depends on the maintenance task how much longer it takes to find the required tool or instrument. Previously every mechanic could use his own special accessories to make maintenance tasks easier to perform. These special accessories are also missing in the current toolboxes.

Spare parts

For most maintenance tasks there are spare parts needed at FrieslandCampina Bedum. These spare parts are stored and sorted in a warehouse. FrieslandCampina Bedum makes use of two different sorting systems. One part of the warehouse sort products by similarity and the other part of the warehouse sort products by parts of machines. To give an example if a bolt is needed of a particular engine, this part can be at the bolts section of the warehouse or it can be at the department of that specific engine inside the warehouse. It often happens that mechanics need a spare part and they start searching in the wrong place because they do not know in advance in which section of the warehouse this spare part is positioned.

(18)

look to see if the spare parts are in stock and where the location of this spare part is. Currently the mechanics have to go to the warehouse by themselves and look up in books what the location of that spare part is. These books are last updated in 2012 and contain many errors and adjustments by hand. After they know the location they have to see at the location itself whether or not there is still that spare part in stock. It could happen though that they search for an hour and come to the conclusion that the spare part they are looking for is not in stock at all. If this happen they have to look for an alternative spare part which can be used as a temporary solution. Often this solution will not be temporary but become a permanent solution.

Transportation of spare parts

The spare parts that are required for the maintenance tasks need to be picked up at the warehouse. At FrieslandCampina Bedum this warehouse is stored outside the factory proximately 400 to 550 meters depending on the location of the maintenance task inside the factory.

The factory of FrieslandCampina Bedum produces cheese which needs to be produced in a very hygienic environment. Therefore there are several rules about hygiene inside the factory. If a mechanic enters the factory he has to decontaminate himself and change his clothes because the clothes worn inside the factory cannot be taken outside. If a mechanics needs to change clothes this includes hair covering, beard covering, a special long coat, special shoes, washing hands and make sure that he do not takes any dirt inside the factory with his tools and spare parts.

Because the location of the spare parts is outside the factory every time a mechanic needs a spare part for his maintenance tasks he need to walk to the factory and change his clothes. This needs to be done twice because on his way back to the factory the mechanic has to put his special clothes back on. At FrieslandCampina Bedum this is seen as waste because on average, a single mechanic on one working day has to walk eight times to the warehouse and back. One single visit to the warehouse takes 23 minutes as can see in table 6-1.

Step Description of step Duration

1 Change of clothes 5 minutes

2 Walking to spare part warehouse 5 minutes

3 Pick up spare parts 3 minutes (can take a hour)

4 Walking back to factory 5 minutes

5 Change of clothes 5 minutes

Total 23 minutes

Table 5-1 Duration of transport of spare parts

(19)

Availability of mechanics

In case a machine breaks down corrective maintenance need to be performed to repair the machine so that the factory can continue operating. When a machine breaks down is not known in advance. Therefore FrieslandCampina Bedum works with a schedule that during the day between 06:00 and 22:00 there is always a mechanic present in case of an emergency and corrective maintenance needs to be performed. During 22:00 and 06:00 there is no mechanic present at the factory. If corrective maintenance is required during this time then the mechanic must come from his home. This means on average an extra 15 minute downtime because a mechanic needs to travel from their home to the factory.

5.2 Solutions

In the literature various examples of SMED implementation are found which proposes solutions (Gilmore & Smith, 1996; Mileham, Culley, Owen, & McIntosh, 1999; Ulutas, 2011; Van Goubergen & Van Landeghem, 2002). Some of these practical examples can also be used for maintenance tasks. In consultation with FrieslandCampina Bedum, the use of the selected SMED techniques and the example solutions in the literature possible solutions were proposed to eliminate the waste found in the maintenance tasks of FrieslandCampina Bedum. These solutions are as follow and will be explained below.

A. Move the warehouse closer to factory B. Assign a warehouse manager

C. Use of a digital spare parts inventory system D. Implement one sorting system

E. Review the content of the toolboxes F. Schedule 24 hours shifts for the mechanics

A. Move warehouse closer to the factory

Moving the warehouse closer to the factory could be a solution for the current long walking distance when the mechanics need a spare part out of the warehouse. The SMED technique “improving transportation” is used for this solution. If the warehouse is moved closer to the factory this can save between 5 to 6 minutes of walking and if the warehouse is inside or adjacent to the factory it could possibly also eliminate the changing of clothes which take 10 minutes. Whether or not the mechanics need to change their clothes is dependent on the policy that is conducted inside the new warehouse location. A suitable location must be chosen in consultation with FrieslandCampina Bedum in order to say how many time exactly can be saved if this solution is implemented.

