Excelling the game of POLCA cards

(1)

– faster and better

(2)

T it le P a g e 2 University of Groningen

Faculty of Economics and Business

Title Page

Excelling the game of POLCA cards – faster and better

Improvement in current lead times of existing POLCA implementation at Bosch Hinges and Metal

Final Master Thesis

Technology Management

Student: Jana Pejchinovska Study programme: Technology Management

Student number: S1946595 Company: Bosch Scharnieren en

Metaal B.V

First supervisor: dr. J. Riezebos Document version: Final thesis Co-assessor: dr. M.J. Land Date: January 2012 Company Supervisor: Godfried Kaanen

(3)

P re fa ce 3

Preface

This thesis is the last task of the learning experience of the Master program in Technology Management at the Faculty of Economics and Business at the University of Groningen. The master itself and the internship project for the thesis have been exciting journey filled with challenges and achievements. Getting here would not have been possible without the help and support of people to whom I wish to express my appreciation and gratitude.

First, I would like to thank Dr. Ir. Jannes Slomp, Director of the Lean Operations Research Center (LO-RC) and Adjunct Professor at the Department of Operations at the Faculty of Economics and Business. His assistance span over the process of introducing me to Bosch Hinges and Metal and provided me with valuable insights about the process of writing a master thesis. I am grateful for the given opportunity and support during the initial start of the project.

This thesis would not have been possible without the support from all the employees at Bosch Hinges and Metal. They have openly welcomed me as part of the team, respected me as a fellow colleague, kept their doors open and were always willing to answer my questions and collaborate with me. Particularly, I would like to thank Ferry Senten, Production Manager at Bosch Hinges and Metal, for his enthusiasm for collaboration and willingness to help me understand their business processes to the finest details.

I would also like to thank Robert Peters, from Soft Tools Manufacturing, the company behind the software solution used at Bosch Hinges and Metal, for providing me with unabridged access to data and many helpful insights.

During the internship it has been a great privilege to be guided by and learn from the Managing Director of Bosch Hinges and Metal, Godfried Kaanen. A true QRM and POLCA champion, Fried has not only taught me valuable insights about the practical implications of QRM, but also helped me to grow and develop as a professional. His support and enthusiastic sharing of knowledge have made this internship a remarkable experience that I will always remember. I am very grateful for the given opportunity and for introducing me to the world of QRM professionals in the Netherlands.

My appreciation goes to Dr. Martin J. Land, Associate Professor with the Department of Operations at the Faculty of Economics and Business, who is my thesis co-assessor. Although our collaboration has been brief, he provided me with valuable comments to my work, which spurred thoughts on how I could improve it.

(4)

P

re

fa

ce

4

me through dilemmas and providing me with excellent feedback. His understanding, encouragement and guidance have been essential to this thesis.

There are few more thanks that I would like to make on a personal note. I would like to thank my colleagues Jens, Joost, Thiemo and Jeroen from the TM master program, with whom learning and teamwork has been fun experience with great achievements.

I have been very fortunate to have people both in my home country and here in the Netherlands that have supported me throughout this journey. Vesna, Cadik, Bibi, Vasko, Deki, Ana, Bela, Jordanka, Kate, Sani, from back home, your inspiring friendship, understanding and support have been very rewarding and I am grateful for that. My gratitude is also due to my fellow friends from Groningen: Jose, Charlotte, Felicity, Tomas, Dejan, thank you for the companionship, entertainment and caring. Lastly, for the people that have become my second family here: Maja, Anca, Alok, Astrid, Jelena, Georgia, Kostas, and Pavle, thank you for the wonderful friendships you have given me.

Lastly, I owe my most profound and deepest gratitude to my parents, Dimitar and Mimoza, and my sister Marija. Their unconditional support and love has brought me here and helped me walk every step of the way. I am indebted to them for raising me with love for science and supporting me in all my pursuits. Although we have been apart in three different places of this world, without their understanding and encouragement it would not be possible to complete this work. It is to them that I dedicate this thesis.

