EXAMENSARBETE INOM MASKINTEKNIK,
Industriell Ekonomi och Produktion, högskoleingenjör 15 hp SÖDERTÄLJE, SVERIGE 2018
Reducing die change time with lean tools
JULIA BJÖRCK ROBIN HAMMAR
Reducing die change time with lean tools
by
Julia Björck Robin Hammar
Examensarbete TRITA-ITM-EX 2018:14 KTH Industriell teknik och management
Tillämpad maskinteknik
Kvarnbergagatan 12, 151 81 Södertälje
Examensarbete TRITA-ITM-EX 2018:14
Reducing die change time with lean tools
Julia Björck
Robin Hammar
Godkänt
2018-02-26
Examinator KTH
Claes Hansson/ Alexander Engström
Handledare KTH
Bertil Wanner
Uppdragsgivare
Crane Currency
Företagskontakt/handledare
Albin Risfelt
Sammanfattning
Kunder numera kräver snabbare och snabbare leveranser, vilket ställer höga krav på företagen. Detta betyder att företagen måste kunna producera snabbt för att inte förlora kunder till sina konkurrenter.
Många företag jobbar därför hårt med att reducera tiden från att kunden lägger en order, till det att kunden får sin produkt. En del av denna tid innefattar ställtiden, vilket är tiden som det tar för företaget att ställa om en given maskin från att producera en produkt till en annan. Under denna tid kan inte företaget producera något, och det är just här som företaget då förlorar mycket pengar.
Detta projekt gick ut på att observera ställtiden på den maskin med lägst kapacitet på företaget för att därefter komma med förslag på hur den ska kunna minimeras med hjälp av olika Lean verktyg.
Under projektets gång har det avgränsats till att problem som förlänger ställtiden, men har sin grund i andra avdelningar inte kommer att beröras, då projektets tid var begränad.
Förslagen som gavs var att företaget borde införa en noggrann beskrivning av varje arbetsmoment med hjälpa av Standard Operating Procedure, genomföra en ordentlig städning och sortering på arbetsstationen med hjälp av 5S, införa ordentliga checklistor så att inte tryckarna faller tillbaka i gamla fotspår när nya implementeringar införs, optimera de moment som omställningen består av samt flytta vissa moment till att göras utanför omställningen, och till sist rekommenderas att företaget tränar upp sin personal för att omställningen ska kunna utföras trots att de inte är fullbemannade.
Detta visade sig ge ett resultat som kan reducera företagets ställtid med 3-4 timmar. Detta skulle påverka företagets lönsamhet då varje förlorad timme för denna maskin kostar företaget 50 000 SEK eftersom den är systemets flaskhals.
Bachelor of Science Thesis TRITA-ITM-EX 2018:14
Reducing die change time with lean tools
Julia Björck
Robin Hammar
Approved
2018-02-26
Examiner KTH
Claes Hansson/ Alexander Engström
Supervisor KTH
Bertil Wanner
Commissioner
Crane Currency
Contact person at company
Albin Risfelt
Abstract
The demand for faster delivery times are increasing, which in turn increases the pressure on
production companies to meet deadlines. This means that companies need to reduce their lead times to avoid losing customers to their competitors. One part of the lead time is the changeover
downtime, which is the time it takes for the company to change a given machine from producing one product to another. As the machine is not producing during this process, it is time where the company is losing money.
The purpose of this project was to attempt to reduce this changeover time through observing the machine in the production line with the lowest capacity. Through these observations, suggestions were generated as to how to reduce the changeover time through the use of an assortment of Lean tools.
During the project, these suggestions were limited to not include other departments at the company.
The suggested improvements were:
- Introduce thorough documentation of each step of the die change through Standard Operating Procedure.
- Perform a thorough cleaning and sorting of the workstation using the 5S-principles.
- Implement checklists to prevent backsliding from occurring.
- Optimize each step of the changeover.
- Train staff to be proficient in operating several machines in the production line to reduce the impact of staff members being absent.
These improvements were shown to be able to reduce the changeover time by 3 to 4 hours. This would have an impact on the profitability of the company, as each hour lost for the observed machine costs the company roughly 50 000 SEK due to it being the bottleneck of the system.
Key-words
Click here to insert text
P
REFACEWe would like to extend our deepest gratitude to anyone who has helped us during our thesis project.
We would especially like to thank our mentor, Bertil Wanner, for his guidance and assistance throughout the thesis.
We would also like to thank our guide at the company, Albin Risfelt, for always being ready to answer questions and discuss the thesis.
Finally, we would like to thank all printers and operators for allowing us to observe and question their methodology.
Stockholm, Sweden January 2018
Julia Björck and Robin Hammar
G
LOSSARYBacksliding The gradual regression of implemented improvements, leading back to earlier habits
Batch size The amount of products that are to be produced before the machine is retooled to produce something else.
Die change The process of changing the machines tools and setting so it produces a different product.
IFS Tracking system for logistics, orders and related things.
Just in Time (JIT) Receiving goods only as they are needed in the production process.
Outsourcing The company allows another company to handle one or more processes.
Post-processing The task required to finish and assemble a product before it is completely ready for delivery.
Prosit Keeps track of the numbering of the pallets.
ProTAK A computer program that tracks production speed and the OEE of the system.
Return of Investment (ROI)
A performance measure used to evaluate the efficiency of an investment.
Scheduled maintenance A planned maintenance that aims to prevent downtime from occurring.
Standard Operating Proceedure (SOP)
A detailed description of how a particular task is to be performed.
Unscheduled maintenance A non-planned maintenance that fixes a particular problem.
