Improving efficiency and responsiveness of the assembly process for solar systems

(1)

Master Thesis – Andre Vollmer ₁

Master Thesis Technology Management

Improving efficiency and

responsiveness of the assembly

process for solar systems

Author : Andre Vollmer

Study : Technology Management (M.Sc.) Student number : s1946587

Email : s1946587@student.rug.nl

Faculty : Faculty of Economics and Business University Supervisor : Dr. ir. D.J. van der Zee

University Co-assessor : Dr. X. Zhu

Company : XXX Company Supervisor : XXX Company Co-assessor : XXX

(2)

Master Thesis – Andre Vollmer ₂

Preface

This report represents my final project for the Technology Management (M.Sc.) masters program at the University of Groningen, Holland. Thereby I had the great opportunity to experience the photovoltaic business as a member of the product management team of the company XXX in XXX, Germany. During a period of 10 months I was involved in conducting a Value Stream Analysis of the first available standardized solar system in the German market.

This research project has been conducted with the help of many people; I would like to take this opportunity to show my appreciation to those who helped me to finish this research successfully. First of all I would like to thank XXX for giving me this opportunity to perform this research and his trust in my required skills. Next to that, I would like to thank XXX for his participative leadership and XXX for his help with several other conducted projects in the company.

Moreover, at the university I would like to show my appreciation to my supervisor Durk Jouke van der Zee for all his comments, his guidance and his constructive criticism. Similar appreciation goes to my co-assessor Stuart X. Zhu for helping me to bring this project to a successful ending.

Last but not least I want to thank my family and friends for all their support during my final research project and study.

(3)

Master Thesis – Andre Vollmer ₃

Management Summary

The motivation of this research project is on the one hand to shed light into the perceived poor efficiency of the assembly process of XXX standardized solar systems when compared to their customized solar systems. Labor costs are critically high and threatening the required net profit margin. On the other hand, customer service is at stake by long response-times in the assembly area. A highly volatile customer demand hampers to match the amount of available company supply with current customer demand.

The major dilemma is that more workforces in the assembly area could cope with upcoming demand peaks, but cost efficiency would even be worse then, especially in times of lower demand. Fewer workforces in the assembly would be more cost effective but for sure could not cope with the upcoming demand peaks and the inability to serve customers when demand reaches a peak is a worst-case scenario.

Therefore the objective is to assess the possibilities for redesigning the assembly system in the company for the standardized solar system in order to improve its efficiency and to ensure customer service.

The approach to find a suitable solution suggests involving two main elements in this project: An Action Research (AR) methodology complemented with different Value Stream Mapping (VSM) methods. Thereby the types of detected and removed waste according to Hines et. al. (1997) are depicted and finally their impact on the performance measures of assembly response-time and labor costs are elaborated.

(4)

Master Thesis – Andre Vollmer ₄ Preface Management Summary List of Abbreviations 1. Introduction ... 6 1.1. Company ... 6 1.2. Product ... 7 1.3. Operations ... 9 2. Research design ... 10 2.1. Problem context ... 10 2.2. Problem statement ... 14 2.3. Research question ... 15 2.4. System boundaries ... 17 2.5. Methodology ... 18 3. Improvement approach ... 19

3.1. Main elements of the approach ... 19

3.2. VSM methods ... 19

3.3. Setup of the VSM application ... 21

3.4. Organization of the project ... 27

3.5. Applied tools within each stage ... 29

4. Stage 1: Assembly process... 30

4.1. Assessment of the current state situation ... 31

4.2. Finding and creating a future state solution ... 35

4.3. Evaluation of the potential ... 39

5. Stage 2: Modules supplies ... 41

6. Stage 3: Inverter supplies ... 46

7. Stage 4: Mounting system supplies ... 58

8. Stage 5: Cables & plugs supplies ... 66

9. Summarized potential of all stages ... 74

9.1. The seven wastes ... 75

9.2. Performance Measures ... 76

9.3. Use and utility of the improvement approach... 80

10. Conclusions ... 81

10.1. Discussion & further research ... 81 References

(5)

Master Thesis – Andre Vollmer ₅

Abbreviations:

AR Action research

BOM Bill of material DC Distribution center

DP Customer order decoupling point FIT Feed-in tariff

FTE Full time equivalent

ICT Information and communication technology KWh Kilowatt per hour

PV Photovoltaic

(6)

Master Thesis – Andre Vollmer ₆

1. Introduction

This chapter will firstly introduce the company in which the project is conducted and secondly will describe the product under consideration. It, thirdly, will close the chapter by providing an overview of the assembly operations in the main distribution center.

1.1. Company

The XXX AG was founded in 2007 as a merger of the companies XXX, XXX and XXX. Each of those companies could look back to a long term of experience in the photovoltaic (PV) market and were among the pioneers of the PV industry.

The company sells PV modules and solar systems via the technical wholesale trade to roofers, electricians and heating system engineers. These tradesmen install that smaller plug- and play systems on the roof of their customers’ homes. XXX, with offices in XXX, XXX and XXX in Germany also supplies custom photovoltaic solutions with crystalline or thin-film modules and special off-grid (stand-alone) PV systems.

The enterprise is also responsible to direct the two distribution centers (DC’s) in XXX and XXX. Those two DC’s operate mainly as warehouses and assembly centers for all products sold to the customers. In those DC’s, all goods are stored and as soon as customer orders are flowing in, the assembly and the shipment get initiated. The figure below describes the strategy formulated by the executive board since the merger in 2007 and highlights the three main pillars:

Figure 1: Corporate Strategy

(7)

Master Thesis – Andre Vollmer ₇

The pillar of cost and quality leader is realized by owning a PV module production in XXX that is one of the largest production facilities in Germany. Economies of scale are constantly reducing unit costs as the size of the production facility is steadily growing. Nevertheless, the operations of the XXX AG can be defined as to be an assembler of solar systems, which is explained in more detail after the next paragraph.

The pillar of selling PV complete systems is realized by owning a vertically integrated supply chain to a very large extend. When complete systems are sold, higher prices for customers are justified by additional benefits like the system design and reduced complexity.

The pillar of focusing on the premium segment of residential homes is justified by a high demand of on-roof solutions. This demand in the premium segment in residential homes is triggered by a complex feed-in tariff (FIT) system of the German government. Those subsidies are steering the market demand away from free-standing solar generators on the fields of farmers towards on-roof solutions installed at residential homes.

1.2. Product

The product under consideration is named XXX and is a solar system that feeds electricity into the power grid. A XXX is available in two power classes; each can be build- up in two versions like shown in the figures below.