Utrecht, January 2012,

(5)

A b st ra ct 5

Abstract

Today’s manufacturing organizations are pressured to perform better and deliver their products to the customers faster in order to survive in the competitive markets. Such phenomenon has spurred many investigations and researches both in the academic environment and in organizations. It is also the main focus of this thesis. Here, the question of improving the lead times has been expanded to accommodate a specific setting: implementation of Quick Response Manufacturing and POLCA material control. It is a strategy for lead time reduction aimed at companies that deliver custom-made products in small quantities.

(6)

< T a b le o f C o n te n ts 6

List of figures

Figure 1. Process flow at Bosch Hinges and Metal ... 15

Figure 2. Example of generic POLCA card ... 19

Figure 3. Example of order flow and usage of POLCA cards© ... 20

Figure 4. Example of order flow and usage of POLCA cards – continued© ... 20

Figure 5. Conceptual Model ... 23

Figure 6. POLCA Audit Steps (adapted from Riezebos, 2010) ... 26

Figure 7. Cellular structure at Bosch Hinges and Metal (July 2011) ... 29

Figure 8. Assignment of workers per machine based on capability to operate ... 45

Figure 9. Distribution of time spent per operation in the measured period ... 47

Figure 10. Schematic illustration of the current order release strategy for product components ... 65

Figure 11. Schematic illustration of proposed order release strategy for assembly operations (hinges) ... 66

Figure 12. Schematic illustration of proposed order release strategy for welding operations ... 67

Figure 13. POLCA cell task ownership protocol ... 71

Figure 14. Example of partial-MCT map for piano hinges from stainless steel, new order ... 72

Figure 15. Example of partial-MCT map for piano hinges from stainless steel, repeated order... 73

Figure 16. An example of the monitoring function of the PROPOS Dashboard ... 81

Figure 17. An example of the authorization function of PROPOS Dashboard ... 82

Figure 18. Example of Touch Screen from PROPOS in front of the cell ... 83

Figure 19. Step 1 in comparison of data findings ... 87

(9)

S e tt in g t h e s ce n e 9

1. Setting the scene

1.1.Brief discussion of current trends in manufacturing

The manufacturing environment of the 21st century and the last decades of the 20th century has been marked by a paradigm shift which involves evolution of the competitive context with organizations operating in increasingly global and uncertain environments. These conditions have put emphasis on customer-driver manufacturing characterized by differentiated products and features, tight delivery performance and production at low cost. All these criteria have become essential and crucial for organizations to survive in such business climate (Zorzini, Corti, & Pozzetti, 2008). To avoid losing important customers, organizations are pressured to quote delivery dates that are both attainable and attractive not only to current, but also to potential customers. Hence, short throughput times can bring business, but if not delivered can also contribute to losing one (Slotnick & Sobel, 2005).

Management of the throughput time, which is related to the criteria of tight delivery performance, is particularly challenging for Make-To-Order (MTO) organizations. These companies cannot use finished goods inventories to buffer against the variation in demand in a way Make-To-Stock (MTS) companies do (Teo, Bhatnagar, & Graves, 2011). MTO organizations are those organizations that produce high variety of products in relatively low volumes. Additionally, to make difference between MTS that have converted to MTO due to change in demand, true MTO companies have always manufactured on this basis, due to the nature of the products and the relationships with their customers (Hendry, 1998). MTO companies are further characterized with unique customer orders that cannot be made in advance and thus hold no inventories (Mestry, Damodaran, & Chen, 2011). While these organizations are often small in size they aim to be capable of competing with the best organizations in their industry. The “best practices” applicable to MTO companies that can be found in literature, are similar to the main principles of World Class Manufacturing and can be classified in several categories:

[1] Workforce empowerment

[2] Design for products, processes and improved supplier relations [3] Simplified shop floor

[4] Capacity issues [5] Improved quality

[6] Appropriate planning and control systems

[7] Performance measurement, benchmarking and continuous improvement

(10)

Manufacturing production philosophies which advocate the cell approach. Performance measurement is crucial element of achieving improvement. The most important factors usually include quality, on-time delivery, reduction of throughput time and production flexibility. This is often enhanced by benchmarking and together they lead to continuous improvement (kaizen) (Hendry, 1998).