Content
Introduction... 1
1.1 Background ... 1
1.2 Problem statement ... 1
1.3 Purpose ... 1
1.4 Limitations ... 1
Method ... 3
2.1 Lean ... 3
2.2 Kaizen ... 3
2.3 Single Minute Exchange of Die ... 3
2.4 Standard operating procedures ... 3
2.5 The seven wastes ... 4
2.6 Ishikawa diagram ... 4
2.7 5-Whys ... 4
2.8 6 Wise Men ... 5
2.9 PDCA Cycle ... 5
2.10 Brainstorming ... 5
2.11 Overall Equipment Efficiency ... 6
2.12 Application of methods ... 6
Analysis of current state ... 7
3.1 Overall picture of the production flow ... 7
3.2 Object of study ... 8
3.3 Die change process ... 8
Problem areas ... 11
4.1 Differences in methodology when performing tasks ... 11
4.2 Unnecessary items in the workplace ... 11
4.3 Backsliding ... 11
4.4 Unnecessary call for the maintenance staff ... 11
4.5 Shift handover ... 12
4.6 Unoptimized die change process ... 12
4.7 Machine downtime due to operator absence ... 12
Improvement suggestions ... 13
5.1 Difference in methodology when performing tasks - Standard ... 13
5.2 Unnecessary things in the workplace – 5S ... 13
5.3 Backsliding ... 15
5.4 Unnecessary call of the mechanics - Checklist ... 16
5.5 Tasks are redone at shift change ... 17
5.6 Unoptimized die change process ... 17
5.7 Machine downtime due to operator absence ... 19
Economic calculations ... 21
6.1 Current values ... 21
6.2 New values ... 22
6.3 Return of Investment ... 22
Conclusions ... 25
7.1 Summary of Suggested Improvements ... 25
7.2 Discussion ... 25
7.3 Future work ... 26
References Appendix
9.1 Specification of Accountability System 9.2 Specification of Checklist Template 9.3 Checklist Template
9.4 Competence matrix
9.5 Current Die Change Process
9.6 New Die Change Process
9.8 Checklist SMED-Process for printer 1 and 2 9.9 Checklist SMED-Process for printer 3 9.10 Checklist for mechanics
9.11 Checklist for shift changes
Introduction
This section includes the thesis and the mission statement, as well as the background for the current situation.
1.1 Background
It is becoming progressively more important for companies to deliver their products Just In Time (JIT). This is in part due to the fact that having large stock requires storage to match, and thus results in increased storage costs. In order for a company to deliver JIT, they need to produce relatively small batch sizes. This requires short and effective die changes in order for it to be profitable. To ensure this, many companies opt to introduce a Lean way of thinking to reduce die change times and other types of resource waste.
This thesis project was performed at a company with high-security production. The company makes their product from scratch, and uses no outsourcing. One machine (SI) in the production line have been identified as a bottleneck. The machine is positioned relatively far down in the chain of production (See Figure 1). If everything works as intended, the die change time is
estimated to be about 20-25 hours. This happens very rarely, and it is not uncommon for the die change to take up to 48 hours.
1.2 Problem statement
When conducting standard manufacturing, there are three reasons for the production to come to a halt:
• Scheduled maintenance
• Unscheduled maintenance
• Die change
These three should equate to the total downtime on a specific machine, assuming a well-planned production schedule. Downtime on a machine is directly related to Overall Equipment Efficiency (OEE), which is the primary measurement for how much a given machine is utilized. Therefore, reducing the time for these three production stops will increase the OEE, and will be the primary purpose of this thesis project.
1.3 Purpose
The purpose for this project will be to identify problems in the production, and suggestions solutions in order to reduce the waste of resources related to die change. This can be done in a variety of ways, such as the reduction of machine downtime, or a decrease in ramp-up time that exists after maintenance or die changes.
1.4 Limitations
2 This thesis project will not include the following:
• Any related problems identified in areas other than the printing area.
• Any machine other than SI.
• Implementation of suggested improvements.
• Inventory control.
• Unscheduled maintenance.
Furthermore, the following assumptions will be made:
• The material delivered to the machine is of adequate quality.
Method
This section details the tools and terms used in this thesis project.
2.1 Lean
Lean is a management philosophy and a systematic method for optimizing production by reducing waste without sacrificing productivity. To accomplish this, Lean principles include the use of several tools, such as SMED, 5S, standard operating procedures, visual management, kaizen and the seven wastes. According to Lean principles, focus should be on the big picture, not just on one part of the company or machine. Solving a problem in one area of the company doesn’t necessarily result in an improvement for the company as a whole. With a focus that is too narrow, one risks solving a problem in one area, only to end up causing new problems in other areas doing so; i.e. the “push down, pop up principle”. The Lean process also emphasizes the necessity of every employee following the newly implemented rules, otherwise the whole process is a waste of time and money. This is often the biggest challenge. (Bicheno & Holweg, 2016)
2.2 Kaizen
Kaizen is a principle for continuous improvement through smaller incremental improvements of each part of an operation. This is done through the use of the PDCA-wheel. The theory behind Kaizen is that larger improvements are difficult to implement due to the fact that they involve many people and several systems. This can also complicate evaluating the results. As a result of this, it is preferable to make small improvements, but more frequently.
2.3 Single Minute Exchange of Die
Single minute exchange of die (SMED) is a method for reducing the die change of a machine.
This is done by dividing the die change into “internal” and “external” activities; where internal activities require for production to be stopped entirely, while external activities can take place during production; i.e. before the old production is stopped or after the new production starts.
Once activities have been categorized, internal activities are minimized by converting them into external activities, as well as reducing waste activities. Finally, external activities are reduced in order to enable the production of small batches.
2.4 Standard operating procedures
A standard operating procedure (SOP) is a set of step-by-step instructions for routine operations.
Lean principles encourage writing a SOP for every machine, with instructions for every step of the die change and/or for when the machine is running. Through this, efficiency and consistency are improved, while miscommunication is reduced.
4
2.5 The seven wastes
In Lean, the seven wastes, also called “Muda” are defined as “Anything beyond the minimum amount of equipment, material, parts, space, and worker’s time absolutely essential in order to add value to the product”. The seven wastes that create no value to the customers are:
• Overproduction
• Waiting
• Transport
• Unnecessary motions
• Over-processing
• Unnecessary inventory
• Defects
Some of these wastes, such as defects and waiting, are impossible to reduce to zero. However, every little improvement is a success for the company, and results in money saved.
2.6 Ishikawa diagram
An Ishikawa diagram, also called fishbone diagram or cause and effect diagram, is a method used for visualizing the causes of a problem connected to a specific machine or part of an operation. It is a useful tool for breaking down a big problem into its many root causes. Once that is done, it is easier to find the original cause of the big problem by applying 5S on the root causes. A common way to find the root causes is to start from the six Ms. These are the Machines, Materials,
Methods, Measurements, Mother Nature (Environment) and Manpower (People).