The power class 3.x with its 16 PV modules can be installed with the module configuration 2x8 and 4x4:

(8)

Master Thesis – Andre Vollmer ₈

The power class 5.x with its 24 PV modules can be installed with the module configuration 3x8 and 2x12:

A solar system includes everything that is necessary to produce electricity out of the sun irradiation and facilitates to feed this electricity into the power grid. Therefore a certain amount of involved product families are responsible for different tasks. Each solar system consists out of a certain amount of photovoltaic modules that are responsible for generating electricity. Also a mounting system is necessary which facilitates the installment of those PV modules on the roof of the customers. Next to that, the inverter is responsible for transforming the produced direct current into alternating current and feeding this electricity into the public power grid. A certain amount of cables and plugs are also necessary in order to connect the PV modules on the roof with the power inverter located in the basement of the house. Such solar systems get mainly installed on the roofs of private homes and are planed on a per project basis with the aim of installing the maximum amount of power capacity in consideration of local conditions like e.g. roof sizes and chimneys at the installation site. Such tailor made systems represent investments of house owners to generate profits by the feed-in tariff (FIT) which is paid by the German government for each kilowatt per hour (KWh) that is fed into the grid.

Nevertheless, a XXX is a product innovation by the company XXX AG that represents now a standardized solar system. A sophisticated analysis of roof types was facilitating to define requirements the system has to fulfill, in order to fit on about 80% of all German roofs. Customers are now able to buy a solar system for which the planning on a per project basis is redundant. This results in lower costs for end customers and justifies higher margins for the provider of such solar systems. The bill of material (BOM) of each product is standardized

(9)

Master Thesis – Andre Vollmer ₉

which means that each system contains exactly the same amount of parts and components, regardless the specific requirements of the installation site.

1.3. Operations

All customized solar systems and standardized XXX get assembled in the main distribution center in XXX and shipped to either a wholesaler or an installer. The main DC in XXX combines inbound logistics, warehousing and the assembly activities. The Inbound Logistics department is responsible to check and procure all parts and components needed for a solar system or a XXX to be on stock. This basically means to organize and control the deliveries of a large amount of parts- and components suppliers that are necessary to assemble a solar system.

Figure 4: Operations in the main DC

A team of two blue-collar workers, who receive a document that is containing the exact BOM of the solar system to be assembled, executes each assembly. This document depicts a barcode the workers have to scan and a wireless scanner guides the team throughout the warehouse. In total, each complete system contains about fifteen different parts or components to be collected out of the shelves while each part or component has to be counted to their exact required amount. Finally everything has to be put within a pre-defined order into especially foreseen boxes and being wrapped. In total such an assembly process requires about two hours depending on the capacity in KWp of the solar system.

(10)

Master Thesis – Andre Vollmer ₁₀

2. Research design

This chapter will set up the research design by firstly describing the problem context and secondly by stressing the problem statement of the project. It will proceed, thirdly, by elaborating the main research question, which will be answered stepwise by employing finer grained sub-questions. Fourthly, the system boundaries of the research will be set and fifthly, the project methodology will close this chapter.

2.1. Problem context

When the company XXX introduced their innovation XXX, the aim was to provide the first standardized solar system in the German PV market as already roughly described in the previous chapter. Now, this chapter is concerning with the two main areas of improving efficiency and ensuring customer service.

I. Efficiency: Labor costs in the assembly area are critically high and are threatening the required net profit margin.

II. Customer Service: Customer performance is at stake. Deliveries have to be within a certain area of tolerance.

Efficiency

(11)

Master Thesis – Andre Vollmer ₁₁

Figure 5: Required time per order in the assembly

Already at the end of 2009 the target of the average amount of FTE per order had to be increased from 100 up to 150 minutes. Nevertheless, at the end of the year 2010, also this aim was impossible to reach. The end of the year average in 2010 unfolded the exhaustive amount of 192,4 minutes per order. The increased time in turn was finally resulting in dramatically increased costs caused by the perceived poor efficiency of the assembly process. If orders take more than those mentioned 150 minutes, the net profit margin (20%, set by the controlling department) are at stake because additional costs appear caused by labor costs. Such exhaustive costs are threatening the competitiveness of a company and are the mayor problem to be addressed in this research.

0 50 100 150 200 250 300 2009 2010 Dez

05 Jan 06 Feb 06 Mrz 06 Apr 06 Mai 06 Jun 06 Jul 06 Aug 06 Sep 06 Okt 06 Nov 06

(12)

Master Thesis – Andre Vollmer ₁₂

Nevertheless, the introduction of the product XXX should relief this problematic cost inclination and standardization is seen as to be the remedy for that. In the table below, a comparison of required process steps between a customized solar system and a standardized XXX shows already first achievements in cost/labor reduction enabled by standardization.

Required processes _{(Standardized)}XXX vs. _(Customized)Solar system

1. Inbound Logistics (Administration) total: 31h per week 31h per week

-Modules 14,5h per week 14,5h per week

-Inverters 10h per week 10h per week

-Mounting System 3,5h per week 3,5h per week -Cables & Plugs 3h per week 3h per week

2. Inventory (Inspection + storing) total: 22h per week 22h per week

-Modules 16h per week 16h per week

-Inverters 4h per week 4h per week

-Mounting System 1,5h per week 1,5h per week -Cables & Plugs 0,5h per week 0,5h per week

3. Tender preparation: 10min. per # 40min. per #

4. Service Management in total: - 60min. per # -Development of the planning sheet - 15min. per # -Development of mounting system sheet - 15min. per # -Development of wiring plan - 20min. per #

-Development of BOM - 10min. per #

5. Inbound Sales Support in total: - 60min. per #

-Integration of BOM in SAP - 20min. per #

-Inverter rating - 20min. per #

-Selection of PV-modules - 20min. per #

6. Assembly process: 105min. per # 105min. per #

Total response time (3-6) 115min. per # vs. 265min. per #

Table 1: XXX vs. solar systems

(13)

Master Thesis – Andre Vollmer ₁₃

hours per week (h per week) while the rest of the resources in the table represent direct costs appearing each time when an order of a solar system is processed in the main DC.

Customer service

Standardization should also address the issue of volatile customer demand, which is characterizing the hyper competitive PV industry. The following chart depicts customer demand by showing the newly installed PV power in megawatt (MW) in the year 2009 and 2010 in Germany and company supply by showing the available FTE in hours (h) in the warehouse where the assembly is executed.

Table 2: Company supply vs. customer demand in the PV industry

Demand peaks are always appearing at the end of the year due to the annual reduction of the feed-in tariff (FIT) for new owners of PV systems. In 2010 the German government even had to decrease the FIT unscheduled in the mid of the year, caused by the surge of newly installed PV power.