Step Description of step Duration before Duration after implementing

1 Change of clothes 5 minutes 5 minutes (possible 0 minutes)

2 Walking to spare part warehouse 5 minutes 2 minutes

3 Pick up spare parts 3 minutes 3 minutes

4 Walking back to factory 5 minutes 2 minutes

5 Change of clothes 5 minutes 5 minutes (possible 0 minutes)

Total 23 minutes 17 minutes (possible 7 minutes)

(20)

B. Assigning a warehouse manager

Assigning a warehouse manager can remove the changing of clothes when a spare part needs to be picked up at the warehouse. Also the walking to the warehouse and searching the spare parts inside the warehouse can be converted from internal activities to external activities if a warehouse manager is assigned to do these tasks. The SMED techniques “improving transportation” and “Implementing parallel operations” are used for this solution.

If a warehouse manager is hired he can assist the mechanics during a maintenance tasks if needed. If the mechanic needs a spare part he can call the warehouse manager and ask for the spare part. The warehouse manager will look for this spare part and bring the part to the place inside the factory where normally the mechanics have to change clothes. At this place the mechanic and the warehouse manager can exchange spare parts. This saves time because the mechanic does not have to walk that far and does not have to change clothes any more.

Currently there are already warehouse employees who are involved with the work inside the warehouse. Further research need to be done to investigate whether these employees can add this task to their current working tasks or if a new employee needs to be hired especially for this job. If a warehouse manager is assigned to help the mechanics to find the spare parts in the warehouse 19 minutes of internal activities can be converted to 9 minutes of external activities. This can be viewed in table 5-3.

Step Description of step Duration before Duration after implementing

1 Change of clothes 5 minutes 0 minutes

2 Walking to spare part warehouse 5 minutes 2 minutes (3 minutes external) 3 Pick up spare parts 3 minutes 0 minutes (3minutes external) 4 Walking back to factory 5 minutes 2 minutes (3 minutes external)

5 Change of clothes 5 minutes 0 minutes

Total 23 minutes 4 minutes

Table 5-3 - Time before and after assigning a warehouse manager

C. Use of a digital spare parts inventory system

The SMED technique used to come up with this solution is “Improving storage and transportation”. The current inventory system of FrieslandCampina Bedum is outdated. Books are used to find the location of spare parts and there is no possibility to remotely view the current stock levels of spare parts. The solution is to transfer to a digital spare parts inventory system. All spare parts should be findable in a digital environment like a pc. In this way the mechanics can remotely check the stock levels of the spare parts and directly find the location where the spare parts are stored. This solution can not only reduce the time of searching for a spare part, but mechanics do not have to go that often to the warehouse anymore to check if spare parts are on stock or not.

D. Implement one sorting system

(21)

more. It does not matter which of the two sorting mechanisms will be chosen as long as the complete warehouse makes use of the same sorting mechanism. For corrective maintenance and sometimes also for preventive maintenance the searching for spare parts is an internal activities. If this solution is implemented this solution can save minutes of internal activities. It is hard to say how many because not all spare parts are divided across the two sorting mechanism and sometimes the mechanic knows the location and does not have to search at all.

E. Review the content of the toolboxes

Current toolboxes are full of equipment that is not used and some special equipment is missing. A solution is to ask the mechanics of what they think what equipment is obsolete and what they think is missing in the current toolboxes. With this solution implemented the mechanics do not have to search that long anymore for the appropriate tools. This solution is found by using the mechanization SMED technique.

F. Schedule a 24 hours shifts for the mechanics

Because there is no mechanic at the factory during the night from 22:00 to 06:00 an extra 15 minutes is needed to travel to the factory if a machine breaks down and corrective maintenance is needed. If FrieslandCampina Bedum implements a 24 hours shift for mechanics which means that there are always mechanics present at the factory, then this can save 15 minutes per corrective maintenance task at night. If FrieslandCampina Bedum has the opportunity then they will delay the corrective maintenance task to a later date where there is a mechanic present. If this is not possible then a mechanic needs to come from his home to perform this corrective maintenance task.

5.3 Review of solutions

All the mentioned solutions could result in a reduction of the duration of maintenance tasks at FrieslandCampina Bedum. Each solution can also be carried out separately, although not all solutions will reduce the same amount of time and not all solutions will cost the same. To make a suggestion of which solution will be the best solution to implement first, the solutions are ranked according to the ease of implementation and amount of time that can be saved by implementing the solution. This is graphically shown in figure 5-1.