2.7 5-Whys
The ”5-whys” approach is an iterative root cause analysis tool that aims to find the original cause of a specific problem through repeatedly asking “why” the problem occurred. (Serrat, 2009) The answer to this question then becomes to subject of the next, until a cause has been identified, whose elimination will prevent the error from occurring again. (Brożyńska, et al., 2016) Usually this is done 5 times, hence the name, but sometimes fewer questions are needed, and sometimes more.
5-whys is useful because most problems are at the end of a long chain of events, where solving the problem at the start of the chain solves every subsequent problem in the chain. The method helps isolate that problem, and ensures that work is not wasted on solving problems that didn’t need solving.
An important fact to keep in mind regarding 5-whys is that “people don’t fail, but processes do.”.
W. Edwards Deming himself once said that “96 % of problems lie with the system, and the remaining 4 % with the people.” (Bicheno & Holweg, 2016)
2.8 6 Wise Men
The 6 wise men analysis is a method for solving problems through defining a set of 6 perspectives regarding a particular problem. These are: What, Why, When, How, Where and Who. By answering these 6 questions in advance, problems can be identified and avoided ahead of time.
2.9 PDCA Cycle
The PDCA-cycle, also known as the Deming Cycle or Shewhart Cycle, is a four-step process for continuous improvement used in many different areas. It consists of the following steps:
• Plan: The first step. Here, the project is scoped, lines of communication are established, and discussions are held in order to reach a consensus about the goal of the project.
Potential solutions are suggested, and root causes are identified.
• Do: The implementation stage. At this point, it is time to start putting the plans in motion. This is done by executing tests and experiments.
• Check: The validation stage. This is done in order to see if the implemented changes actually had the desired result, and if not, why? This also checks whether an observed improvement in the process is stable or a product of the Hawthorne Effect.
• Act: The final step, where the new and improved process is standardized, and systems are put in place to make sure the process does not revert to a previous stage.
Once this cycle has been completed, it starts all over again with a new area of investigation, resulting in a steady and continuous improvement.
2.10 Brainstorming
Brainstorming is a method for generating ideas. It “combines a relaxed, informal approach to problem solving with lateral thinking.” (Mind Tools Content Team, 2017) Simply put, it is about creating an atmosphere where the participants can feel comfortable throwing around both practical and crazy ideas, and attempting to combine them in order to craft unique and useful solutions.
Brainstorming is done in sessions of about an hour, and usually has a designated leader that attempts to guide the discussion and employ various techniques to increase the amount of ideas generated.
6
During these sessions, it is important to avoid talking negatively about any of the ideas that are presented, as this can affect the atmosphere that is at the core of brainstorming.
At the end of a brainstorming session, the ideas generated are evaluated, and a few are singled out for further analysis. (Företagande.se, 2010)
2.11 Overall Equipment Efficiency
Overall Equipment Efficiency, or OEE, is a key performance indicator that measures how well utilized a given machine is. It is calculated from three factors:
• Availability (A): what percentage of the planned production time is actually utilized for production.
• Performance (P): what percentage of maximum speed the machine is operated at.
• Quality (Q): what percentage of the produced product is actually shipped to customer as opposed to discarded due to quality concerns.
Combined, this gives the following formula:
• OEE = A * P * Q
2.12 Application of methods
The workflow of this project will begin with a thorough investigation of the current situation, as well as the recent past of the processes in focus. This will lead to the identification of key areas for improvement, which will then be visualized using an Ishikawa diagram, with “Where does the die exchange process lose efficiency?” as the main question to be answered.
The identification process will be conducted in two parts, using the theory of 7 wastes as a basis for the process, and Brainstorming as the method for generating ideas.
The first part will involve going through the individual parts of the SMED process, and generating ideas regarding where time is lost due to inefficiency.
The second part will focus on the process as a whole, and generate ideas regarding where time is lost due to confusion and unnecessary movements.
Once these areas have been identified, the PDCA methodology will be applied to the Ishikawa diagram as a whole, with the goal of reducing the time it takes for the die exchange to be completed.
Analysis of current state
In this section the current situation, including the method of operations during die changes, is presented. The data used in this section has been gathered through interviews, observations, and information provided by the company.
3.1 Overall picture of the production flow
The company employs a pull-based production system that produces banknotes. These
banknotes are produced using several different machines that each print a specific feature, most of them for the sake of security and preventing forgery.
The production facility is comprised of a banknote printing unit, a paper mill and a set of administrative units. However, this thesis will deal exclusively with the printing unit.
The machine observed are relatively far ahead in the process (See Figure 1). Most of the products go through all of the machines in the production flow, but there are some products that do not need to be processed in NS and/or LACK. If this is decided by the customer, the products just skip these steps and go directly to the next machine in line.
If all steps of the die change work as intended, and the operators receive all the materials they need from other departments in time, the die change is calculated to take about 20-25 hours.
Currently, this happens very rarely, and die changes often take up to 40-50 hours.
Figure 1. Describes where is the process the machine is located.
Cotton
preperation Papermaking Printing Post-
processing
SUSI NS SI SOI SCNU NOP
8
3.2 Object of study
The machine that has been studied is a Super Intaglio, produced by KBA-NotaSys – a company that produces printing machinery especially tailored towards creating high-security banknotes and documents.
The Super Intaglio utilizes the Intaglio printing technique in order to create sharp, textured images on the target paper. This is done by incising the image onto the surface of a plate using acid or tools. The surface is then coated with ink, and wiped clean so that the ink remains only in the incised areas. Finally, paper is pressed onto the surface, transferring the indentations and ink from the plate to the paper. (Wanczura, u.d.)
The machine is operated by three staff members referred to as “printers”. These manage the machine on a daily basis, and make sure it runs smoothly. The machine runs 24 hours per day.
During weekdays, this is split up in three shifts of 8 hours: 06:00 – 14:00, 14:00-22:00 and 22:00- 06:00. During weekends, it is done in two shifts: 06:00-18:00, and 18:00-06:00, starting on Saturday mornings. These shifts are staffed by five different teams, each consisting of three printers.
3.3 Die change process
During the die change process, each of the three printers have a different set of tasks to perform.
Currently, there are no descriptions of how these tasks are to be performed. This has resulted in each team having a different method to complete them. As such, there is no accurate data on how much time these tasks take to perform. The process, along with an estimation of how much time each task takes, can be seen in chapter 9.5. Studying this, it becomes clear that the printer designated “printer 3” has a set of tasks that take less time than the others. As such, the total time it takes for the die change to be completed is dependent on the sequence of tasks completed by printer 1 and 2.