The choice of appropriate production planning and control system is especially important for MTO organizations. These companies cannot rely on the universal tools available due to their specific nature. Moreover, these organizations are usually constrained by limited financial resources, putting additional pressure on the choice of the adequate planning and control system (Stevenson, Hendry, & Kingsman, 2005). Lastly, in the plethora of available planning and control systems (MRP, MRP II, WLC, Kanban, JIT, CONWIP, e-SCM just to name a few), how can an efficient decision that will satisfy all the objectives, be made?

1.2.Quick Response Manufacturing as strategy for addressing the

current challenges in manufacturing

At the end of 1990s, an approach for improvement in manufacturing that was targeted for MTO organizations in particular, characterized by increased degree of customization, and low order volumes was introduced. The significance of this approach was that it combined several of the categories for “best practices” discussed above. Quick Response Manufacturing (QRM) is a “company-wide strategy for reducing the lead times in all aspects of the company, both externally and internally”. The external aspect involves design and manufacturing of products that satisfy specific customer needs. The internal aspect is concentrated on reducing the lead times, resulting in improved quality, lower costs and quicker response. This system has been designed to overcome the limitations of implementing lean production in organizations that pursue “strategic variability”, a term coined by the author of QRM, Rajan Suri to describe organizations that operate in markets with highly unpredicatble demand and offer custom-engineered products with large variety of options. As a result QRM offers methodology and tools to exploit this strategic variablity to achieve competitive advantage (Suri, 1998).

QRM has been built around four pillars:

[1] The power of time [2] Organizational structure

[3] Understanding system dynamics [4] Enterprise-wide application

(11)

principles of QRM is redesign of the organizational structure to accommodate exploitation of strategic variability and quicker response. For that purpose the QRM proposes cellular structure to be applied not only to the shop floor, but also to the entire organization (Suri, 2010; Suri, 2003). The implementation of the cells should be further supported by cross trained workforce and self managed teams that retain strong focus on the throughput time reduction. This involves multi skilled operators, which are required to offset the fluctuations in utilization due to variable product mix and setting ownership and responsibility for jobs. QRM also calls for change in the perception of the system dynamics, in particular capacity and utilization. According to the QRM principles, production should be set at 80% of the avaialable capacity to accommodate changes and provide capacity availablity for rush-orders. Lastly, QRM goes beyond the optimization of the shop floor involving changes in the office, the supply chain, customers, material planning, shop floor control, and new product development (Suri, 2010).

To satisfy the new challenges of implementing QRM in organizations, a new material control system had to be developed. QRM is supported by Paired-cell Overlapping Loops of Cards with Authorisation (POLCA). POLCA is a hybrid system that combines the best of both worlds of card-based pull systems and push systems, while overcoming the limitation posed by each when applied to engineer-to-order business model (Suri & Krishnamurthy, 2003).

1.3.Topic choice decision and relevance

QRM and POLCA have been sucessfully implemented in variety of manufacturing organizations, dominantly in United States. These implementations range from large-scale manufacturing companies to small companies. However, there have also been implementations in Europe, mostly in the Netherlands and Belgium. One of the most prominent examples of application of QRM and POLCA is to be found in the company Bosch Hinges and Metal, the first company to implement QRM in Europe. While Bosch Hinges and Metal is small manufacturer of custom-made hinges, QRM practices can be found implemented in larger companies, such as Harol, in Belguim, manufacturer of custom-made rolling shutters.

In the decade and more since the inception of QRM and POLCA, literature has documented variety of research on the topic. There have been numerous case studies of implementation (Krishnamurthy & Suri, 2009; Suri & Krishnamurthy, 2003), research on designing POLCA systems (Riezebos, 2010; Krishnamurthy & Suri, 2009), and variations of POLCA (Generic POLCA, Load-based POLCA, e-POLCA) (Fernandes & Carmo-Silva, 2006; Vandaele, Van Nieuwenhuyse, Claerhout, & Cremmery, 2008). The unifying element is that all these researchers have analyzed and discussed the transition from traditional job-shop environments to QRM. Such condition revealed a potential for exploring an organization that has implemented this manufacturing system for several years now and it is evidencing the benefits.