Nevertheless, customer service is jeopardized each time when under- or over utilization of workers appears. When demand is too low workers suffer from waiting and boredom, which in turn might influence product quality. Besides that, the company remains to pay wages while generating almost no turnover. Contrary to that, demand reaches a peak and workers have to suffer from high pressure caused by too many orders that cannot be fulfilled anymore. Also that can influence product quality and for sure customer service is jeopardized by long assembly lead-times of the product. Not only those costs in wages of

(14)

Master Thesis – Andre Vollmer ₁₄

employees in the administration and assembly setting appear, also costs determined by customers that cannot be served evolve.

Therefore customer service has to be within a certain area of tolerance, in particular in times where customer demand is dramatically high. Especially in those times it is crucial to respond to as much orders as possible to assure customer service.

2.2. Problem statement

De Leeuw (2002) states that a problem statement consists of a research objective and a research question. On the one hand the research objective is the objective of the case study, on the other hand the research question describes the knowledge that is needed in order to achieve this objective. Based on the previous paragraph, in this paragraph the research objective and the research question are now elaborated. This research objective is dealing with the two main areas of concern and is formulated as follows:

Assessing the possibilities for redesigning the assembly system for a standardized solar system in order to improve its efficiency and to ensure customer service.

In order to finally assess for the benefits of those two areas of concern, performance measures are needed. Those performance measures reflect two vantage points:

Company point of view: Labor costs

(15)

Master Thesis – Andre Vollmer ₁₅

Customer point of view: Response-time

The response time addresses the customer service, which is dramatically at stake as explained in the problem context. Measurement starts when an order is flowing into the company and ends when the product assembly is completely finished.

2.3. Research question

The required knowledge to finally reach the above elaborated research objective is addressed in the main research question, which is formulated as follows:

How can the value creation process of a standardized solar system be improved?

The finer grained sub-questions, which are used to structure the following stages of this project and finally answer the main research question, therefore are:

SQ1: Which improvement approach suits the research objective?

This sub-question complements the project methodology with VSM methods and is thereby setting-up the research approach into sequential stages. Furthermore the research project is organized and the applied tools are outlined in further detail.

SQ2: What can be improved in the assembly process?

This sub-question begins at the customer end of the value stream and provides insights into the main value creation operation, which is the assembly process of the standardized solar system. Finally it requests to assess the potential to contribute to the problem statement.

SQ3: How can the product family XXX contribute?

(16)

Master Thesis – Andre Vollmer ₁₆

SQ4: How can the product family XXX contribute?

This sub-question is proceeding with the second upstream value creation process, which in turn are the inverters of the standardized solar system. Finally it requests to assess the potential to contribute to the problem statement.

SQ5: How can the product family XXX contribute?

This sub-question is proceeding with the third upstream value creation process, which in turn is the mounting system of the standardized solar system. Finally it requests to assess the potential to contribute to the problem statement.

SQ6: How can the involved cables & plugs contribute?

This sub-question is proceeding with the fourth upstream value creation processes, which in turn are all the necessary cables and plugs of the standardized solar system. Finally it requests to assess the potential to contribute to the problem statement.

SQ7: What is the overall potential of all improvements to the assembly system and the evaluation of the improvement approach?

This sub-question summarizes the improvements of all involved product families that are flowing into the assembly process and stresses the two main performance measures. Moreover it creates space for the researcher to evaluate the chosen improvement approach.

(17)

Master Thesis – Andre Vollmer ₁₇

2.4. System boundaries

In the first chapter, the company, their assembly operations and the product under consideration were introduced. Now, the system within the company in which the research is conducted is clearly outlined and in particular the boundaries of the research are stressed. The figure below shows those system boundaries and moreover depicts the relevant departments involved in the project:

The system under consideration contains the Inbound Logistics departments of each involved product family, their inventory in the main DC and the final assembly process. Complementary to that, the Distribution department is involved which receives the inflow of customer orders. The Warehouse Logistics department further processes those orders via SAP before the assembly starts. Not parts of the system are the value creation processes and logistical processes of all the parts and component suppliers due to time constraints. Also not a part of this project is the logistical process after the final assembly of the product, also caused by time constraints of this project. Due to the standardized BOM of

(18)

Master Thesis – Andre Vollmer ₁₈

the solar system, the amount of parts can exactly be derived from the amount of forecasted solar systems. For example when customized solar systems are forecasted, the positions of the BOM are not clear until the system was finally sold and finally planned to the particular customer requirements. This standardization reduces complexity for the Inbound Logistics department dramatically and therefore the forecast technique of the Distribution department itself is also not addressed in this research.

2.5. Methodology

An action research (AR) methodology (Checkland, 1991) was used, in which the researcher was involved in the improvement process as a participant-observer. This is an interactive inquiry process that balances problem solving actions implemented in a collaborative context with data-driven collaborative analysis or research to understand underlying causes enabling future predictions about personal and organizational change (Reason & Bradbury, 2002).

Figure 7: Action Research model

(19)

Master Thesis – Andre Vollmer ₁₉

3. Improvement approach

This chapter will firstly stress the main elements of the approach, which will be the Action Research (AR) methodology and the Value Stream Mapping (VSM) methods and, secondly, will justify the contingency approach of the VSM application in each of the following chapters. Thirdly, the setup of the VSM application will be depicted in a table and the match of the AR steps with the VSM methods will be explained, including the detailed actions of the project. Then, fourthly, the organization of the project will give ample insights about the required project teams per project stage and fifthly, an explanation of the applied tools in each project stage will close this chapter.

3.1. Main elements of the approach

The main elements of the improvement approach are, firstly, the Action Research (AR) methodology according to Brannick et. al. (2001) which sets out the main steps of the project e.g. diagnosing, planning action and evaluation. The second element, which are Value Stream Mapping (VSM) methods according to Hines et. al. (1997) extended with the visual methods of ‘Current and Future State’ mapping presented by Rother and Shook (1998) and Jones and Womack (2002), further detail each stage/cycle. Those elements focus on ensuring customer service and improving efficiency of the assembly process as described above in the problem context by removing the seven commonly accepted wastes of the Toyota Production System (TPS) according to Hines et. al. (1997).

3.2. VSM methods

According to Hines et. al. (1997) the difference between the traditional supply or value chain and the value stream is that the former includes the complete activities of all the companies involved, whereas the latter refers only to the specific parts of the firms that actually add value to the specific product or service under consideration. As such the value stream is a far more focused and contingent view of the value- adding process.