Die change process
Printer 1 Printer 2 Printer 3
Rengöra färgverk Cleaning the inking device
Rengöra färgverk Cleaning the inking device
Banda ihop sista pallar och rapportera
Tie together the last pallets and report
Uttag wiping och rengöra grottan
Remove the wiping cloth clean the “cave”
Uttag schbloner och rengör cylindrar
Remove the templates and clean the cylinders
Packa ihop Makpall, rapportera i Prosit, plomba/följesedel och
lämna till logsitik Package the revocation pallet, report, plomb, add a waybill and
turn over to logistics.
Uttag plåtar och rengöra cylindrar
Remove the plates and clean the cylinders
Uttag plåtar och rengöra cylindrar
Remove the plates and clean the cylinders
Rengör färgpumpar och byt fat.
Rapportera kvarvarande färg i IFS.
Clean the inkpumps and change the inkdrum. Report amount of
remaining ink Byt dukar
Change the cloths Byt dukar
Change the cloths Ta fram SMED-vagn Bring out the SMED-trolley Nolla linjaler och sätt i plåtar
Reset rulers and insert plates
Nolla linjaler och sätt i plåtar Reset rulers and insert plates
Slå för intagspall och ställ in iläggare
Smoothen out the input pallet and configure the feeder Nolla linjaler och sätt i
schabloner Reset rulers and insert
templates
Nolla linjaler och sätt i schabloner
Reset rulers and insert templates
Säkerställ att arket får bra gripartag i svänggriparen Confirm the pivoting gripper gets a good grip on the sheet Nolla fronthållare
Reset frontal grip Nolla fronthållare
Reset frontal grip Grovställa avläggare Roughly adjust the layer Justera arkbanan och
kontrollera gripartag Adjust the sheet path and
confirm grip
Justera arkbanan och kontrollera gripartag
Adjust the sheet path and confirm grip
Kontrollera IFS/Prosit/ProTak Confirm IFS/Prosit/ProTak Ställ valsar och tryck
Adjust cylinders and pressure Ställ valsar och tryck
Adjust cylinders and pressure Kontrollera Make-Ready sensorn Control Make-Ready sensor Rikta schblon mot plåt
Align template and plate Rikta schablon mot plåt
Align template and plate 5S Rond 5S Round Ställ färger och finlira
Configure ink and improve settings
Ställ färger och finlira Configure ink and improve
settings
Köra iläggare och finlira avläggare
Run the feeder and configure the layer
Godkänning
Approval Godkänning
Approval Intagspallar körda
Run input pallets Ställ in NotaSave
Configure NotaSave
Ställ in NotaSave Configure NotaSave
Kör iläggare och slå för upplaga Run the feeder and smoothen
the pressing Starta produktion
Start Production Starta produktionen
Start Production Starta produktionen Start Production During the production
Särtryck/Godkänningsark
Offprint/Approvalsheet Särtryck
Offprint Rapportera avvikelser
Report deivation Veckans FU Weekly FU Tabel 1. All moments during the die change.
10
Problem areas
In this section the problems observed during the die change are presented.
4.1 Differences in methodology when performing tasks
Several operations during the die change are performed differently depending of which shift is working, and which person is performing it. This creates problems for new employees, as they can only learn from the other printers in the same shift. It also makes it difficult to generate any accurate data on how much time is required for each operation, making it impossible to
determine where in the process most time is wasted. An example of this was when a shift removed the finger protection during fastening of the printing plates. The next shift that then arrived shortly after had no idea how to put them back, which resulted in the process taking more time than was necessary.
4.2 Unnecessary items in the workplace
The surfaces and areas around the workplace are currently filled with needless, unutilized papers and objects. This makes it seem very cluttered, which can make things hard to find, and has the potential to cause stress-related errors in judgement.
4.3 Backsliding
While studying the workspace around the designated machines, it was revealed that many previously implemented changes and solutions to problems had stopped being used over time.
The skill and knowledge gained in the organization over time had been lost due to a lack of proper documentation. This was found to be caused by a lack of feedback between the production floor and management about whether or not implemented solutions and
improvements were actually being utilized, resulting in any effort on the part of management to be entirely wasted, as it did not result in any meaningful change in worker methodology.
4.4 Unnecessary call for the maintenance staff
During the observation of the die change, the operators were unable to reach the goal pressure of the machine. They tried to solve it themselves, but failed. That led to the mechanics being called in, in hopes that they could solve it. The mechanics tried, but could not. When the time came for shift handover, the new shift solved the problem immediately. One operator noticed that the temperature was set to 60 degrees, when it should be 80 degrees. They adjusted the temperature, and once the machine had heated up, the pressure reached the goal value. After that, the new shift had to readjust all settings, wasting another 2-3 hours. Based off of data from interviews, mechanics are generally called at least once per die change.
12
4.5 Shift handover
When shift handover occurs in the middle of an operation during a die change, the shift taking over will often redo the entire operation or parts of it. This is due to the fact that the new shift is uncertain whether the shift before has performed every step correctly. Because of the fact that there is no standard operating procedure for the die changes, there is no established “correct”
way to perform the die change, and every shift performs the change in their own way. When a new shift takes over, they thus redo the operation using their own method in order to make sure it is performed correctly.
4.6 Unoptimized die change process
Because there are no Standard Operating Procedure established for the individual moments in the die change process it is impossible to determine just how much time each moment should take. In addition to this, the distribution of work between the three printers is uneven, resulting in printer 3 having tasks equal to roughly 11 hours whereas printer 1 and 2 have tasks equal to roughly 23 hours. Furthermore, as the moments have not been thoroughly documented, there may be activities in each moment that could be done in preparation for a die change or after the die change, as opposed to being done during it.
4.7 Machine downtime due to operator absence
Machines can only operate when there are two or more knowledgeable operators. This results in complications related to meal breaks and personnel shortages. Often, only two operators man each machine, leading to the machine having to shut down entirely when one or both of them leaves for lunch or dinner. This particular scenario, per example, result in an average of 40 minutes of lost time. This causes a loss of production while operators are away as well as while the machine is ramping back up to full production. If a personnel shortage arises, the machine will often be offline for up to 8 hours.
