Production Line Improvement at Bosch Thermotechniek
Carly Overmars - MSc Thesis Industrial Engineering & Managemen t
Public Version
University Supervisors Dr. Ir. Leo van der Wegen Dr. Peter Schuur
University
University of Twente Drienerlolaan 5 7522NB Enschede The Netherlands
Company Supervisor Bram Viskaal
Graduation Company Bosch Thermotechniek B.V.
Zweedsestraat 1 7418BG Deventer The Netherlands
i
ii
Summary
The master thesis in front of you is about the improvement of production lines at Bosch Thermotechniek. Bosch Thermotechniek is an industrial leader in heating and hot water products. The production facility located in Deventer, the Netherlands, holds multiple production lines that are dedicated to product families of boilers. The facility in Deventer competes with facilities in low-wage countries. Moreover, the assembly of boilers is labour-intensive and the organization has to deal with an aging work force. This thesis focuses on the TrendLine A production line and proposes ways to improve the production line.
The goal of this research is to propose changes to the production line, such that costs are decreased while productivity levels are at least maintained. Changes to the production line are restricted by a pay-back period of 3 years and ergonomic considerations. The research question is formulated as follows:
‘How should the TrendLine A production line be redesigned to decrease costs?’
The research question is answered in the sections of this research. The first section analyses the current situation and presents the current performance of the production line. At this moment, the TrendLine A production line consists of 11 workstations that assemble a boiler from start to finish with a cycle time of 238 seconds. A 12th workstation, located outside the U-shaped production line, prepares kits with parts that are supplied to the other workstations. All workstations are occupied by one employee and parts are supplied to the line by means of a manual Kanban system. In terms of performance the assembly line balancing efficiency (LBE) is 90.9%. Analysis of the internal KPIs over the past year evidences that bad performance is mostly explained by the number of hours used for training and boiler losses as a consequence of technical and material problems.
With use of literature, the current situation is analysed for problems and inefficiencies. The methods presented in the literature study are applied to the current situation. This analysis finds that the vertical balance of the workload over the workstations is not optimal. The better the vertical balance, the less unnecessary waiting time there is. Use of waste identification techniques shows that 45.9% of the total work content consists of non-value added activities. Including waiting time until the cycle time is reached, this is even 57.9%. The workstation with the most waste is Workstation 12 (kitting area).
Analysis of the production losses shows that 19% of the boiler losses is caused by reasons subject to this research. Of this total, 9% is caused by material shortages and 10% by assembly mistakes. A deeper look into the data, finds that material shortages are mostly due to Kanban related problems such as late deliveries. Factory-wide, almost 6,700 Euros was lost on boiler losses due to the Kanban system during the past year. Finally, the section on the current situation presents an analysis of the part allocation and picking flow on a workstation. A spaghetti diagram is used to see where waste occurs in the picking flow. The diagram shows that randomly allocating parts is not efficient.
The next section discusses the ideas to improve productivity. These ideas originated from discussions with colleagues, visits to trade shows and visits of companies to the production facility. The first improvement results from the creation of a line balancing application by the researcher. This application finds the near-optimal task division that maximizes the line balancing efficiency by minimizing the product of the cycle time and the number of workstations.
While performing the research the cycle time reduced to 232 seconds due to changes of the technical department. The cycle time is further decreased to 226 seconds with use of the line balancing application. The resulting task division and performance measurements are taken as the baseline for
iii evaluation of other improvement proposals. The performance measurements of the feasible ideas are presented in Table 0-1 and discussed thereafter.
Optimization problem Cycle Time (sec) # of Workstations LBE (%)
Current situation (CS) 232 11 91.4
Line balance CS 226 11 95.2
Line balance CS + partial removal kitting area 226 11 94.5 Line balance future situation with test of 188
sec. (FS)
219 11 97.8
Line balance FS + partial removal kitting area 216 11 98.0
Line balance CS + palletizing robot 219 11 97.3
Table 0-1 Summary of Line Balancing Application Results
The study evaluates the partial removal of the kitting workstation (Workstation 12) and improvement of part allocations in the racks. These latter improvements result from use of the 5S and spaghetti diagram methods. Moving certain parts to the production line, does not impact the cycle time of 226 seconds in the current situation, but it does save one FTE. However, next year, the technical department implements a new final test in the production line with a shorter duration. Recalculation of the near-optimal task division results in a cycle time baseline of 219 seconds. The study also examines this situation in combination with the improved part allocation and partial removal of the kitting area. Implementation of the part allocation changes further reduce the cycle time to 216 seconds. This shows that changes to the production line not always yield the same result when applied to different situations.
Secondly, the study examines the possible implementation of a palletizer robot. This robot should take over the task of palletizing the finished boiler from the employee at Workstation 11. This change achieves a cycle time reduction of 7 seconds compared to the baseline of 226 seconds.
Thirdly, the study looks into automation of the Kanban system. This change does not affect the cycle time as the process is independent from production itself. However, it does save boiler losses by preventing the loss of Kanban cards and material shortages from happening.
Finally, the study finds that the implementation of robots that can work in close collaboration with humans in assembly (cobots) is infeasible. As the assembly of the boilers requires a lot of different screws, nuts and bolts, the part supply of these parts already requires a significant investment without gaining time or productivity. Moreover, a cobot is not able to work on non-fixed locations. The boilers are sent through the production line on a carrier with a non-fixed location which makes it impossible for a precise cobot to be implemented. Other ideas that involve structured part supply are also infeasible due to the high cost for vibratory fillers and other part suppliers.