(12)

(13)

B o sc h H in g e s a n d M e ta l: c o n te x t a n d o rg a n iz a ti o n a l ch a ra ct e ri st ic s 13

2. Bosch Hinges and Metal: context and organizational characteristics

2.1.Brief introduction of the company

Bosch Hinges and Metal (original name: BOSCH Scharnieren en Metaal B.V) is a metal-producing company located in Doetinchem, the Netherlands. The main activity of the company is production of metal custom-made hinges and light metal construction for industrial applications. The materials used in production are mostly steel, stainless steel and less frequently brass, aluminum or other metals. The company’s end products are parts (hinges or metal constructions) that are further used by their customers in production. The company serves around 500 active clients who are solely industrial customers in several different industries, such as: metal industry, trucking industry, marine hinges, machine building, interior contractors, trains and metros, industrial hardware, and individual applications (Metaal, Over Ons, 2011).

The production environment at Bosch Hinges and Metal can be best characterized as Make-To-Order (MTO), meaning that production does not take place until the customer has confirmed the order (Stevenson, Hendry, & Kingsman, 2005). The MTO production strategy implies that the company manufactures high variety of customer made products in small volumes, a situation known in literature as low-volume/high-variety environment (Slomp, Bokhorst, & Germs, 2009). Bosch Hinges and Metal annually produces around 500 different products, a number that is constantly growing over the years, which illustrates the high-variability. The orders size can vary from 1 – 2000 pieces (limited by the management), where the average order size is 350 pieces. The average order size is decreasing over the years (it was greater than 500 in 2006) and it is common occurrence that order sizes are between 50 – 100 pieces. These numbers attest to the low-volume production environment (Metaal, Company-wide transformation using QRM and POLCA, 2011).

(14)

D e sc ri pt io n o f m a in pr o d u ct f lo w s a n d m a in pr o ce ss e s 14

Another interesting characterization of MTO production strategy is that it is mostly adopted by Small and Medium Size Enterprises (SMEs), defined as companies with less than 50 employees (Stevenson, Hendry, & Kingsman, 2005). Such is the case with Bosch Hinges and Metal, which has only 25 full time employees, and 80% of them are workers on the shop-floor. Given that SMEs are often dealing with limitations of financial resources, choosing the appropriate production planning and control approach is vital to the success of the company (Stevenson, Hendry, & Kingsman, 2005).

2.2.Description of main product flows and main processes

The previous section indicated that Bosch Hinges and Metal annually produces approximately 500 different products, implying that each customer requires different and unique product. Within such setting it is difficult to deduce specific product groups or families. However, evidence from the observations shows that there are four different product flows that can be distinguished. Each of the different products can belong to one of these product flows:

[1] Light metal construction [2] Heavy hinges

[3] Piano (continuous) hinges [4] Special hinges

Light metal construction product flow groups together all metal construction products, such as racks, tables or other custom-made products. Heavy hinges product flow is a joint label for all customer-made hinges that use materials thicker than 6mm. Piano hinges is a product flow category for continuous custom-made hinges that have length up to 2.5m for standard piano hinges or 4.2m for bespoke piano hinges. Lastly, special hinges is a product flow cluster for the remainder of bespoke hinges that do not fall in the previous two product flow categories.