Improvement suggestions
In this section, solutions to the issues discussed are presented.
5.1 Difference in methodology when performing tasks - Standard
In order to make all the printers perform every operation during the die change the same way, there needs to be detailed instructions to follow. There are currently five different shifts working with the machine, and they are all performing several operations in different ways. This creates problems for new employees, as they can only learn from the other printers in the same shift.
One particularly clear example of this is the pattern plates. Currently, some shifts clean the paint off of these plates when taking them out during the die change process. This is entirely pointless, as they are immediately discarded once they are used.
In order to make all the shifts perform every step the same way, it is recommended to introduce a Standard Operating Procedure. This document should specify in detail how every step is performed.
Implementing these will give the company a solid foundation on which to begin implementing improvements, and will simplify the introductory period for new hires.
5.2 Unnecessary things in the workplace – 5S
In order to improve the structure of the workplace, it is recommended to utilize the 5S-theory.
Below follows a suggestion on how this could be implemented:
1. Sort: Begin by setting aside a table where all items can be collected (this includes
everything in the locker, on the desk and in the desk drawers and all papers on the tables), along with three sorts of tags (green, yellow and red). Each shift should then place these tags on each item on the table and write their shift number on it, where the chosen color indicates how useful it is to them. Red tag means useless, green tag means useful and yellow tag indicates uncertainty.
Once all items have been tagged appropriately, all items with yellow tags should be investigated to determine whether they should be converted to red or green.
The items marked exclusively with red tags should immediately be removed from the workplace, while items marked green by every shift should be set aside. This leaves only items with a mix of red and green tags. These should be marked with a special tag that indicates they are only used by some shifts.
2. Set in order: Once the sorting phase has been completed, the remaining items should be group up according to function and positioned close to where in the workplace they are used. It is important that each item has a unique place for when it is not currently being
14
used. Start by decide what is used daily, what is used weekly, what is used monthly and what is used only during a die change. Items that are only used during a die change should be grouped together and placed somewhere out of sight. For example, in the locker, as these items will be used relatively rarely. When everything has a unique
placement, every placement should be marked with a piece of tape which indicates what item belongs there.
By the machine labelled SI, there are currently four small whiteboards, and one large whiteboard, which is used to detail the die change procedure. The small whiteboards have no specified use, but have various procedures and documents posted on them. The problem is that many of these documents are either outdated or unused. This can lead to confusion for the workers as they do not know what should be followed and what should not. One of the interviewed employees even revealed that they had no idea what was on the boards. It is recommended that all outdated documents posted to these whiteboards are removed and archived, and replaced with updated versions where necessary.
Furthermore, there are a set of shelves above the desk that contain binders that are rarely used. These should instead be moved to drawers so the space can be used for more current and useful items.
It is also useful to whenever possible try to move items off the floor and up on the walls to maximize space. For example, the coffee table could be replaced with a shelf, as there is no need for it to ever be moved. Clearing as much floor area as possible gives
employees an easier time to move items around the production floor.
There are currently two printers connected to the computers by SI, with one being connected to a standard workstation and the other connected to the computers that control the production machine. Removing one of these printers would further save space, as there is no need to have two things performing the same function in the workspace.
It is also recommended to put all documentation related to the computers in a binder and place it next to the machine control station.
The pallet delivery notes that is on the desk should each have separate folder on the wall above the desk.
To save space in the locker, the employees boxes could be half the size as they currently contain items that shouldn’t be there. For example, several of the workers have put in a packet of plastic gloves in their boxes. These should have their own unique placement.
3. Shine: A round of cleaning should be done before every shift handover, and all consumables should be checked to see if they need refilling. Anything found lacking should be noted in a designated place on one of the whiteboards. Once a week, a more thorough cleaning should be conducted based on a checklist which should be readily available to the employees.
4. Standardize: To facilitate the cleaning and establish a routine, there should be pictures above every area to be cleaned that shows how it should look like when the cleaning is completed.
5. Sustain: The cleaning routine should be checked during shift changes. It is important that every shift leader control these checklists and make sure they are being followed.
5.3 Backsliding
The proposed solution for handling backsliding is to implement a stricter documentation and accountability process, that clearly indicates where backsliding was occurring. The system would attempt to secure the implemented solutions and changes in the long-term by clarifying how checklists and similar documents would move through the organization. Based on these needs, a Product Specification was created (See Appendix 1). From this specification, the Accountability System was created.
The Accountability System is a system that defines how documents would move through the organization, and who would be accountable for a given document at a given point in the chain.
The system serves as a guideline for how to design and utilize documents in a way that makes them work with the rest of the system, and becomes a basis for implementation of future improvements. This is done by assigning every improvement or solution one or several associated Points of Control that, upon being controlled, will reveal whether or not a given solution or implantation has been utilized. These Points of Control are then gathered in the form of a checklist, which gets assigned a User, a Controller (if necessary) and a Collector.
User
The User is the person who uses the checklist to perform a given task.
Controller
The Controller is the person that makes sure that the checklist has been utilized for a given task.
This should ideally be someone unrelated to the User (per example, for a checklist that ensures good communication at a shift change, the controller would be a person from the arriving shift), but can also be a co-worker at the same station.
Collector
The Collector is the person that gathers the document and archives it, so that it can be used to create statistics on how a given change has affected Key Performance Indicators in the
production.
16 This results in a circle of accountability,
where each person in the chain requires the previous person to perform their task in order to perform their own task. If the User doesn’t use the checklist, the Controller will notice. If the Controller hasn’t signed the checklist, the Collector will notice. If the Collector hasn’t created any visualizations from the data, the User will notice.
In order to work, a system like this requires each document that is a part of it to be designed in a manner that makes it clear how this chain looks for that particular
document, and what stage in the chain a document is in at any point. Based on this, another product specification was created (See chapter 9.2) for a checklist template.
The resulting template (See chapter 9.3) contains clear fields where the names or titles of the User, Controller and Collector are listed. It also lists where the checklist should be stored until the Collector gathers it, in addition to any other system-specific information that is required (such as document ID). The method for describing the process is derived from a 6 Good Men analysis, and lists what should be done, how it should be done, and why is should be done. It also lists why the checklist exists, and when it should be utilized. Through this, all 6 perspectives of the 6 Good Men analysis are covered. Each field that needs to be filled in prior to usage has also been marked with yellow to ensure it is not missed.