(22)

compared to solution A, B and C. Solution E has medium priority because it is very easy to implement although the time that can be saved with this solution E is not much. Solution F has a medium priority because it is fairly easy to implement but the time that can be saved is relatively low. Solution F can save 15 minutes but this is only for corrective maintenance tasks and only for the corrective maintenance tasks that are urgent and need to be performed at night. For solution F new mechanics should be hired to cover the night shift. Or the current mechanics should be assigned more hours.

Figure 5-1 - time that can be saved vs ease of implementation matrix

Solution B thus emerges as the best solution to implement at first. This solution can save the most time compared to the ease of implementation of this solution. A combination of solution A and C can make a good replacement compared to time saving although these solutions are not that easy to implement. Because solution D has low priority it is not the best solution to start with. Although because solution D is easy to implement solution D can be implemented if there is some time available.

An estimation can be given on how many downtime can be saved, and the associated costs, by applying the proposed solutions. FrieslandCampina Bedum has two production lines ‘Cheese 1’ and ‘Cheese 2’ which are fairly similar concerns maintenance tasks. Over 2013, production line ‘Cheese 1’, which among other things includes the Conomatic, had 331.7 hours of unplanned downtime, for which corrective maintenance is needed. These 331 hours are divided among 522 corrective maintenance tasks with an average time of 38 minutes per task. Not all tasks needed spare parts and the proposed solutions are not beneficial for all corrective maintenance tasks. Out of these 522 tasks 130 can be shortened with the proposed solutions. The proposed solutions can on average reduce fifteen minutes per corrective maintenance task, which give a total of 130 * 15 minutes = 32.5 hours downtime that can be saved per year. FrieslandCampina Bedum estimated that one hour of downtime cost 5000 euro. So a total of €162.500 can be saved per year with the proposed solutions in section 5.2. FrieslandCampina Bedum has two production lines and it is likely that for production line ‘Cheese 2’ similar results can be obtained.

Medium priority High priority

E B

F

Low priority Medium priority

D A

C

Low High

Time that can be saved

(23)

5.4 Review of selected SMED techniques

Six out of the eight selected SMED techniques are used for the proposed solutions in this study. Thus it can be stated that these six techniques are suitable for reduce the duration of maintenance tasks. These six techniques are:

 Using a checklist

 Performing function checks

 Improving transportation of tools and other parts  Improving storage and transportation

 Implementation of parallel operations  Mechanization

“Preparing operating conditions in advance” and “Elimination of adjustments” are not used in this study. This does not mean that they are not suitable for maintenance tasks. They are indeed, however it became clear during interviews that the mechanics of FrieslandCampina Bedum are already using these techniques by themselves. Good mechanics know what they have to do when a preventive or corrective maintenance tasks need to be done. They already start thinking how they can execute that task in the best way. This is also what the “Elimination of adjustments” technique proposes. The other technique “Preparing operating conditions in advance” is done by the insights of the mechanics. The mechanics are currently estimating by themselves if preparing tasks in advance leads to a reduction in the downtime of the system.

So preparing operating conditions in advance and Elimination of adjustments are suitable for maintenance tasks. Because the diverse of maintenance tasks it is not possible to analyse every maintenance tasks and see if these techniques are implementable. A good way is to inform the mechanics and teach them the overall technique so that they can assess themselves whether or not the techniques are useful for their current maintenance tasks.

(24)

6 Conclusion

The goal of this study is how the SMED concept can be applied to reduce the duration of maintenance tasks. This study found that the SMED concept is not only good for setup time reduction, but SMED is also applicable to reduce the duration of maintenance tasks. A case study is done to determine which techniques of the SMED concept are applicable. This case study showed not all twelve SMED techniques can be used for maintenance tasks because maintenance tasks compared to setup tasks overlap a wider part of the machines, maintenance tasks do not occur that frequently over time, there is more variation in maintenance tasks and corrective maintenance tasks are not known in advance when to happen. Therefore eight out of the total of twelve SMED techniques were found to be applicable to reduce the duration of maintenance tasks. These techniques are: Using a checklist, Performing function checks, Improving transportation of tools and other parts, Preparing operating conditions in advance, Improving storage and transportation, Implementation of parallel operations, Elimination of adjustments and Mechanization.