(20)

Master Thesis – Andre Vollmer ₂₀

(21)

Master Thesis – Andre Vollmer ₂₁

3.3. Setup of the VSM application

The following table provides an overview of the content of the following chapters in which VSM will be applied to the solar system under consideration. Thereby the distinctive stages of VSM are stressed including the detailed actions within each stage, complementary the linkage to the AR methodology is depicted.

3.3.1. AR Steps

The match of the diagnosing step of AR and the current state VSM is given when initially a problem is identified and data is collected for a more detailed diagnosis. The match of the planning action step of AR and the future state VSM is given by a collective postulation of several possible solutions, from which a Table 3: VSM execution guideline

AR Steps Diagnosing Planning Action (Taking Action) Evaluation VSM Stages Assessment of the current state situation

Finding and creation of solutions

Estimating the potential of solutions

Detailed actions

o Selecting the product family o Identify material and

information flows o Learning to see value o Taking a structured walk

along the value stream o Current state VSM

o Drawing the lean future state map

o Creation of solution(s)

(22)

Master Thesis – Andre Vollmer ₂₂

single plan of action emerges and is implemented. The AR step of taking action is not part of this project caused by a lack of time resources of the project. Finally, AR adds additional value to the VSM methods by extending them with a final evaluation phase where data on the results of the intervention are collected and analyzed and the findings are interpreted in light of how successful the action has been. This is necessary to present the estimated managerial impact of the proposed actions to the executive board of the company in order to trigger the implementation of the changes, which could follow after this project. At this point, the problem can be re-assessed and the whole process begins another cycle. This process is iterative and focuses finally on ensuring customer service and improving efficiency of the assembly system. If after a certain amount of time e.g. some components of the product or employees in the assembly are changing, the following cycles of improvement are capable of considering such evolutions of the framework of the problem and customer service and efficiency in the assembly are also assured in the long run.

3.3.2. Detailed actions

(23)

Master Thesis – Andre Vollmer ₂₃

The figure below depicts how the overall material flow takes place and how the VSM stages of current and future state mapping and their evaluation relate to those:

The assembly process (stage 1) merges all the supplies of the product families together and therefore a separate stage is devoted to shed light into the detailed value creation processes here. In this manner it is facilitated to acquire particular knowledge about each part of the jigsaw stepwise by applying the VSM tools as shown below:

VSM always begins at the shipping end and is working upstream in order to examine firstly those processes that are linked as close as possible to the customer. Therefore the first stage is concerning with the assembly process, followed by the steps of the individual product families. When the current state

Stage 5: Cables & Plugs supplies Stage 4: XXX supplies

Stage 3: XXX supplies Stage 2: XXX supplies Stage 1: Assembly process

Current State Map Future State Map

Evaluation

Figure 9: Stages and main tools of the project

Warehousing (5) Cables & Plugs

Warehousing (4) Mounting System Warehousing (3) Inverters Warehousing (2) Modules Evaluation Current State

Maps Future State Maps

(1) Assembly Customer

(24)

Master Thesis – Andre Vollmer ₂₄

situation is assessed and the detected wastes are outlined, the future state situation is drawn up including suggestions of the redesign of the system by removing waste. The evaluation step includes a clear outlined explanation of the removed waste and the impact on the elaborated performance measures of assembly lead-time and labor costs in the assembly.

Assessment of the current state situation

Each stage maps the material and information flow by firstly drawing up a current state value stream map, which is created by using a predefined set of standardized icons (Appendix). While doing so, different process owners at the shop floor are able to contribute to the diagnosis of the different material and information flows. Nevertheless, taking the value stream viewpoint means to work on the big picture and not on individual processes. Below an example is given by mapping the value creation of a letter:

Figure 10: Example current state VSM

(25)

Master Thesis – Andre Vollmer ₂₅

customer wants to pay for. This current state map in turn provides the basis for reducing the seven commonly accepted wastes according to Hines et. al. (1997) as explained below. He describes the seven commonly accepted wastes from a production environment, specifically from the automotive industry, and from a Japanese perspective. As a result some translation of the general terminology was required to adapt it to the particular assembly setting in a non-Japanese setting as follows:

Waiting

When time is being used ineffectively, then the waste of waiting occurs. In an assembly setting, this waste occurs whenever goods are not moving or being worked on. This waste affects both goods and workers, each spending time for waiting. The ideal state should be no waiting time with a consequent faster flow of goods.

Inappropriate Processing

Inappropriate processing occurs in situations where overly complex solutions are found to simple procedures such as using a large amount of resources to process simply small parts. Such over-complexity generally assigns a large amount of labor costs to a relative small fraction of material costs while no particular value is created for which the customer is willing to pay for. The ideal therefore is to invest the smallest amount of resources, capable of producing the required quality, located next to preceding and subsequent operations.

Unnecessary Inventory

(26)

Master Thesis – Andre Vollmer ₂₆

Transport

The fourth waste, transport, involves goods being moved about. Taken to an extreme, any movement in the factory could be viewed as waste and so transport minimization rather than total removal is usually sought. In addition, double handling and excessive movements are likely to cause damage and deterioration with the distance of communication between processes proportional to the time it takes to feed back reports of poor quality and to take corrective action.

Overproduction

Overproduction tends to lead to excessive lead and storage times. As a result defects may not be detected early, products may deteriorate and artificial pressures on work rate may be generated. In addition, overproduction leads to excessive work-in-progress stocks that result in the physical dislocation of operations with consequent poorer communication. This state of affairs is often encouraged by bonus systems that encourage the push of unwanted goods. The pull or kanban system was employed by Toyota as a way of overcoming this problem.

Unnecessary Motion

Unnecessary movements involve the ergonomics of production where operators have to stretch, bend and pick up when these actions could be avoided. Such waste is tiring for the employees and is likely to lead to poor productivity and, often, to quality problems.

Defects

The bottom-line waste is that of defects as these are direct costs. The Toyota philosophy is that defects should be regarded as opportunities to improve rather than something to be traded off against what is ultimately poor management. Finding and creation of solutions

(27)

Master Thesis – Andre Vollmer ₂₇

basis for the implementation of those changes in a later stage whereby the future state VSM serves as a blueprint. Also here the linkage of material and information flow shows the linkage of processes and first drafts should not take longer than a few days whereby fine-tuning always takes place during the implementation (Rother and Shook, 2003).

Evaluation

The proposed organizational changes in every stage get evaluated by accounting on the reduction of waste within each stage as described above. Here, every stage and therefore each involved product family of the solar system contribute to the problem statement of improving the efficiency of the downstream assembly process and to ensure customer service of the solar system.