The economic and ergonomic evaluation of all beneficial proposals are presented in Table 0-2. The economic evaluation includes the cycle time reductions as calculated.
Optimization problem Pay-back period (time- value of money incl.)
Ergonomics (--/-/0/+/++) Use of line balancing application current situation (CS) 129 days 0
Line balance CS + partial removal kitting area 1 year and 341 days -/0 Line balance FS + partial removal kitting area 1 year and 226 days -/0 Line balance CS + Palletizing robot 3 years and 121 days +
Automated Kanban system 3 years and 316 days 0
Table 0-2 Economic and Ergonomic Evaluation
iv The study finds that the use of the line balancing application is highly beneficial. The partial removal of the kitting area including the improved allocations also has a pay-back period of less than 3 years.
Not only in the current situation, but even more when the new final test is implemented. Then, the palletizing robot violates the pay-back period of 3 years. This investment will most likely not prove beneficial as the structured supply of boilers to the palletizer robot has not been taken into account yet. It might be beneficial if the palletizer robot can serve multiple production lines at the same time.
However, this requires significant changes to the current lay-out of the plant. Finally, the implementation of an automated Kanban system also violates the pay-back period of 3 years.
However, this investment also involves some significant soft wins and additional chances to save money. First, if the Kanban system is automated, data is available on part level. This information can be used to reduce inventories and get insights into part use. Moreover, the system is very transparent and prevents further discussions between the production and logistics department. These soft wins are well worth the additional 316 days until pay-back. A combination of the palletizing robot and removal of the kitting area shows worse results than implementation of either one of the two.
To conclude, we recommend Bosch Thermotechniek to:
- On the short term: improve the line balance with use of the Line Balancing Application;
- On the short term: improve the part allocations in the racks and partially move the parts from the kitting area to the production line;
- On the medium to long term: implement an automated Kanban system;
And, moreover:
- Study the possibility of a palletizing robot that serves multiple lines;
- Study the possibility of fixed sub-assembly locations and reduction of the number of parts used for sub-assemblies. This would make the implementation of cobots possible.
v
Preface
Dear reader,
With great pleasure I present to you my Master Thesis. The final hurdle I had to take before obtaining my master’s degree in Industrial Engineering and Management. After a visit to the production facility of Bosch Thermotechniek in 2016 I knew where I wanted to write my Master Thesis, and so it happened.
I want to express my gratitude to my supervisors Leo van der Wegen and Peter Schuur who have provided me with a lot of constructive feedback. They helped a great deal with the structure of my report and steered me in the right direction whenever I lost sight of the most important content. Your exquisite eye for detail and text-structuring greatly improved readability.
I would like to thank the people from the TEF department at Bosch for welcoming me and making time for me when I needed help. I also want to thank you for guiding me towards the correct people within the larger organization. I’ve learned that it is quite a challenge to work in such a big company. There is a lot of information available, but it is a challenge to get a hold of it. Special thanks to Bram Viskaal for creating this thesis opportunity.
Most of all, I would like to thank my family and loved ones for their endless support. Not only did you provide me with information that I couldn’t get a hold of myself, but you were also willing to discuss my thesis and listen to my concerns whenever I needed it.
Finally, I would like to thank my university colleagues without whom I wouldn’t have had such a great study time. I especially want to thank Thom, Niek and Sebastiaan for finishing the degree with me.
Working with you was a pleasure!
All that remains for me is to wish you a pleasant read.
Carly Overmars
Deventer, January 21, 2019
vi
Contents
Summary ...ii
Preface ... v
List of Abbreviations ... viii
List of Figures ... viii
List of Tables ... ix
1 Introduction ... 1
1.1 Company Background ... 1
1.1.1 Bosch Thermotechnology ... 1
1.1.2 The Deventer Production Plant ... 2
1.2 Research Plan ... 3
1.2.1 Problem Description ... 3
1.2.2 Objective of the Research ... 3
1.2.3 Research Scope ... 3
1.2.4 Research Questions ... 4
1.2.5 Deliverables ... 6
2 Current State and Performance of the Production Line ... 7
2.1 The TrendLine A Production Line ... 7
2.2 Detailed description of the workstation operations ... 8
2.3 Supporting practices ... 11
2.4 Current Performance ... 12
2.4.1 Takt Time ... 13
2.4.2 Performance Measurements ... 15
2.5 Conclusion ... 20
3 Literature Review... 21
3.1 Lean Management ... 21
3.1.1 Muda, Muri and Mura ... 21
3.1.2 5S and Spaghetti Diagrams ... 22
3.1.3 Kanban ... 22
3.1.4 Assembly Line Balancing... 23
3.2 Solving a Combinatorial Optimization Problem with Simulated Annealing ... 24
3.3 Automation in Production ... 25
3.3.1 Definition and Measurement of Automation ... 25
3.3.2 Risk of Automation ... 26
3.3.3 Automation options ... 27
3.4 The Value of an Investment ... 28
3.5 Conclusion ... 29
4 Current problems and inefficiencies ... 31
4.1 Balancing Loss ... 31
4.2 Waste analysis ... 32
4.3 Defects ... 33
vii
4.3.1 Losses due to material shortage... 34
4.3.2 Losses due to assembly mistakes ... 34
4.4 Part Allocation at Workstations ... 35
4.5 Conclusion ... 36
5 Possible improvements ... 37
5.1 Assembly Line Balancing ... 37
5.1.1 Assembly Line Balancing Application ... 38
5.2 Sub-assemblies ... 43
5.2.1 Assembly by or Support of Cobot ... 44
5.2.2 Gathering Parts with Vibratory Fillers ... 44