Finally, to tie it all together, a communications plan needs to be in place to follow up on any misses that occur in the chain. If a checklist is missing, it needs to be followed up on to ensure consistency in the system. It should also include a schedule for when the collector gathers any relevant documents, and what sort of visualizations should be created as feedback based on these checklists.
Implementation of this system will result in a clear line of accountability, which gives
management a clear indication whether implemented solution or improvements are actually being utilized or not, and will form a basis for future continuous improvements.
Without a good system for following up and monitoring implementations, it can never be certain that the work management puts in actually translates to an increase in product quality.
5.4 Unnecessary call of the mechanics - Checklist
To solve this problem, one possibility is a checklist for operators to go through before mechanics are called. Through this, common causes of problems can be identified and eliminated by the operators themselves. The goal of this is for mechanics to be called only when their specific expertise is required, or when there is a task that only they are allowed to undertake. This means
Collector gather and adds any data points to statistics and creates visualisations
User receives feedback thanks to visualisations
User fills in checklist Controller
ensures checklist is properly used
that the mechanics’ competence and skills will be utilized fully, and reduces the waiting times for help.
An example of such a checklist can be seen in chapter 9.10. The list is intended to be an ongoing project, where operators and their shift leaders can add things depending on new situations that arise, or changes to the process.
5.5 Tasks are redone at shift change
To prevent moments being wholly or partially redone during shift changes, it is worth
investigating if the SMED-process can be split up into smaller parts. This is also an area where implementing Standards of Operation will help, as it will make it clearer what has and has not already been done.
Furthermore, improving the communication between shifts that occurs during shift changes should also reduce the amount of work that is redone during the SMED-process. The recommended solution for this is the implementation of a checklist with a list of topics that should be discussed during the shift change. This checklist should be specified using the process described in chapter 9.2.
An example of such a checklist can be found in chapter 9.11.
5.6 Unoptimized die change process
To optimize the die change process (see chapter 9.5), it is necessary to implement SOPs. Without these, it is impossible to know how much time each task in the SMED-process actually takes.
Implementation of SOPs will allow for further optimization work to proceed, which will result in a lower overall die change time. Furthermore, the SMED-process should be investigated
thoroughly with the goal of finding partial tasks that can be done in advance while the machine is running, or tasks that can be transferred from printer 1 and 2 over to printer 3.
One such task is the changing on the cloth that presses the paper onto the plate. This process is not dependent on any of the other tasks, and while it would take a bit more time, it can be handled by a single person. Moving this task over to printer 3 would shorten the overall time required to complete the entire process.
Additionally, it is recommended that the printers log the progress of the SMED-process in ProTAK. This would be done by adding a comment for when each moment of the process starts and ends. This is, per example, already done with breaks and when the machine is being cleaned.
If the time for a particular task takes considerably longer than what is specified in the SOP, it should be investigated as an anomaly.
Another suggestion is to implement a checklist where the printers can fill in at what time each task in the SMED-process starts, and what time it finishes. As tasks are sometimes started by one shift and finished by another, the checklist needs to allow for each shift to sign which tasks they start and which tasks they finish, along with the option to add comments regarding for far into each task they got before handing over to the next shift, or if any problems arose during the task.
18
An example of how this checklist could look can be found in chapter 8.8 and chapter 9.9.
Another problem with the die change process is that it currently is considered completed when the machine is able to produce a single sheet of banknotes that is of adequate quality. However, even after this is done the printers still modify and do small changes to the colours and pressure of the machine in order to make sure it keeps printing well. This is done at a lower speed than what the machine is capable of. This is referred to as the “ramp-up time”. This time is not considered to be a part of the die change process, and therefor there is no data collected on how long it should take or what circumstances cause it to take longer than needed.
In order to solve this problem, it is recommended to add a final task to the SMED-process labelled “full production”, which becomes completed once the machine reaches full capacity.
This turns the otherwise unspecified time into a part of the die change process, and as such becomes the responsibility of person in charge of that. This will allow for easier data collection and turns reaching full production capacity into a more visual goal for the printers.
Another way to save time on the die change process is to purchase an additional set of certain parts that require cleaning during the die change process. This will enable the printers to simply swap out the dirty parts for another set of pre-cleaned parts, and save the cleaning process until later or move the workload over to another part of the company. The parts that have been identified as candidates for this are:
• The Paint Drip Basin.
This part is what collects the paint that drips off the plates during production.
• The Gables.
This part is what locks the seal for the plates in place.
• Stirrers.
This part stirs the paint around to make sure it keeps an even texture.
• Plate seals.
This part locks the plates onto the cylinders.
• Screws.
The screws used to lock the seals onto the cylinders.
These should be stored somewhere near the machine for easy access during the die change.
In addition to this, it was observed that the printers reset the liners for the plate seals during the insertion of the printing plates. This could instead be moved up to be done during the extraction of the previous printing plates. This would not necessarily save time, but would rather lessen the impact of other delays.
Finally, a new method of resetting the front holder allows it to be done by one person as opposed to the previous two. This means it can be moved from being a task done by printer 1 and 2 to being a task done by printer 3.
The new die change process that incorporates these changes can be found in appendix 6.
5.7 Machine downtime due to operator absence
A possible method for reducing interruptions in production is for the company to train the operators to use several different machines. This reduces the necessity of having to shut down production during meal breaks and in the case of personnel shortages, as operators from other machines could take over the tasks of the operators that are absent. Despite the fact that most machines require significant amounts of training to operate, the company still stands to gain from this in the long run; especially for the bottleneck at machine 3. Machine 3 is shut down at least once a week due to personnel shortages. If more operators are taught how to operate it, that problem would disappear entirely.
In order for the shift leader to be able to assign personnel to where they are needed the most, a competence matrix is to be put up on the wall (See chapter 9.4). The matrix shows which machines each operator is trained to use, and to what degree.
Another suggestion is to let the printers choose which machines they are to learn next. It is impossible for all the printers to learn every machine, but letting printers learn three or four machines each should yield a notable improvement in machine uptime. Rotating employee on different machines should also result in improved cohesion among the staff as they gain more understanding of each other’s circumstances.
However, as it stands, there exists a “us and them” feeling in the production department.