In the chapter of the summarized potential of all stages, those outputs of each stage merge together and are evaluated by employing the proposed performance measures of labor costs and assembly lead-time. As already mentioned, the company point of view, which are labor costs, play the most important role. Moreover the customer point of view is accounted by ensuring the lead-time within the area of tolerance as described in the problem statement.

3.4. Organization of the project

The six months case study in an industrial environment is undertaken by the M.Sc. student in order to graduate as a Technology Manager at the Rijksuniversiteit Groningen (Netherlands) in the faculty of Economics & Business. Therefore the student is involved in a project team in the company whereby a supervisor at the university guides the theoretical considerations of the project. The project is undertaken in the product management department and the student is responsible to report every four weeks to the head of the department whereby the team leader is responsible for the guidance of the project more on a daily basis.

Project teams

(28)

Master Thesis – Andre Vollmer ₂₈

mentioning the particular system boundaries. This system includes functions from the Inbound Logistics departments of all involved product families, the warehousing and the assembly departments. The assembly stage differs from all other stages by describing the assembly operations in the main DC and therefore forces to include personnel that can provide knowledge about the processing of orders (Warehouse Logistics), the assembly process itself (commissioning team) and the coordination of both (head of warehousing).

Stages Required project teams

1. Assembly Commissioning Team, Head of Warehousing, _{Warehousing Logistics}

2. Supply of modules Controlling, SCM, Distribution, Product Manager, SAP Specialist, Inbound Logistics

3. Supply of inverters Controlling, SCM, Distribution, Product Manager, SAP Specialist, Inbound Logistics

4. Supply of the

mounting system Controlling, SCM, Distribution, Product Manager, SAP Specialist, Inbound Logistics 5. Supply of cables &

plugs Controlling, SCM, Distribution, Product Manager, SAP Specialist, Inbound Logistics

Table 4: Required project teams per project stage

All other remaining stages are structured around the respective product families and provide the physical input for the assembly process. Each of those stages required data input about wages of involved personnel, prices and margins of components (Controlling), the organization of the physical flow of goods (SCM) and how the customer orders are flowing into the system (Distribution). Furthermore the knowledge about technical details, life cycles and market knowledge about each product family (Product Management) needs to be considered and also access to SAP is needed (SAP specialist) in order to analyze the stocks in the warehouse. Finally the intersections with suppliers, facilitated by the Inbound Logistics department, sheds light into the product information flow into the system under consideration.

(29)

Master Thesis – Andre Vollmer ₂₉

3.5. Applied tools within each stage

The main toolboxes are the current and future state VSM methods by Rother and Shook (1998) and Jones and Womack (2002). Rother and Shook (1998) define their toolbox as the simple process of observing the flows of information and material as they now occur, summarizing them visually, and then envisioning a future state with much better performance. The whole point of value stream mapping is to disaggregate operational issues to the level of specific products where managers can more easily act on them. This is because the first objective of VSM is to achieve a breakthrough in shared consciousness of waste and to identify systematic opportunities for eliminating the waste.

Stages Current State _{VSM toolbox} Future State _{VSM toolbox} Evaluation

Workshop Workshop Assessment of

1. Assembly Power Point Power Point the seven

Stopwatch wastes

Web workshop Web workshop Assessment of

2. Supplies of

modules Power Point Power Point the seven

Net Viewer wastes

3. Supplies of

inverters Power Point Power Point the seven

Net Viewer wastes

4. Supplies of the

mounting system Power Point Power Point the seven

Net Viewer wastes

5. Supplies of

cables & plugs Power Point Power Point the seven

Net Viewer wastes

Table 5: Applied toolboxes and tools per stage of the project

(30)

Master Thesis – Andre Vollmer ₃₀

workshop tool is also used to identify material and information flows in the assembly processes and to learn to see value by the introduction of the seven commonly accepted wastes. Next to that, a stopwatch is introduced as a tool to measure the time of the individual assembly processes. All this is facilitated by a power point presentation, followed by a small discussion to clarify eventual questions. The toolbox of future state VSM contains also the workshop and power point tools where the proposed actions are introduced and discussed for their feasibility in practice with the respective team members. Finally the researcher is assessing the proposed actions with regards of the seven commonly accepted wastes.

The next stages differ slightly caused by their direct relation to a particular product family. Here the current state VSM additionally includes the Net Viewer tool, which facilitates it to conduct a web workshop employing ICT (Information and Communication Technology). Team members of the respective product families are geographically spread in other offices in Germany and need to elaborate the product and information flows on a whiteboard by webcam and voice over IP. Finally the researcher is also here assessing the proposed actions with regards of the seven commonly accepted wastes.

This chapter was answering the first sub-question, which in turn was asking for an improvement approach that suits the stated research objective. This was answered by merging the AR methodology with the VSM techniques and setting up the research approach in the above stated manner.

4. Stage 1: Assembly process

(31)

Master Thesis – Andre Vollmer ₃₁

4.1. Assessment of the current state situation

The assembly process merges four product families into a XXX and contains in total fifteen different parts and components (Appendix I). As an introduction, the following chart is depicting the value distribution of the different product families within a XXX. Value refers to the purchase price that the respective inbound logistics departments need to spend for the procurement of the required parts and components.

Figure 11: Value distribution of the distinctive product families in a Cenpac

Assembly always begins with an inflow of a customer order (1) into the system. Then the warehouse logistics personnel create a transport document via SAP (2) that is given to a commissioning team that in turn consists of two blue-collar workers. This team is able to scan a barcode on the document with a wireless scanner device and is then guided by an optimized computed route throughout the warehouse (3) for the picking of all parts and components (1-15) out of the shelves. Here, the VSM technique according to Rother and Shook (1999) is adapted from a production to an assembly setting by removing the different process steps. The assembly setting in the company forces the commissioning team to walk throughout the warehouse and collect fifteen parts and components out of the shelf. This activity is split in value creation time (picking parts out of the shelf) and non-value creation time but necessary to perform (walking throughout the warehouse or wrapping). In the appendix, information about the names of the involved numbered parts/components (1-15) and their product family is depicted in a table.

7.506 € 1.163 € 1.085 € 56 € 4.782 € 991 € _{750 €} 56 € 0 € 1.000 € 2.000 € 3.000 € 4.000 € 5.000 € 6.000 € 7.000 € 8.000 €

(32)

Master Thesis – Andre Vollmer

(33)

Master Thesis – Andre Vollmer ₃₃

The main activities of the assembly can be distinguished in five major process steps:

I. a.) Collection of nine small parts of the BOM: 0-20 min. I. b.) Collection of six components of the BOM: 20-40 min. II. Packing small parts in boxes and bags: 40-53 min.