5.2.3 Screwing unit ... 45
5.3 Bin picking ... 45
5.3.1 Gathering Parts with Vibratory Fillers ... 46
5.3.2 (Partially) Remove Kitting Area ... 46
5.3.3 Use Kitting Area for Other Purpose ... 53
5.3.4 Use Cobot for Kitting ... 53
5.4 Testing ... 54
5.4.1 Analyse Operations to Fill Time ... 54
5.4.2 Visual Checks and Placement of Test Heads by Cobot ... 54
5.5 Packing ... 54
5.5.1 Palletizing Robot ... 55
5.6 Kanban ... 56
5.7 Conclusion ... 58
6 Conclusions and Recommendations ... 61
6.1 Conclusions ... 61
6.2 Implementation & Recommendations ... 62
6.3 Discussion ... 63
Bibliography ... 65
Appendix 1 Spaghetti Diagrams ... 67
Appendix 2 Simulated Annealing Code ... 71
Appendix 3 Application Manual ... 75
Appendix 4 Assembly Line Balancing Application Solutions ... 82
Appendix 5 Overview of changes when parts are stored in the racks ... 89
viii
List of Abbreviations
Abbreviation Definition Introduced on page
BPS Bosch Production System 4
HCM Heat Control Modular 8
KIF Ketel Identificatie Formulier (Boiler Identification Form) 11
KPI Key Performance Indicator 4
LBE Line Balancing Efficiency 17
LOG Logistics 2
MFO Manufacturing Operations 2
MOE Manufacturing Operations and Engineering 2
MTM Method Time Measurement 14
NPV Net Present Value 28
OEE Overall Equipment Effectiveness 16
SA Simulated Annealing 24
SALBP Simple Assembly Line Balancing Problem 23
SSL Supplier Service Level 18
SX Smoothness Index 24
TEF Technical Functions 2
TPS Toyota Production System 21
List of Figures
Figure 1-1 Overview of the production locations of Bosch Thermotechnology GmbH ... 1
Figure 1-2 One of the TrendLine boilers: TrendLine HRC30 - CW5 ... 2
Figure 1-3 Problem Cluster ... 3
Figure 2-1 Layout of the TrendLine A Production Line (Source: TEF Department) ... 7
Figure 2-2 Flow Chart of TrendLine A production line ... 7
Figure 2-3 Frame with Supply Connection, Return Manifold and Pump ... 8
Figure 2-4 Frame with Control Box of Workstation 2 and Pipes of Workstation 3 ... 8
Figure 2-5 Sub-assembly of Heatcell assembled at Workstations 4 and 5 ... 9
Figure 2-6 Frame with heatcell ... 9
Figure 2-7 Boiler ready for Packing and Shipping ... 10
Figure 2-8 Aid to carry the boiler frame (Dolly) ... 11
Figure 2-9 Averages of clocked operating times in the current situation ... 14
Figure 2-10 MTM measurements corresponding to current assembly situation ... 15
Figure 2-11 Development of the Efficiency VT KPI during the past year ... 16
Figure 2-12 Development of the OEE KPI during the past year ... 17
Figure 2-13 Development of the Production Quality KPI during the past year ... 18
Figure 2-14 Development of the SSL KPI during the past year... 18
Figure 2-15 Development of the productivity during the past year ... 19
Figure 3-1 Example of a Spaghetti Diagram (Source: Bureautromp.nl) ... 22
Figure 3-2 Prob of computerization for wage and education level (Source: Frey & Osborne, 2013) ... 26
Figure 4-1 Identification of Value and Non-Value-Added operations ... 32
Figure 4-2 Classification of operations into value-added, support and waste activities ... 33
Figure 4-3 Causes of Boiler Losses ... 34
Figure 4-4 Space for Parts ... 35
Figure 4-5 Sequence of Picking at Workstation 4 ... 36
Figure 5-1 Print Screen of the Line Balancing Application ... 39
ix
Figure 5-2 Validation of the Line Balancing Application ... 40
Figure 5-3 Progress of Line Balancing Application for the Current Situation ... 41
Figure 5-4 Task Division Result of Line Balancing Application applied to Current Situation ... 42
Figure 5-5 Vibratory Filler with Dispense Cup ... 43
Figure 5-6 Example of a Rivet Dispenser ... 44
Figure 5-7 Improved lay-out at Workstation 4 with parts from kit in racking ... 49
Figure 5-8 Kit for sub-assembly Heatcell ... 53
Figure 5-9 FANUC CR-35iA ... 55
List of Tables Table 0-1 Summary of Line Balancing Application Results... iii
Table 0-2 Economic and Ergonomic Evaluation ... iii
Table 2-1 Division of workstations with 50% occupation ... 10
Table 2-2 MTM Analysis of a small part of Workstation 1 ... 14
Table 2-3 Performance Measurements relevant to the research ... 15
Table 3-1 Mechanization and Computerization Scale (Source: Frohm, 2008) ... 26
Table 4-1 Summary of SI and Line Balancing Efficiency for the current situation ... 32
Table 5-1 Economic Evaluation of Implementation of the Line Balancing Application ... 43
Table 5-2 Part Information of Sub-assemblies ... 44
Table 5-3 UAS-Norm Time Table (Source: TEF Department) ... 48
Table 5-4 Time and Cost calculations for Workstation 4 when Kitting is not done anymore ... 48
Table 5-5 Time and Cost calculations for New Lay-out ... 49
Table 5-6 Total time and cost savings of placing certain parts in rack ... 50
Table 5-7 SALBP solution (partial) removal of kitting area ... 51
Table 5-8 SALBP solution with future test of 188 seconds ... 51
Table 5-9 SALBP solution with future test and (partial) removal of kitting area ... 51
Table 5-10 Economic Evaluation of Partial Removal of Kitting Area with 5 PTL Systems ... 52
Table 5-11 Economic Evaluation with 3 PTL systems and 2 Active Assists ... 52
Table 5-12 Economic Evaluation with Cycle Time Reduction and 5 PTL Systems ... 52
Table 5-13 Economic Evaluation with Cycle Time Reduction, 3 PTL systems and 2 Active Assists... 53
Table 5-14 SALBP solution with robotic palletizer ... 55
Table 5-15 Economic Evaluation of Palletizing Robot ... 56
Table 5-16 Economic Evaluation of Combination of Investments ... 56
Table 5-17 Economic Evaluation of an Automated Kanban System ... 57
Table 5-18 Summary of Line Balancing Application Results... 58
Table 5-19 Summary of Pay-back period results ... 59
1
1 Introduction
This chapter introduces the research performed at Bosch Thermotechniek to complete my master’s degree in Industrial Engineering and Management. This research analyses the current performance of one of the production lines and seeks ways to improve performance. Section 1.1 provides a description of the company, situation and the production plant. Section 1.2 describes the research plan.