Everyone does their own thing, with little help being provided across different positions even if opportunity exists. Moving staff around will let them work with new people and on different machines which should reduce that feeling. The company has to make the printers all work towards a common goal, which is to maximize the output to post-processing. Right now, most of them have that goal, but only for their own machine, as opposed to looking at the production at large.
20
Economic calculations
To show an example of how much time and money the company can save, the OEE and ROI were calculated for a particular change. The chosen change was that the company should train their printers (chapter 5.7) so that they will not have to shut down the machine during lunch breaks when they are not fully staffed.
6.1 Current values
The company employs two different methods of calculating the OEE: one with the die change times included, and one without. Normally it is only done with the die change times included.
The calculation without die changes are done in order to clarify the effect unscheduled
maintenance has on the production, as it would otherwise be dwarfed by the die change time.
In this case, the OEE-value that includes the die change process will be investigated, as it is this process is the topic of the thesis.
The current values can be seen in Table 2 and Table 3 nedan. Currently, there is a large difference between these values. This is primarily due to the massive difference in availability. The OEE- value without die change times was calculated based on production during 10 days from 14/10- 17 to 24/10-17, and the OEE-value with die change times was calculated based on production during a month from 1/11-17 to 1/12-17 and includes 3 die change processes that total to 93 hours and 46 minutes.
With die changes included
Availability Performance Quality OEE
27.8 % (70 %) 85.4 % (90 %) 97.3 % (99.2 %) 23.1 % (62.5 %) Table 2. Current OEE-calculation with die changes included.
Without die changes included
Availability Performance Quality OEE
63.1 % (70 %) 87.6 % (90 %) 99.2 % (99.2) 54.9 % (62.5 %) Table 3. Current OEE-calculation without die changes included.
As evidenced by these numbers, the quality and performance are at roughly the same levels during both of these time periods. The main factor that drives the differences between these two calculations is the Availability.
22
6.2 New values
The current values can be seen in Table 4 and Table 5 nedan. After the suggestion has been implemented, OEE should improve from 54.9 % to 58.3 % when die changes are excluded from 23.1 % to 24.2 % when die changes are included.
With die changes included
Availability Performance Quality OEE
29.2 % (70 %) 85.4 % (90 %) 97.3 % (99.2 %) 24.2 % (62.5 %) Table 4. New OEE-calculation with die changes included.
Without die changes included
Availability Performance Quality OEE
67.1 % (70 %) 87.6 % (90 %) 99.2 % (99.2) 58.3 % (62.5 %) Table 5. New OEE-calculation without die changes included.
6.3 Return of Investment
In order to further visualize the value of reducing machine downtime, a Return of Investment (ROI) calculation was performed (See Table 6), detailing the monetary cost and gain from educating more printers on how to operate SI.
As SI is the bottleneck of the system, the downtime cost for the machine is equal to the
downtime cost of the entire system. According to the economy department, this equals to 50 000 SEK per hour. And an hourly cost for a worker is 300 SEK. During the 10-day conducted during the calculation of the OEE-value, it was determined that the machine has roughly 56.5 minutes of downtime due to preventable lack of printers per day. During the course of a year, this means that SI has a total of 324 hours of preventable downtime.
The estimated time required for education was determined to be 6 months. However, as the printers also have other duties to perform, it was estimated that they could only spend 50 % of their time on this education. Which means they will be fully trained after 12 months. As such, the time required becomes 855 hours over a time period of a year per employee. As the company employs 5 shifts, this means that a minimum of ten printers need to be educated (two in each shift). The cost for this was calculated to be equal to the wage cost of the employees being educated.
The calculation was performed on a 2-year basis, and assumes that a newly educated printer starts being useful after the entire educational period has been completed. In reality, the printer can probably begin assisting after 5-6 months.
𝑮𝒂𝒊𝒏 𝒐𝒇 𝒊𝒏𝒗𝒆𝒔𝒕𝒎𝒆𝒏𝒕 − 𝐂𝐨𝐬𝐭 𝐨𝐟 𝐢𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭
𝐂𝐨𝐬𝐭 𝐨𝐟 𝐢𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭 ROI
324 ∗ 50000 – 300 ∗ 855 ∗ 10
300 ∗ 855 ∗ 10 5.3
Table 6. ROI-calculation.
24
Conclusions
The main problem identified during this thesis project was the lack of standardization and data collection in the production. The foundations for continuous improvements are not quite there, which creates problems for both the printers and the management, as they cannot truly know whether any implemented improvements or changes have any real effect on the production.
However, implementing a stricter documentation process and standardizing each task in the SMED-process should give the company the basis it needs in order to truly begin improving the efficiency of the production. It should be stressed that any other attempts at optimizing the SMED-process cannot produce tangible or reliable results until these problems are solved. It is therefore our recommendation that this work is given the highest priority.
Despite this, there are several things that can be improved upon even though Standard Operating Procedures are not implemented. Training the printers, for example, on other machines can be started. This will lead to a higher uptime on low-capacity machines. This should increase the OEE of SI by 3,4%. A 5s-procedure can also be done to reduce the clutter of the workspace, and the changes to shift change procedures can be implemented.
7.1 Summary of Suggested Improvements
In summary, the suggested improvements are:
• Implementing SOPs for all tasks during the SMED-process.
• Implementing the Accountability system as described in section 5.3.
• Performing a 5S-procedure to clean up the workspace.
• Training printers on how to work other machines and constructing a competence matrix.
• Reorganizing the SMED-process to make the workload more even between the printers.
• Purchasing additional sets of parts that are cleaned during the SMED-process.
• Implementing a checklist that details what information should be communicated during a shift change.
• Implementing a checklist detailing things to check before calling in the mechanics.
7.2 Discussion
The chosen process and methodology worked well for investigating the process and generating solutions. However, a lot of time had to be spent on identifying problems. This could have been done more efficiently and systematic.
The interviewing process could have been done more efficiently. Most of the information we used from the people we talked to was not documented in any systematic manner, but was rather just written down as notes in various notebooks. A more thorough investigation into how
interviews are conducted and better documentation procedures would have given more useful information and better references for the rapport.
Furthermore, it was very difficult for us to fully observe the changeover process as it took over 24 hours. This means that a large part of it was done during the night. As we were not allowed to
26
film the process due to regulations, we were unable to ever observe a changeover from start to finish. There were also not that many opportunities for us to observe the changeover procedure.