III. Wrapping and fixing PV modules: 53-59 min.

IV. Arranging rails and boxes on pallet: 59-72 min. V. Adding product documentation and final wrapping: 72-75 min. Those process steps are conducted in the assembly system, which also contains inbound logistics and warehousing activities as explained in the first chapter. Now, additionally to the above-mentioned purchasing costs, the costs of labor hours thereby incur. The controlling department calculated for each labor hour in the assembly €26,69 including labor fringe costs. Moreover inventory costs of twenty-two hours weekly for the incoming goods inspection and putting the parts into the shelves occur. Thereby the costs of warehouse space is not considered due to the complexity that would be required for making an educated guess of such costs. The following picture depicts the assembly process steps within the system under consideration and highlights how the company was placing the customer order decoupling point (DP).

This DP separates the part of the supply chain that responds directly to the customer (pull) from the part of the supply chain that uses forward planning and a strategic stock (push) to buffer against the variability in the demand of the supply chain (Naylor et. al, 1999).

Inventory

Parts & Components

Inbound Logistics

- Modules - Mounting System - Inverters - Cables & Plugs

Assembly

Process steps: Ia + Ib + II + III + IV + V

Customer order decoupling

point

(34)

Master Thesis – Andre Vollmer ₃₄

A finer grained view of the assembly process depicts the value inclination with the picking of each part or component that is assembled over the time.

Figure 14: Current state value inclination in the assembly

The shaded area at the front end of the curve (Ia) is mapping the picking of small parts of the product families XXX and Cables & Plugs out of the shelves. The next shaded area (II) represents the wrapping of all those small parts in boxes and bags. The ratio of added value and assembly time between small parts and other components is shown below.

Total time Total value

3.x 5.x in minutes in % in € in % in € in % Small parts (Ia+II): 32 42,38% 275,19 3,87% 410,02 3,85% Components (Ib+III+IV+V): 43 57,62% 6.842,29 96,13% 10.244,79 96,15% Total: 75 100,00% 7.117,48 100,00% 10.654,81 100,00%

Table 6: Value/time ratio of involved parts in the assembly

It unfolds that the assembly of the small parts represent only 4% of the total value of the system, but require 42% of the total assembly response time. Moreover, the assessment of the current state situation of the assembly process unfolds the following wastes as to be detected, each to a certain degree.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 II III IV V ,0€ 2000,0€ 4000,0€ 6000,0€ 8000,0€ 10000,0€ 12000,0€ 0 4 6 8 10 12 14 16 17 20 22 24 28 32 35 42 52 59 69 75 Sy st em v al ue

Response time in minutes 3.x

5.x

(35)

Master Thesis – Andre Vollmer ₃₅

The table below is summarizing the detected degree of waste. Thereby a scale from very high (+ +) to intermediate (o) and very low (- - ) is used:

Type of waste Degree Overproduction

₊

Waiting

₊

Transport

₊

Inappropriate processing

+ +

Unnecessary inventory

- -

Unnecessary motion

- -

Defects

_{- -}

Table 7: Degrees of detected wastes within the assembly process

Inappropriate processing is detected by assigning a relative large amount of labor

hours to process only small parts like screws, cables and plugs. Overproduction is detected in times when demand is low and products were produced to stock in order to avoid under utilization of workers. Waiting is detected by under utilization of workforce in times of low customer demand. The waste of

Transport is detected by the non-value creating activity of walking throughout

the warehouse for the assembly of the small parts and cables & plugs, which could be avoided by shifting their assembly upstream of the DP.

4.2. Finding and creating a future state solution

(36)

Master Thesis – Andre Vollmer ₃₆

The picture below shows the shift of the process steps Ia and II upstream of the DP:

Assuming that the small parts are now available in a pre-assembled component, the future state value inclination curve therefore results by subtracting the shaded areas (Ia and II) out of the current state value inclination curve as shown below:

Figure 16: Future state value inclination in the assembly

As soon as a customer order flows in (1), the commissioning team receives a transport document (2) and can start the picking of the parts and components as usual. Then the scanner device (3) guides the commissioning team in the shortest computed distance throughout the warehouse. Now, the picking of the small parts 1-9 is facilitated via one ready-made component and executed within seconds. The picking of parts 10-15 conforms the process step Ib as described above and remains unmodified as described in the current state situation. It would result is an exhaustive amount of bounded capital and warehouse space to shift also their assembly more upstream of the DP. Process step II (Packing

Ia+II are pre assembled 10 11 12 13 14 15 III IV V ,0€ 2000,0€ 4000,0€ 6000,0€ 8000,0€ 10000,0€ 12000,0€ 0 3 7 9 11 15 22 27 37 41 Sy st em v al ue

Response time in minutes 3.x

5.x

Inventory

Parts & Components Process steps Ia + II

Inbound Logistics

- Modules - Mounting System - Inverters - Cables & Plugs

Assembly Process steps: Ib + III + IV + V Customer order decoupling point

(37)

Master Thesis – Andre Vollmer ₃₇

(38)

(39)

Master Thesis – Andre Vollmer ₃₉

4.3. Evaluation of the potential

In the assembly process waste of overproduction by the employees in times when demand of the market is low was detected. Shifting the pre-assembly of components upstream the DP to the first tier supplier reduces assembly lead-times and therefore reduces the need to buffer against upcoming demand peaks, which finally results in reducing such overproduction. The waste of waiting by customers is removed by the acceleration of the assembly process. Especially when customer demand reaches a peak, customer service is assured and the likeliness of not being able to serve a customer caused by constraints in the assembly area is reduced dramatically. Reducing the picking of nine different parts per XXX down to one pre-assembled component minimizes the waste of

transport. Inappropriate processing is reduced whereby a relative large amount

of labor hours was needed before to process only small parts like screws, cables and plugs. The outsourcing of those components to the first tier component supplier now reduces this waste. Next to that, unnecessary inventory is reduced by storing the components at the warehouse of the external service provider. Hereby storage space at the warehouse is saved and also a reduction of WIP inventory is realized. Transport is also decreasing dramatically by reducing the picking route from nine parts down to one single component. The waste of

unnecessary motion cannot be found because activities of workers neither

(40)

Master Thesis – Andre Vollmer ₄₀

The table below is summarizing the reduction of waste described above by depicting the degree of waste before and after the suggested implementations of redesigning the assembly system. Thereby a scale from very high (+ +) to intermediate (o) and very low (- - ) is used:

Type of waste Degree: Current state Degree: Future state Overproduction

₊

- -

Waiting

₊

- -

Transport

₊

- -

+ +

- -

Unnecessary motion

- -

Defects

_{- -}

- -

Table 8: Degree of waste reduction within the assembly process

(41)

Master Thesis – Andre Vollmer ₄₁

This chapter was answering the second sub-question by providing insights into the reduction of wastes in the assembly process and the improvement of efficiency and ensuring customer service.