1.1 Company Background
Section 1.1.1 describes the relation of Bosch Thermotechniek Deventer in the context of the wider Robert Bosch Concern and the situation they are currently in. Section 1.1.2 describes the Deventer production plant and the context for this research.
1.1.1 Bosch Thermotechnology
Bosch Thermotechniek B.V. is one of the twenty production locations of Bosch Thermotechnology GmbH (Figure 1-1). Bosch Thermotechnology is part of the wider Robert Bosch Concern and is an industrial leader in heating and hot water products with an approximated turnover of €2.8 billion (Bosch Thermotechnology, 2018). Bosch Thermotechniek B.V. is located in Deventer, the Netherlands, and sells the from origin Dutch Nefit boilers as well as other brands such as Junkers and Buderus. Nefit is part of Bosch Thermotechnology since 2004 and its strong brand name is commonly known.
Deventer also has a Research & Development department that develops and researches hybrid sustainable solutions, as well as a sales and aftersales department. The location in Deventer employs 700 people (Nefit, 2017).
Figure 1-1 Overview of the production locations of Bosch Thermotechnology GmbH
As can be observed from the map, the production facility in the Netherlands must compete with production facilities in low-wage countries where the cost of labour covers a smaller part of the total costs. Productivity is therefore highly important and deserves ongoing attention.
In 2018 the media in the Netherlands headlined ‘Old-fashioned boilers disappear from 2021’. However, fifteen parties proposed not to forbid the use of boilers. Among these fifteen parties are energy suppliers, environmental organizations and the business association for the installation branch.
Instead they asked for higher efficiency requirements and the replacement of old boilers with high efficiency boilers. The latter boilers work in combination with hybrid systems such as heat pumps and
2 sun boilers (Uneto-Vni, 2018). The production of high efficiency boilers is thus guaranteed for another 10-15 years and improvements in the production of the boilers are still of great importance.
Formerly, Bosch Thermotechniek B.V. produced all boiler types on a single production line. As a consequence of economic depreciation and high maintenance costs the production line was replaced by separate production lines for every product family. The production lines for the product families are similar across the various production facilities worldwide. This way, production disruptions and thus losses in one production facility can be compensated for by other production facilities. This strategy is called a second-source strategy. Due to this strategy, there is no wish to go back to a single production line as this would hamper information sharing and similarities in production.
1.1.2 The Deventer Production Plant
The production facility in Deventer is filled with production lines that are dedicated to production families for the lifetime of the product. This production line is constructed on the work floor whenever a batch of a product family has to be made (this can occur multiple times over the lifetime of the product). At the moment of writing there are five production lines on the work floor for different product families. Every production line has multiple workstations and the workstation that needs the longest time to fulfil the operations determines the cycle time of all work stations. The supply of the parts for production is done by milk-run trolleys on the basis of a
two-bin Kanban system.
This research focuses on the TrendLine A production line for reasons that are to be explained in the section about the scope of this research. The TrendLine A production line produces four types of the TrendLine boiler. These types differ in their heat output and quality mark. The Dutch certification company KIWA established a quality mark ranging from CW1 to CW6. CW is an abbreviation for Comfort Warm Water and the level indicates the use for which the boiler is most suitable, with 6 being the most advanced. The Trendline series consist of a CW4, CW5 and two CW6 boilers with heat outputs of 25 or 30 kW (Figure 1-2). The various types do not differ in production time as the same operations are performed.
The only difference between the various types are the parts that go into the frames (e.g. different gaskets). The policy of Bosch is to make a different production line whenever the types do differ in required production times.
There are four departments that are closely linked to the actual production being Manufacturing Operations and Engineering (MOE), Technical Functions (TEF), Manufacturing Operations (MFO) and Logistics (LOG).
MOE is the head department for production and is responsible for the management of production. TEF has 16 FTE and is responsible for maintaining and improving the factory processes and equipment. The department consists of a manager, technicians, maintenance technicians, industrial and project engineers and technical specialists. This thesis is performed as part of the TEF department, but requires interaction with the other departments as well. The team leaders and planners of the production lines are part of MFO. They are responsible for the day-to-day management of the production lines and are located close to or in the production lines. The LOG department ensures the correct supply of parts to the production lines.