(15)

Figure

2.3.Summary of improvement activities

Initially the company operated in a typical job shop setting with low volume high variety type of production. The lead times offered to

delivery was 8 weeks. In 2004 the company was faced with production challenges resulting in long and unreliable lead times, back orders, errors in production, overproduction, excessive inventory and lack of skills. Such challenges lead many of their clients to cancel orders or leave. The situation compelled the management to consider improvement strategy. The solution was found in adopting the principles of lean manufacturing. By applying 5Ss improvement program and Single

managed to fix the lead time at 6 weeks and reduce the inventory. While some improvement was achieved, the company identified that some of the lean principles

and elimination) did not apply well to their MTO production environment. After examining the different production control system and solutions that would be suitable, in 2007 the company implemented the Quick Response Manufacturing (

POLCA material control system. The implementation was successful, resulting in reduction of lead time to 4 weeks and increase in both delivery reliability and productivity.

the introduction of QRM and POLCA on the production managing the visual signaling and fine

engineering company (Soft Tools Manufacturing) they have developed the monitoring software Production and POLCA Observation System (PRO

2.4.Current problem areas and challenges

An order winning criteria in MTO organizations that have low

production environment are product variety, ability to offer customized products with additional features and delivery speed

inherent to the business model applied at Bosch Hinges and Metal. The company has identified the delivery speed, or more specifically competitive lead times, as their competiti advantage. Furthermore, the proliferation of the PROPOS monitoring software has enabled them to gain insights on how the system functions. This

Raw material Metal-sheet cutting and punching

S

u

p

l

i

e

r

s

Figure 1. Process flow at Bosch Hinges and Metal

Summary of improvement activities

Initially the company operated in a typical job shop setting with low volume high variety type of production. The lead times offered to customer were set at 6 weeks, however, typical delivery was 8 weeks. In 2004 the company was faced with production challenges resulting in long and unreliable lead times, back orders, errors in production, overproduction, excessive kills. Such challenges lead many of their clients to cancel orders or leave. The situation compelled the management to consider improvement strategy. The solution was found in adopting the principles of lean manufacturing. By applying 5Ss

am and Single-Minute Exchange of Dies (SMED), (Nicholas, 2008) managed to fix the lead time at 6 weeks and reduce the inventory. While some improvement was achieved, the company identified that some of the lean principles (variability reduction and elimination) did not apply well to their MTO production environment. After examining the different production control system and solutions that would be suitable, in 2007 the company implemented the Quick Response Manufacturing (QRM) accompanied with POLCA material control system. The implementation was successful, resulting in reduction of lead time to 4 weeks and increase in both delivery reliability and productivity.

the introduction of QRM and POLCA on the production floor, the company worked on managing the visual signaling and fine-tuning the system. In 2010 together with a software engineering company (Soft Tools Manufacturing) they have developed the monitoring software Production and POLCA Observation System (PROPOS).

Current problem areas and challenges

An order winning criteria in MTO organizations that have low-volume/high

production environment are product variety, ability to offer customized products with additional features and delivery speed (Jina, Bhattacharya, & Walton, 1997)

inherent to the business model applied at Bosch Hinges and Metal. The company has identified the delivery speed, or more specifically competitive lead times, as their competiti advantage. Furthermore, the proliferation of the PROPOS monitoring software has enabled them to gain insights on how the system functions. This in turn has led to recognition that

sheet cutting and punching Parts machining Assembly S u m m a ry o f im pr o v e m e n t a ct iv it ie s 15

Initially the company operated in a typical job shop setting with low volume high variety type customer were set at 6 weeks, however, typical delivery was 8 weeks. In 2004 the company was faced with production challenges resulting in long and unreliable lead times, back orders, errors in production, overproduction, excessive kills. Such challenges lead many of their clients to cancel orders or leave. The situation compelled the management to consider improvement strategy. The solution was found in adopting the principles of lean manufacturing. By applying 5Ss (Nicholas, 2008) they managed to fix the lead time at 6 weeks and reduce the inventory. While some improvement (variability reduction and elimination) did not apply well to their MTO production environment. After examining the different production control system and solutions that would be suitable, in 2007 the QRM) accompanied with POLCA material control system. The implementation was successful, resulting in reduction of lead time to 4 weeks and increase in both delivery reliability and productivity. Following floor, the company worked on tuning the system. In 2010 together with a software engineering company (Soft Tools Manufacturing) they have developed the monitoring

volume/high-variety production environment are product variety, ability to offer customized products with (Jina, Bhattacharya, & Walton, 1997). The first two are inherent to the business model applied at Bosch Hinges and Metal. The company has identified the delivery speed, or more specifically competitive lead times, as their competitive advantage. Furthermore, the proliferation of the PROPOS monitoring software has enabled turn has led to recognition that Finished product

C

(16)

M a n a g e m e n t q u e st io n s 16

there are areas for further improvement with respect to the lead times. Their current strategic orientation is to further reduce lead times by 1,5 – 2 weeks.