5. Stage 2: Modules supplies

This stage will firstly assess the current situation of the supplies of the product family XXX, which are the photovoltaic (PV) modules. Thereby the physical flow of goods and the information flow will be described and result in a current state value stream map. Secondly, an improved future state value stream map will be drawn up and thirdly evaluated.

PV modules are responsible for generating electricity out of the sun irradiation and made for the on-roof installation on residential homes. The involved mono- or polycrystalline PV modules are available in five power classes from 205-225 Watt peak (Wp) and have all the same outer dimensions, which is important to consider when a logistical focus is taken as it is the case here. Each customer order of a XXX contains only mono- or polycrystalline modules, each in the same power class.

(42)

Master Thesis – Andre Vollmer ₄₂

leaves the production site, the next lorry already arrives to pick up the next production output. From there the modules get delivered to the main DC in XXX and stored again in the shelves and are waiting for their assembly in a XXX as already described in chapter four.

(43)

(44)

Master Thesis – Andre Vollmer ₄₄

Moreover, the assessment of the current state situation of the modules unfolds the following wastes as to be present, each to a certain degree. The column with the indication of the degree describes a scale starting with (+ +) for very strong, the digit zero (o) for intermediate and (- -) for not found.

- -

Waiting

- -

Transport

- -

Unnecessary motion

- -

Defects

- -

Table 9: Degrees of detected wastes within the product family of the modules supply

The logistics of the modules that are delivered from the XXX in XXX/Germany via XXX to XXX is executed via a third party logistics provider and not part of the system under consideration. The election of this logistics party was facilitated by a call for tenders in the respective geographical region and the examination of those processes are outside of the scope of the managerial impact of this project. While taking a structured walk along the value stream within the assembly system, no potential was unfold concerning the seven wastes in order to contribute to the optimization of the solar system and therefore the future state map for the product family of the modules remains unaltered as shown in the current state.

(45)

Master Thesis – Andre Vollmer ₄₅

assigned to each customer order and therefore the accuracy of demand is increasing dramatically. Waiting is not found in the assembly system because orders from suppliers arrive on a daily basis and the module processing in the assembly follows immediately. Therefore neither workers nor material suffer from waiting. Concerning transport, the wireless scanner of the commissioning team is connected with SAP and computing the shortest distances of picking routes of each order in the warehouse. Also double handling or excessive movements are not found within the system. Inappropriate processing is not the case because the actual state of the module assembly is not executed in overly complex solutions or requiring large machines. The commissioning team is simply using a small forklift for the pallets and the wireless scanner device.

Unnecessary inventory is not held caused by the already explained standardized

BOM. Now, a particular amount of modules can be assigned to each order and strategic inventory to buffer against variability is not necessary anymore.

Unnecessary motions concerning ergonomics in production where operators

have to stretch, bend and pick up do not exist. The commissioning team consists of two workers and the heaviest component, which is the module, can be moved easily from the pallet of the storage to that is being shipped. Defects are also not appearing because the assembly process itself is not employing large machines to process those goods. Also modules are stored only on the lowest level of the high racks in order to avoid accidents with the forklift.

Efficiency: Due to the fact that no waste was detected and no suggestions are made for redesigning the system, no additional efficiency gains can be realized by the modules inflow to the solar system.

(46)

Master Thesis – Andre Vollmer ₄₆

This chapter was answering the third sub-question by providing insights into the potential of the modules to contribute to the problem statement. The fact that no waste is currently in the system, the modules are not contributing to optimize the current state value creation processes.

6. Stage 3: Inverter supplies

This step will assess the current situation of the supplies of the product family

XXX, which are the inverters. Thereby the physical flow of goods and the

information flow will be described and result in a current state value stream map. Here the chapter differs from other chapters in which VSM is applied by additionally introducing a postponement strategy to reduce the detected waste. Secondly, an improved future state value stream map will be drawn up, assuming the postponement strategy will be applied and, thirdly, evaluated.

An inverter is responsible to transform the generated direct current from the photovoltaic (PV) modules into alternating current and to finally feed this electricity into the public power grid. Each power class of a XXX (3.x and 5.x) needs also a distinctive inverter (3.6 or PS 5.5) and each inverter also includes a pre-installed maximum power point (MPP) tracking software. Such software optimizes the feed-in efficiency of the inverter when the irradiation of the sun in de- or increasing during the normal daytime from dawn to dusk.

(47)

Master Thesis – Andre Vollmer ₄₇

(48)

(49)

Master Thesis – Andre Vollmer ₄₉

Moreover, the assessment of the current state situation of the inverters unfold the following wastes as to be present, each to a certain degree. The column with the indication of the degree describes a scale starting with (+ +) for very strong, the digit zero (o) for intermediate and (- -) for not found.

_{- -}

Waiting

_{- -}

Transport

_{- -}

- -

+ +

Unnecessary motion

- -

Defects

_{- -}

Table 10: Degrees of detected wastes within the product family of the inverter supply

The examination of the value stream of the inverters was unfolding in particular the waste of unnecessary inventory. The proposed solution below was found while conducting a collaborative group session which is proposed by the Action Research (AR) methodology according to Trist (1979). In this group session, the researcher took the role of the observer and forced collaborative learning between a group member (the respective Product Manager) and himself.

Postponement strategy

(50)

Master Thesis – Andre Vollmer ₅₀

a specific customer demand until the latest possible point in the supply chain (Van Hoek, 2001). Here Lean thinking should raise the question: Why already differentiate a product when its real demand is not even clear yet?

Considering the XXX (PS X.X KWp) inverters used for the XXX, it is unfold that the application of the postponement strategy is technically feasible and, moreover, raises potential for cost reduction in logistics. The company holds in total six power classes (3.0KWp, 3,6KWp, 4,2KWp, 5,5KWp, 8,3KWp, 10,1KWp) of the

XXX inverters in their product portfolio. Whatever, those six power classes are

constructed out of three different hardware types. The installed MPP tracking software customizes each hardware type into two power classes as depicted below.