Figure 1-2 One of the TrendLine boilers:
TrendLine HRC30 - CW5
3
1.2 Research Plan
This section presents the research plan. First, Section 1.2.1 describes the problem as presented by Bosch. The subsequent sections present the problem cluster, the objective of the research, the scope and the research questions with an approach of how to solve them. Finally, section 1.2.5 presents the deliverables.
1.2.1 Problem Description
The existing production lines at Bosch Thermotechniek involve a lot of manual labour. The financial controller stated that the personnel costs comprise 30% of the value-added costs of the production and logistics department (total costs related to production and logistics – material costs). Therefore, there is a wish to reduce the expenses on personnel. In addition, the workforce of Bosch is aging. In 5 years, the average age of the production employees will be 53 years. It is expected that this may lead to ergonomic problems and thus problems such as absence and health problems may arise. Moreover, these employees will leave the company for retirement in a couple of years. Therefore, right now is the perfect moment to consider if production can be done with less people. In addition to this, Bosch has a hard time attracting new employees, because there are few practical people in the Netherlands that are willing to do the work. It is thus of utmost importance that the number of employees needed for production decreases and/or the current older employees will stay. To this end, the long-term employability has to improve. The loss of knowledge is not of importance as operations are easily learned within two to ten days. The problem cluster as described is presented in Figure 1-3.
Figure 1-3 Problem Cluster
In summary, the motivation for this research is founded on two aspects. 1) the prospect of attaining higher levels of productivity through improved processes and the implementation of machines as well as 2) the increasing cost of attracting new employees on a tight labour market.
1.2.2 Objective of the Research
The goal of the research is to provide Bosch Thermotechniek with a plan on how to transform their current production line into one that costs less in the long run. Restrictions to the solution are an ergonomic design and a payback period of three years for investments.
1.2.3 Research Scope
The scope of this research is limited to the TrendLine A production line as this line produces most products and has most data available. However, this does not exclude generalizability to other production lines if similar problems or losses occur.
4 Every production line is supplied by milk-run trolleys. If the analysis shows that problems occur due to insufficient supply, the milk-run trolley supply is analysed as well. However, if supply is not found to have impact on the productivity of the production line, the process of supply is left out of consideration. The scope is further limited to the actual production and testing of the boilers. The inbound and outbound logistics are not part of the analysis.
Moreover, as Bosch works with the Lean Management philosophy already, recommendations and solutions should fit this philosophy. Bosch is already familiar with the concepts of Jidoka, 5S etc. and improvements supported by this philosophy most likely gain easier support.
1.2.4 Research Questions
The description of the problem and the objectives lead to the following questions. The main question answered in this thesis is as follows:
RQ ‘How should the TrendLine A production line be redesigned to decrease costs?’
To answer this question, we need answers to the following sub questions:
SQ 1 What is the current state of production on the TrendLine A line?
a) What does the production line look like?
b) What operations are performed at the various workstations?
c) What KPIs are currently used and how is the performance?
This sub question is answered in Chapter 2. Mapping the current TrendLine A production line requires process descriptions of the operations. To visualize and validate these process descriptions observations are made along the production line and interviews are held with production employees and their team leaders. With this information I can give a full description of the production processes that are used to produce a TrendLine A boiler. The Trendline A production line consists of 12 workstations, this is the highest level for a flow chart. Within the workstations, multiple operations are performed by a single employee and these can be described on a more detailed level. Bosch Thermotechniek also ranks the difficulty of the workstations on a continuum. This sub question should also evidence the meaning and application of this ranking in and for production.
Key Performance Indicators (KPI) are often related to productivity or efficiency. In the latter case one wants to produce the same with less input, whereas productivity is concerned with increasing output from the same level of input (Fried, Knox Lovell, & Schmidt, 2008). Relevant KPIs to this research are related to costs, ergonomics, productivity and efficiency and are determined in consultation with the supervisor and other stakeholders.
SQ 2 What does literature say about Lean Management, Automation of Production Lines and the economic evaluation of alternatives?
This research seeks ways to improve productivity and efficiency of production lines. The Lean Management philosophy is known to be focused on continuous improvement and Bosch employees are already familiar with it. Bosch has its own ‘Bosch Production System’ (BPS) as variant on the ‘Toyota Production System’. Lean Management tools are widely used in the company and it is likely that proposals supported by Lean Management tools are accepted more easily. Therefore, the first section focuses on Lean Management, useful tools to identify waste and tools that can help resolve identified issues. Keywords that are used in Google Scholar and Scopus are among others: ‘Lean Management’,
‘Value-added, Business Value-added & Non-value-added waste’, ‘Identifying waste’ and ‘Reducing
5 waste’. The second section discusses automation in production. The research is conducted in a production location and I require an idea of what tasks can be automated and supported by automation and which options there are to do so. Bosch is especially interested in the use of machines and robots in its production lines. Keywords used for this section are among others: ‘automation in production’, ‘automation of tasks’, ‘automated assembly’ and ‘robotic applications’. As business cases are not always scientifically documented, the world wide web is searched in full for similar cases.
Finally, the research proposes alternatives for current processes that have to compared for their economic value and effect on the line balance. Keywords used are among others: ‘investment value’
and ‘line balancing efficiency’. The literature review can be found in Chapter 3.