Additionally, in late 2010 the company acquired small family owned metal producing factory, Huls Metaal, from Etten, the Netherlands. The new operations (light metal constructions) were incorporated in the existing production system in April 2011. These operations are currently serving the interior design industry clients and are characterized with very competitive lead times of 1 – 2 weeks. The light metal construction operations have been organized to conform to the existing QRM principles. However, in practice the transition revealed challenges resulting in errors in production, variable lead times (from 2 – 4 weeks), and backorders. Consequently, the company is focused on re-evaluating the existing production strategy, the POLCA material control system and investigation of the nature of metal construction operations, to ensure seamless integration with the existing system and reduce lead times.

2.5.Management questions

The above discussed company situation, their challenges and problem areas was the starting point of this research. The start of this research project was initiated by two management questions that are outlined below.

How can the existing QRM implementation and POLCA material control system be further improved to facilitate lead time reduction and achieve lead times of 2 weeks?

(17)

T h e o re ti ca l fo u n d a ti o n o f th e t h e si s 17

3. Theoretical foundation of the thesis

3.1.The building blocks of QRM: Cellular Manufacturing

Cellular Manufacturing Systems emerged from the application of Group Technology principles that were designed to search for similarity within the production system. This general concept has been applied to factory reconfiguration and shop floor redesign. Consequently, cellular manufacturing is a model for layout design and a “production system decomposed into cells, able to produce several operations per work order” (Riezebos, 2001). Consequently, the cells are not responsible for producing only one product, but rather a family of products. Such mix leads to imbalances in machine utilization within the cell. Hence, the resources assigned to cell should be multi-skilled and able to operate all the machines, to handle these fluctuations. Ergo, a cell can be defined as “organizational unit consisting of a set of resources (workers, equipment or information), capable of processing several operations per work order” (Riezebos, 2001).

The design and implementation of cell-based manufacturing system is a complex process that can be best described as consisting of three building elements: system design, system control and system integration.

The system design element refers to the cell design and as such it should satisfy three criteria. The first criterion is that cells should be organized around identifiable product groups or product families that can be manufactured within the identified manufacturing area, or the cell. Generally, such design decisions involve relying on modular product architecture and/or standardization of parts. However, these are more general guidelines, as the actual application is organization specific. Next, each cell should have clearly demarcated boundary of operations. Such boundary allows easier control of the operations performed within the cell as well as monitoring of the production. However, the cell boundary should also exhibit flexibility to accommodate changes in the cell dynamics. Lastly, the third criterion proposes that each cell should have dedicated set of workers, who can work with all the flexibility of a team. It is important to note, that although cells are designed as autonomous units, with clear boundaries and dedicated staff, some movement of workers between cells is unavoidable and can foster exchange of good working practices.

The second building element, the control of the operations in a cell must be applied from within the cell. This involves control of the work, machines, tooling and maintaining the schedule. As such, it can be assisted by use of visual boards or software. Similarly, performance should be measured within the cell. While the cell functions as separate entity when it comes to control, it cannot function independently and represents a part of a bigger system, the organization.

(18)

cells, production priorities, and creation of strategy. For a successful application of cellular manufacturing, in addition to following the guidelines about system design and system control, cells must have corresponding interfaces to enable integration and functioning system-wise (Prickett, 1994).

The benefits from implementing cellular manufacturing can be summarized in the following list:

Decompose large and complex manufacturing system into several smaller and more manageable systems

Simplified quality control Reduced material handling

Reduced throughput times of successive operations that reduces the WIP Reduced transport time, due to proximity

Reduced waiting time

Overall reduce of throughput time and lead time in manufacturing organizations (Irani, 1999).