A case study is done at FrieslandCampina Bedum to test the SMED techniques in practice. Maintenance tasks at FrieslandCampina Bedum are more diverse compared to setup tasks, but also maintenance tasks contain components which are common among almost all different tasks. At FrieslandCampina Bedum the components spare parts and the transportation of spare parts and tools of maintenance tasks were the components where the most waste could be eliminated. Six out of the eight SMED techniques are used in this study to come up with various solutions to reduce the duration of maintenance. These solutions are: A. Move warehouse closer to the factory, B. Assigning a warehouse manager, C. Use of a digital spare parts inventory system, D. Implement one sorting system, E. Review the content of the toolboxes, F. Schedule a 24 hours shifts for the mechanics. Where the solution B “Assign a warehouse manager“ emerged as the best solution to implement based the ease of implementation and time that can be saved with this solution followed by solution A and C. If the solutions are implemented the downtime can be reduced with 32.5 hours per year and a cost saving of €162.500 per year.

Two out of the eight techniques are not used in this case study. These two techniques, preparing operating conditions in advance and elimination of adjustments were already implemented at the current way of performing their maintenance tasks at FrieslandCampina Bedum.

(25)

7 Discussion

The main objective of this study is to test whether SMED techniques are usable to reduce the duration of maintenance tasks. A case study at FrieslandCampina Bedum has shown that SMED techniques can be used to reduce the duration of maintenance tasks. This statement is only based on this study because this is the first study that investigates the applicability of SMED to reduce the duration of maintenance tasks. No other research is done and to validate the results of this study it is advisable that more research will be done at other companies to confirm the applicability of the SMED concept to reduce the duration of maintenance tasks. The SMED techniques proposed in this research can possibly reduce the duration of maintenance tasks amongst other factories which have similar characteristics but this need to be validated with further research.

The solutions proposed to FrieslandCampina Bedum are prioritized according to the ease of implementation and the amount of time that can be saved. An estimation of the cost that can be saved is given but the costs of implementation are not given. This is because it is very complicated to calculate the implementation costs for every proposed solution. Many different costs need to be included and currently these costs are not known at FrieslandCampina Bedum. It is deliberately chosen to not include a cost implementation estimation because it would be inaccurate due to the missing information. But the implementation costs are very important and it is advisable before executing the solutions to first define the cost associated with the solutions. This has to be done as a further research.

(26)

Reference

Ashayeri, J., Teelen, A., & Selenj, W. (1996). A production and maintenance planning model for the process industry. Journal of Production Research, 34(12), 3311–3326.

Bäckert, W., & Rippin, D. (1985). The determination of maintenance strategies for plants subject to breakdown. Computers & Chemical Engineering, 9(2), 113–126.

Benjamin, S. J., Murugaiah, U., & Marathamuthu, M. S. (2013). The use of SMED to eliminate small stops in a manufacturing firm. Journal of Manufacturing Technology Management, 24(5), 792– 807.

Dekker, R., & Smeitink, E. (1994). Preventive maintenance at opportunities of restricted duration.

Naval Research Logistics: An International Journal, 41(3), 335–353.

Dekker, R., Wildeman, R. E., & Duyn Schouten, F. a. (1997). A review of multi-component maintenance models with economic dependence. Mathematical Methods of Operations

Research, 45(3), 411–435.

Gilmore, M., & Smith, D. J. (1996). Set-up reduction in pharmaceutical manufacturing: an action research study. International Journal of Operations & Production Management, 16(3), 4–17. Kanawaty, G. (1992). Introduction to work study (p. 524). International Labour Organization. Karlsson, C. (2009). Researching operations management (p. 322). New York: Routledge. Laggoune, R., Chateauneuf, A., & Aissani, D. (2009). Opportunistic policy for optimal preventive

maintenance of a multi-component system in continuous operating units. Computers &

Chemical Engineering, 33(9), 1499–1510.

Mileham, a. R., Culley, S. J., Owen, G. W., & McIntosh, R. I. (1999). Rapid changeover – a pre-requisite for responsive manufacture. International Journal of Operations & Production Management,

19(8), 785–796.

Pandey, D., Kulkarni, M. S., & Vrat, P. (2012). A methodology for simultaneous optimisation of design parameters for the preventive maintenance and quality policy incorporating Taguchi loss function. International Journal of Production Research, 50(7), 2030–2045.

Rao, A., & Bhadury, B. (2000). Opportunistic maintenance of multi‐equipment system: a case study.

Quality and Reliability Engineering International, 16(6), 487–500.

Shingo, S. (1985). A revolution in manufacturing: the SMED system. Cambridge, MA.: Productivity Press.

Ulutas, B. (2011). An application of SMED Methodology. World Academy of Science, Engineering and

Technology, 5(1), 67–70.

(27)

Waeyenbergh, G., & Pintelon, L. (2004). Maintenance concept development: A case study.

International Journal of Production Economics, 89(3), 395–405.

Reducing the duration of maintenance tasks by using SMED techniques