SQ 3 Which problems and inefficiencies occur along the line and where do losses occur?
Analysis of the collected data can evidence operations or processes where losses occur and/or improvement is possible. One can think of wastes and balancing losses. It might also shed light on which workstations have the possibility to become more ergonomically designed. Additionally, employees on the work floor often have a better idea of problems that occur than management. That is why I would like to gain input from these employees as well. Data collection might be necessary where data is not gathered yet. Chapter 4 presents this analysis.
SQ 4 Which possibilities for improvements exist and what are the consequences in terms of operating times, ergonomics and costs?
Chapter 5 discusses possible changes to the production lines that can result in an improved productivity. This is done in response to the identified problems and inefficiencies in Chapter 4. The Manager Manufacturing Engineering and Technical Functions at Bosch expressed the wish for implementation of robots and automation along the line to increase the productivity rate. However, the solution should not be restricted to the deployment of robots and machines. Other sources of productivity gain and ergonomic gain such as changes in operating sequences, operating procedures and adjustments to machines are just as important. In terms of robots and machines one can think of:
1) supporting employees in their operations by robots or 2) the replacement of manual operations with machines where machines are either faster, cheaper or more precise. It might also be wise to reconsider the task assignment to the various workstations. Additionally, if balancing losses occur, I see an opportunity for the use of a line-balancing program that can help design and balance a line whenever significant changes occur. The chapter is structured according to the processes on which the changes have an impact.
Sources of solutions are the technical department, production employees, literature and other documented real-life examples. All Bosch Thermotechnology facilities share knowledge about the deployment of robots in their production lines. This knowledge as well as experience with the application can be used for the research.
All possibilities have to be evaluated for positive and negative consequences. One can think of time reductions in operations, time additions, personnel cost reductions and additions and ergonomic improvements or problems that will occur. Both costs and benefits have to be analysed for every possible improvement and the payback period and NPV have to be calculated. Those opportunities that have a payback period of less than 3 years can be considered for implementation. The finance department of Bosch will be asked for their approach towards investment decisions, after which I decide on the suitability of this approach. This chapter concludes with a list of changes that can be implemented to improve productivity.
6 1.2.5 Deliverables
The deliverables of this research are:
• A conclusion on ideas that should not be pursued in the current situation.
• An implementation plan as part of the conclusion on the improvements and automation steps that should be made to improve productivity and/or achieve cost reductions.
• A line balancing application that can be re-used for future production line creation and rebalancing of the existing production lines when significant changes occur.
7
2 Current State and Performance of the Production Line
This chapter describes the current production line and performance. Section 2.1 describes the current production line on a high-level and the functions of the workstations. Section 2.2 goes into depth by describing the operations on the various workstations in detail, Section 2.3 describes the supporting practices that enable production. Finally, Section 2.4 describes the current operating times and the current performance in terms of KPIs. The chapter closes with a conclusion in Section 2.5.
2.1 The TrendLine A Production Line
This section presents the production line under consideration on a high-level. It presents the lay-out, the various workstations with its functions and the flow.
The TrendLine A production line presented in Figure 2-1 is a so called Mixed-model U-line as it produces more than one product (Miltenburg, 2002). The tasks at the first 11 workstations are performed within the cycle time. The 12th workstation prepares part kits for the other stations in batches of 6 and has sufficient time to perform the operations to keep up with the takt time. A test on the boiler or any of its sub components is performed on Workstations 5, 8, 9, and 10.
Controle Workstation
Figure 2-2 Flow Chart of TrendLine A production line
2 1 4 3
5 6 7 8 9 10 11
12
1
Figure 2-1 Layout of the TrendLine A Production Line (Source: TEF Department)
8 Figure 2-2 presents the flow of production of a boiler on the TrendLine A production line. Every task represents a workstation where multiple operations are performed to fulfil the main task of the workstation (e.g. frame mounting). Every workstation is occupied with one employee who performs the operations in a pre-defined sequence described in the work sequence cards. The work sequence cards describe the actions to take, the necessary parts and equipment and the complementary work instructions.
The following section describes the operations on the various workstations in detail to evidence the function of the workstations.
2.2 Detailed description of the workstation operations
This section describes the operations performed at the workstations as described in the work sequence cards and observed from the videos of the various workstations.
Workstation 1 – Frame mounting
This workstation ensures that the floor plate is mounted onto the empty boiler frame. Insulation is placed all around the boiler frame and the supply connection and return manifold sub-assembly are assembled. The three-way valve and the flow regulator are installed after some checks for fabrication mistakes and installation of the flow sensor. The flow regulator is different for the various TrendLine types and can be distinguished by colour. Finally, the pump is mounted and secured and the supply connection and return manifold with pump are installed in the frame (Figure 2-3).
Workstation 2 – Electronics mounting
The employee at this station checks the work done at Station 1. Then the employee attaches the low-voltage cable, the modulating plug and connects the three-way plug and flow-, pressure- and heat sensor. Finally, the high-voltage cable, the 230V cable and covers are mounted. All cables are then rolled together and secured (Figure 2-4).
Workstation 3 – Pipe mounting
Again, the work of the previous workstation is checked.
Then the HCM (Heat Control Modular) is programmed and the gas pipe is mounted and secured. The low- voltage and high-voltage cable are put correctly in the frame and the earth plug and pump cable are also placed correctly. The supply and return pipe are mounted in the manifold subsequently and sensors are attached. The
gas valve bracket is installed and the HCM is taken out of the programming unit and installed.