The customer order decoupling point (DP) separates the part of the supply chain that responds directly to the customer (pull) from the part of the supply chain that uses forward planning and a strategic stock (push) to buffer against the variability in the demand of the supply chain (Naylor et.al, 1999). In the current situation, planning, forecasting and strategic stock keeping have to be made considering a product variability of six different power classes. Nevertheless, in order to increase responsiveness to changing customer demand, to minimize stock keeping and to reduce complexity for forecasts, the implementation of the software more downstream in the supply chain would improve all those areas

XXX XX Customer Hardware type #1 Hardware type #2 Hardware type #3 PS 3.6 KWp PS 3.0 KWp PS 4.2 KWp PS 5.5 KWp PS 8.3 KWp PS 10.1 KWp Customizing via

software decoupling point Customer order

PS 3.0 KWp PS 3.6 KWp PS 8.3 KWp PS 10.1 KWp PS 4.2 KWp PS 5.5 KWp PS 3.0 KWp PS 3.6 KWp PS 4.2 KWp PS 5.5 KWp PS 8.3 KWp PS 10.1 KWp

(51)

Master Thesis – Andre Vollmer ₅₁

simultaneously. In order quantify the described problem of unnecessary inventory, the inverter stock in the year 2010 is depicted below:

Figure 21: Inverter stocks in 2010 per power class

The unexpected feed-in tariff (FIT) reduction in July 2010 was affecting tremendous uncertainty in the whole PV market as already explained in chapter two by describing the problem context. There in Table 2: Company supply vs. customer demand in the PV industry, the second demand peak resulted in a dramatically dropdown after the announcement of the unscheduled FIT reduction. This volatile customer demand was finally triggering the overall decreased inverter stock of the company after July 2010 as depicted above. The PS 3.6 stocks were increasing up to 312.000€ in November 2010 and up to 360.000€ in December 2010, but readability of the chart above forced to cut the graph in October 2010. Below, the future state supply chain of the inverters is drawn up assuming that the postponement of the software implementation is executed: 0,00€ 20.000,00€ 40.000,00€ 60.000,00€ 80.000,00€ 100.000,00€ 120.000,00€ 140.000,00€ 160.000,00€ 180.000,00€ 200.000,00€ In ve n to ry v al u e PS 3.0 PS 3.6 PS 4.2 PS 5.5 PS 8.3 PS 10.2

Figure 22: Future state of the inverter postponement strategy

XXX

XX Customer

Customer order decoupling point &

(52)

Master Thesis – Andre Vollmer ₅₂

Now, the DP and the point of customization via proprietary MPP tracking software merge together more downstream to the customer. Customization is only executed when the customer has already placed his particular orders of solar systems and therefore the real demand of the inverter power class is already clear. The stock in 2010 is again drawn up below, but now per hardware type, and shows that the combination of the two power classes into their harware type already smoothens the graph of the hardware type #1 and #2:

Figure 23: Inverter stock in 2010 per hardware type

The following example assumes that the software changes could be undertaken by pressing a switch on the circuit board of the inverter. That is also feasible in reality, only an R&D project had to be triggered for that. Nevertheless, there are two opportunities:

(1) The first opportunity would be to press the switch in the main DC in Paderborn where also the assembly of the XXX takes place. On the one hand it would be invisible for installers and end-customers that they might buy a downgraded product (e.g. Hardware class #1 is bought, but the MPP tracking software of the PS 3.0 is installed which optimizes the feed-in of max. 3000 Watt. The same hardware class would be able to optimize the feed-in of 3600 Watt, which would be the more expensive PS 3.6 inverter). On the other hand the assembly lead-time of the complete system would increase dramatically. The blue-collar workers had to unwrap and open the box of the inverter and lift it out of the box. After that they would have to release two screws of the inverter enclosure and press the hidden switch

(53)

Master Thesis – Andre Vollmer ₅₃

on the circuit board. Finally, the enclosure had to be fixed again and everything had to be wrapped again.

(2) The second opportunity would be to instruct the installers that they can press the switch at the installation site where the PV system is finally delivered, unpacked and installed. On the one hand the assembly lead-time would remain untouched and the task of pressing the button would also not overburden an installer. On the other hand when an installer would know that e.g. the inverter PS 3.0 (max feed-in of 3000 Watt) could also be used to feed-in 3600 Watts by only pressing a button, it would be impossible to expect someone to still buy the more expensive PS 3.6 inverter.

Whatever, it is further assumed that an appropriate solution would be at hand for implementing the software later in the supply chain neither on the expenses of the assembly lead-time nor by cannibalizing the own products as described above.

(54)

Master Thesis – Andre Vollmer ₅₄

121.082,10€ x 7% p.a. = 8.475,75€ p.a.

In order to study the variation of the savings when the 50% of stock reduction is fluctuating, the following sensitivity analysis is clarifying this dependency:

Stock reduction Excessive bounded capital Costs of capital p.a. (7% interest) 10% €24.216,42 €1.695,15 20% €48.432,84 €3.390,30 40% €96.865,68 €6.780,60 60% €145.298,51 €10.170,90

Table 11: Sensitivity analysis of the inverter postponement savings

(55)

Master Thesis – Andre Vollmer ₅₅

The next problem is how to conduct the individual sales of the inverters. Inverters are not only sold in solar systems but also individually to solar installers. The ability of inverters to operate also within a higher power class with another software needs to be invisible for installers and customers, otherwise they would perceive to pay too much for a “downgraded” product. In order to summarize all aspects of the inverter postponement strategy, a SWOT analysis is used as shown below:

Strengths Weaknesses Opportunities Threads

Less complex planning and forecasting

R&D project

needed Increased availability for customers (inverters were a serious bottleneck in 2010) Customers don't want to buy a downgraded product Reducing risk of inverter stock-out

Who and where to implement the software?

How to realize individual inverter sales that refer to KWp prices? Cost reduction in

logistics How to label the product and the outer packing? Increased flexibility to fluctuating customer demand How to realize individual inverter sales that refer to KWp prices? Customization according to real demand, not on forecasts anymore How to avoid cannibalizing own products when postponing?

Table 12: SWOT analysis of the inverter postponement strategy

Until now, this chapter was concerning exclusively with the waste of unnecessary

inventory. The waste of overproduction was not found in the assembly system

(56)

Master Thesis – Andre Vollmer ₅₆

was also not found because the inbound logistics, the warehousing and the assembly processes of the inverters are not requiring overly complex solutions or large machines as it is the case in a production setting. Unnecessary motion concerning ergonomics in production where operators have to stretch, bend and pick up do not exist. The commissioning team consists of two workers and the inverter can be moved easily, even by one person, from the pallet of the storage to the pallet that is finally being shipped to the customers. Defects are also not appearing because the assembly process itself employs no large machines to process the inverters. The involved inbound logistics, warehousing and assembly processes require only the movement of a well wrapped box of the inverter and information from the past about caused defects confirm this assertion.