7.3 Future work
• Sometimes the ink is to be produced in time. This leads to the die change process not being able to be completed, and the printers need to wait until the ink arrives.
• Due to the problems stated regarding observation of the die change process, the earlier parts of the process were only observed once. Due to this, there are likely more
improvements that can be discovered here.
References
Bicheno, J. & Holweg, M., 2016. The Lean Toolbox. 5th ed. Buckingham: PISCIE Books.
Brożyńska, M., Kowal, K., Lis, A. & Szymczak, M., 2016. 5xWhys. Method First Handbook. Łódź: 2K Consulting.
Företagande.se, 2010. Brainstorming är kreativt tänkande i grupp, övningar. [Online]
Available at: https://www.foretagande.se/brainstorming-kreativt-tankande-i-grupp/
[Accessed 8 November 2017].
Mind Tools Content Team, 2017. Brainstorming - Creativity Techniques from Mindtools.com. [Online]
Available at: https://www.mindtools.com/brainstm.html [Accessed 8 November 2017].
Serrat, O., 2009. The 5 Whys Technique. [Online]
Available at: https://www.adb.org/sites/default/files/publication/27641/five-whys-technique.pdf [Accessed 6 November 2017].
Wanczura, D., n.d. Ingalio Printmaking - artelino. [Online]
Available at: https://www.artelino.com/articles/intaglio_printmaking.asp [Accessed 30 November 2017].
Appendix
9.1 Specification of Accountability System
Creator Information
Product Information
Product Specifications
Example
Company: KTH Stockholm
Contact Person: Robin Hammar & Julia Björck Telephone: 072-5218941/076-9250199
E-mail: rhamm@kth.se/juliabjo@kth.se
The Accountability System is a system of checklists and similar tools that work together to create a concrete line of accountability throughout an entire process. This is done in order to maintain improvement and solutions that have been implemented in the process, and to prevent backsliding from occurring.
The idea of the system is to analyse every point where information is transferred, or where a new methodology has been implemented and determine a set of Points of Control that will serve as checkpoints to make sure that the new methodology is actually being utilized. These Points of Control will be collected and organised in the form of checklists, that will each have a controller and a user assigned to them. The user performs the checks specified in the checklist, and the controller makes sure that the checklist is filled out.
Orignal version: 16/11-17 Current version: 20/11-17
The system MUST:
• Assign the administrative burden evenly across the system it is applied to.
• Not be time-consuming.
• Clearly assign two people for each checklist: one controller and one user.
• Clearly describe who is responsible for each step in the handling of the checklists.
• Make sure that there is exists a personal reason for everyone in the chain to do their part.
• Visualise both how well the checklists are being utilized and what effect they are having.
The system CAN:
• Be either digital or analogue.
Checklist is needed for making sure communication is working from one shift to another.
What does the checklist contain? Topics that should be covered during the shift change.
How is the checklist visualised? By connecting to Key Performance Indicators.
When is the checklist collected? During Gemba walks.
Where is checklist stored until collection? In a compartment by the machine.
Who is the user? The leaving shift.
Why should they do it? Because the arriving shift needs the information.
Who is the controller? The arriving shift.
Why should they do it? Because they need the information.
9.2 Specification of Checklist Template
Creator Information
Product Information
Product Specifications
Company: KTH Stockholm
Contact Person: Robin Hammar & Julia Björck Telephone: 072-5218941/076-9250199
E-mail: rhamm@kth.se/juliabjo@kth.se
The checklist is a series of Points of Control that work together to ensure a specific regulation or procedure is followed. It does this by being part of a chain that makes sure that each
checklist is controlled twice, and then archived and used to create visualisations of the progress in the area the checklist affects.
The checklist template is a base for how these checklists can look. It will through its design make sure that it is clear who uses the checklist, who controls it and so on.
Orignal version: 20/11-17 Current version: 20/11-17
The checklist template MUST:
• Include a place for the controllers’ name or title.
• Include a place for the controller to sign the checklist.
• Include a place for the controlees’ name or title.
• Include a place for the controlee to sign the checklist.
• Include a place for the collectors’ name or title.
• Include a place for the collector to sign the checklist.
• Include a place for any other system-specific information (document ID, date etc.)
• Include clear steps on how to perform the check.
• Include a clear description for when the checklist should be used.
9.3 Checklist Template
9.4 Competence matrix
9.5 Current Die Change Process
CURRENT Before stop
- Printer 1 Time
(h) Printer 2 Time
(h) Printer 3 Time
(h) Rengöra färgverk
Cleaning the inking device
3
Rengöra färgverk Cleaning the inking
device
3
Banda ihop sista pallar och rapportera Tie together the last
pallets and report
0,5
Uttag wiping och rengöra grottan
Remove the wiping cloth clean the “cave”
2
Uttag schbloner och rengör cylindrar
Remove the templates and clean
the cylinders
3
Packa ihop Makpall, rapportera i Prosit, plomba, följesedel och
lämna till logsitik Package the revocation
pallet, report, plomb, add a waybill and turn
over to logistics.
0,5
Uttag plåtar och rengöra cylindrar
Remove the plates and clean
the cylinders
3
Uttag plåtar och rengöra cylindrar Remove the plates
and clean the cylinders
3
Rengör färgpumpar och byt fat. Rapportera kvarvarande färg i IFS.
Clean the inkpumps and change the inkdrum.
Report amount of remaining ink
3
Byt dukar Change the
cloths
0,33
Byt dukar
Change the cloths 0,33
Ta fram SMED-vagn Bring out the SMED-
trolley
0,17
Nolla linjaler och sätt i plåtar Reset rulers and
insert plates
4
Nolla linjaler och sätt i plåtar Reset rulers and
insert plates
4
Slå för intagspall och ställ in iläggare Smoothen out the input pallet and configure the
feeder
0,33
Nolla linjaler och sätt i schabloner Reset rulers and insert templates
2
Nolla linjaler och sätt i schabloner Reset rulers and insert templates
2
Säkerställ att arket får bra gripartag i svänggriparen Confirm the pivoting gripper gets a good grip
on the sheet
0,083
Nolla fronthållare Reset frontal grip
0,33 Nolla fronthållare
Reset frontal grip 0,33 Grovställa avläggare
Roughly adjust the layer 0,33