Figure 2-3 Frame with Supply Connection, Return Manifold and Pump
Figure 2-4 Frame with Control Box of Workstation 2 and Pipes of Workstation 3
9 Workstation 4 – Sub-mounting heatcell 1
The employee at this workstation does not have to perform any checks on the work of other workstations. He or she takes the exchange and places it in the dolly. The ignition seal is placed in the exchange and the pipes are sealed with a mould. Then the ignition, the transformer bracket and the condensate collector are mounted to the exchange. Finally, the gas burner and the air vent are placed and the dolly is passed on to the next work station.
Workstation 5 – Sub-mounting heatcell 2
At this workstation no checks on previous work are performed. A fan is placed in the mould and the venturi adapter is placed on the fan with help of a torque wrench. A white cap is mounted on the measuring nipple of the venturi adapter. Then the burner hood connector is placed in the mould and the fan is mounted on it. Then a gasket is placed on the fan. The burner hood is prepared with a flap guard and a flap valve and attached to the fan. Subsequently, the burner seal is placed on the burner hood and the burner assembly is placed on the exchange. Everything is secured with clamps and crown nuts and finally the air vent and the clamping plate are mounted. If all operations are done, a test is started to check the working of the heatcell (Figure 2-5).
Workstation 6 – Heatcell mounting 1
At this workstation the gas valve and the heat exchange are placed into the boiler frame. While doing this, the work of workstation 4 and 5 are checked as well as the heatcell test results. After placement of the exchange, the flow and return pipes are placed and the pipes are secured with a clip.
Workstation 7 – Heatcell mounting 2
The employee at the workstation places the gasket on the pump and attaches the pump plug. Then, the exchange is secured to the backside and the burner protection is installed.
The cables are then mounted on the burner protection and ionisation pens. The earth plug is connected to the earth ignition, the red ignition cable is connected to the ignition of the WW and the ionisation plug to the ionisation. The spark transformer is installed in the transformer clamp. Then, two red ignition cables are connected to the spark transformer and two cables are mounted on the ventilator. Finally, the residual gas analyser sensor is installed on the s-pipe and located in the frame (Figure 2-6).
Workstation 8 – Hydro High-Voltage Test
After checking the correct assembly of the former workstation, the 230v-cable is put in the wall outlet and the test heads for the leakage test are attached to the boiler. Then the hydro test and gas test are started. The hydro test is performed to test the pressure vessels for strength and leaks. Water is sent through the pipe system to detect problems. Then a test pen is used to perform the electrical safety test through touching 5 points on the boiler. Subsequently, the employee places insulation over the
Figure 2-5 Sub-assembly of Heatcell assembled at Workstations 4 and 5
Figure 2-6 Frame with heatcell
10 exchange, and installs the siphon on the condensation tray. The HCM is fitted into the control box and the venture is sealed and cased. Then the gas hose and the suction pipe are attached. The power cable plug and the air vent cap of the pump are removed and the hydro test and gas test are finished.
Workstation 9 – Final test
At this workstation the boiler is tested. The 230V cable and the flue are coupled to the test machine and the test heads are attached to the boiler. Then the test is started and this test takes 180,5 seconds.
While the test runs, the employee checks whether the bolts and tulles are mounted correctly and if there is a water or gas leakage. If the test is successful, the machine is detached and passed on to the next workstation. The employee is often waiting for the test to finish as his operations in the meantime have run out.
Workstation 10 – Cover mounting
Here, the cover receives the Nefit logo and the inside is covered with foam. The cover also receives an inner insulation seal to make the boiler air-tight. The boiler from Workstation 9 receives a few caps on the connections and the control box, exchange and venturi adapter are checked for missing parts.
Finally, the front case is fitted to the front of the boiler and the CO2 Capture and Storage (CCS) test cycle is started.
Workstation 11 – Packing
When the boiler arrives, the boiler is completely finished (Figure 2-7). The boiler is packed in Styrofoam for protection, boxed and accessories and instruction manuals are added. The box is then sealed with strapex and placed on a pallet with a suction machine. This machine carries the full load of the boiler and is steered by the employee. If 6 boxes are placed on the pallet, the boxes are sealed to the pallet. The employee brings the freed dolly back to workstation 1 and hangs a new frame on the dolly.
Workstation 12 – Picking
At this workstation, the part kits for workstation 6 are prepared. An employee gets 8 crates and fills them manually with the various screws, bolts and other parts that are required. Another 6 kits are prepared for the production line following the picking lists including the type specific parts. This employee also ensures that empty crates are removed from the line and other boxes are opened. In case the covers at workstation 10 have run out, the employee also brings new covers to this workstation.
In case of 50% occupation the division of workstations among the employees is as presented in Table 2-1. This division ensures that checks are not performed by the same employee.
WS1 WS2 WS3 WS4 WS5 WS6 WS7 WS8 WS9 WS10 WS11 WS12 Employee 1
Employee 2 Employee 3 Employee 4 Employee 5 Employee 6
Table 2-1 Division of workstations with 50% occupation
Figure 2-7 Boiler ready for Packing and Shipping
11
2.3 Supporting practices
This section describes some of the supporting processes that enable production. It includes the transportation of the boiler frames from one workstation to another, employment of new production employees, ergonomic considerations, the tools and part delivery at the workstations and the procedure that is used for line balancing.
All boilers are sent into the production line on a so-called dolly with a corresponding KIF (Ketel identificatie formulier) that presents the type of the boiler and a boiler specific barcode. A dolly is an aid that can carry a boiler frame (Figure 2-8). The KIF is placed on the dolly next to the boiler. On the KIF every workstation has space for the employee to sign off on the workstation.
Every new employee is trained on a workstation for two to ten days depending on the difficulty of a workstation. After this training, a certified employee takes the final test. Passing the test allows an employee to start working on the workstation alone. As a consequence, not every employee can work on all workstations.
Over time, employees get to learn multiple workstations to be able to rotate between workstations. Rotation is done once a day.
Moreover, every workstation has its own score of difficulty in which ergonomic considerations are taken into account. According to
rules from the Arbo Unie in the Netherlands, points are given on a scale of 1 to 5 for various movements such as walking, carrying etc. A score of 1 for walking would mean that there is few walking done, 1 is thus the best score and 5 is the worst. A lower total score means a better workstation in terms of ergonomics. Then, the number of work sequence cards and lines of instructions for a workstation are also considered. This way, both physical and mental difficulty are considered. If employees have an injury that restrains them in their abilities, they can be allocated to a workstation that does not worsen the injury.
Ergonomic data cards are used across all Bosch divisions. Employees that need to ensure good ergonomic conditions make use of these cards. They prescribe instructions for the design of the workspace. The contents of the cards are as follows:
• Blue Card: workplace measurements
• Yellow Card: manual transfer of load in standing position
• Yellow-Brown Card: handling of loads at storage racks and at material supply systems
• White Card: vision and illumination
Ergo checks check adherence to these cards and should be performed at the production lines every 2 years and whenever a change is made.
At every workstation there are:
• Racks that carry the parts required for the operations,
• Railings with tools such as drilling machines and torque wrenches,
• Machines and jigs required for testing and carrying of the boilers or subsystems.
All workstations are supplied by a two-bin Kanban system. Whenever a bin or crate is empty the employee removes the bin or crate from the immediate part location and places the Kanban card in
Figure 2-8 Aid to carry the boiler frame (Dolly)
12 the Kanban card bin. Milk-run trolleys drive around and pick-up these cards and scan the barcode to signal logistics that a refill is needed.
All parts are stored in one of 5 reusable packaging units.
Grey Bin S (300x200x120) Grey Bin M (400x300x150) Folding Crate Grey/Blue (600x400x220)
ESD Box (600x400x220) DIN Pallet (1200x1000x122)
At the moment, the line is rebalanced not as often as it should. Changes in the production procedures cause changes in times required for the operations, but the line is not always rebalanced when a change occurs. When the technical department works on the design of a new production line, the starting point for the line balancing problem is an overview of all operations that have to be performed with real-time measurements of the time requirements. Then, intuitive decisions are made on the sequence of the tasks that have to be performed. Whenever the technical department decides to rebalance a line, they often have an overview of the total time required at every workstation. Work content is then intuitively replaced from a workstation that has an overload, to a workstation that is underloaded without violating precedence constraints. The question is whether it is possible to make this process less intuitive and more exact.
2.4 Current Performance
This section presents the current performance of the TrendLine A Production Line.
A description of the current performance of production requires historical information. On a daily basis, Bosch collects information about the various production lines which consists of, but is not limited to:
• Clocked hours
• The number of employees on the line
• The number of products that were lost due to
o Organisational interruptions (e.g. meetings of production employees) o Technological problems
o Quality problems o Material problems
13
• The hours of direct support
• The planned number of boilers for a day and the actual produced number of boilers (Supplier Service Level (SSL))
• The number of boilers that are accepted after the final test (failure rate)
• The efficiency rate based on the planned hours and used hours
• The OEE rate
In the past year (2 October 2017 – 28 September 2018), the TrendLine A production line produced boilers on a total of 228 days. The line was always employed with either 6 or 12 employees with one instance of 9 employees. Operating hours varied between 4.7 and 8.5 hours. The production plan is made based on a yearly forecast. Every year an estimate is made taking into account sales figures of the past year and trends. This way, the production levels are kept relatively stable over the year and seasonal peaks are prevented as much as possible. Normally, employees work one shift from 07:00 to 16:00, in times of increased production a second shift is added from 16:30 to 00:30. Small deviations such as working to 16:30 for the first shift can occur.
Section 2.4.1 explains the use of a takt time and presents the current state of the division of the operations over the workstations. Section 2.4.2 presents performance measurements that are currently used.
2.4.1 Takt Time
Production systems that work according to a takt time, normally establish this takt time as follows (Theisens, 2016):
𝑇𝑎𝑘𝑡 𝑡𝑖𝑚𝑒 =𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 𝑡𝑖𝑚𝑒 𝑓𝑜𝑟 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝐶𝑢𝑠𝑡𝑜𝑚𝑒𝑟 𝐷𝑒𝑚𝑎𝑛𝑑
The customer takt established for the TrendLine A production line was 250 seconds based on a situation with 11 workstations. However, as the technical department is aware of production losses that occur, they have decided to work with a cycle time of 238 seconds to hedge against these losses upfront.
To determine the state of the current line balance the cycle time of each workstation has been measured 20 times by me and two other employees for the purpose of this research. As operators differ a lot in age and experience, the time they need on workstations differs substantially. To take into account this inter-operator variability, multiple operators were clocked. The result of these measurements can be found in Figure 2-9. This analysis evidences that the workload is not evenly distributed and that the cycle time of 238 seconds might be too long. The current bottleneck is workstation 6 as it contains the longest work content according to the measurements. There is no specific task that acts as bottleneck. The longest task is the final test at workstation 9 and lasts 180.5 seconds.