Furthermore, a study of 30 companies who have adopted cellular manufacturing (Irani, 1999), has shown that the most motivating reason for changing the shop floor to cellular manufacturing was reduction of the throughput time, chosen by 83% of the respondents. Such characteristic of the cellular manufacturing is aligned with the main principle of QRM, which is why this is one of the building blocks of this strategy. However, it is not the only element.

3.2.Managing the cells: Material control in QRM

POLCA system is a material control system designed in accordance with the building principles of QRM, with objective to guide materials through production systems that have cellular shop floor layout (Vandaele, Van Nieuwenhuyse, Claerhout, & Cremmery, 2008). POLCA is designed to accommodate manufacturing environments with high variety and it combines features of both push and pull material control systems. It is designed to overcome the drawbacks of pull systems when applied to custom product environments and of push systems which can create large WIP and long throughput times (Krishnamurthy & Suri, 2009).

An important dimension of the POLCA material control system is the authorization mechanism. This mechanism has double function – to determine the release of an order into the production and to determine the progress of the orders on the shop floor (Riezebos, 2010). The release decision of orders is executed by High-Level MRP (HL/MRP) system that provides high level planning and coordination across the cells, without micromanaging them. The HL/MRP system used the throughput times per cell to back-schedule the start dates for order. The main difference between such system and traditional push mechanism is that the

Release Authorization determines the possible dates when an order may be processed,

(19)

released, the flow on the production is regulated by using release lists. These lists are a planning instrument, aimed to control the progress of the orders (Riezebos, 2010). The release lists typically contain information about authorization date, next cell in the routing, and in some instances may contain additional information. The information is often listed in increasing order (Krishnamurthy & Suri, 2009).

Within the POLCA system, two successive cells in the routing of a product are considered as a loop. Hence, the Overlapping Loops in the abbreviation of the system name (Vandaele, Van Nieuwenhuyse, Claerhout, & Cremmery, 2008). These loops are also known as control loops and are the foundation of the mechanism of POLCA. The design of the control loops is an important issue that should be addressed in POLCA systems (Riezebos, Design of POLCA material control systems, 2010). Each of these loops is adjoined by control cards, known as POLCA cards. Each POLCA card represents a quantum of work that can express either workload (hours) or number of orders (Krishnamurthy & Suri, 2009). Moreover, the POLCA cards serve as a visual signal as they contain information about the originating cell and the destination cell and card identification (Suri, 1998). An example of generic POLCA card is shown in Figure 2, below.

Figure 2. Example of generic POLCA card

The assignment of the POLCA cards can be calculated by using formula that is dependent on the throughput times of the cells in the loop and the number of orders exchanged in the loop over a period of time (Suri, 1998). As such, the POLCA cards limit the amount of orders that can circulate through the given loop and limits the WIP in that loop (Vandaele, Van Nieuwenhuyse, Claerhout, & Cremmery, 2008).

(20)

then attached to the order. This signals that the order can be processed in the first cell in the routing (the Metal Sheet Processing cell in this example).

Figure 3. Example of order flow and usage of POLCA cards©1

Once all work for this order has been completed in this cell, the card indicates availability to process to the next cell. Before continuing, however, another card for the next loop must be available and attached to the order, as it can be seen of the picture (bottom squares marked with K-M). Then the order is being processed in the next cell (the Rolling cell in Figure 2). Upon completion, and before continuing to the third cell in the routing (the Assembly cell), the first card (marked U-K in the picture) is returned to the originating cell (Sheet Metal Processing), and a card (M-E) for the next loop is selected and attached to the order, as it can be seen on Figure 4.

Figure 4. Example of order flow and usage of POLCA cards – continued©2

1

©Image property of Bosch Hinges and Metal

2

(21)

These steps are repeated until all the cells in the order routing have been circulated and the order is completed. The logic behind this mechanism is that an order can be completed only if there is an available capacity downstream, so that the upstream cell can continue processing the order. In case of breakdowns and bottlenecks there will be no return of cards to the upstream cell, despite released order, which will indicate that there is no available capacity to process upcoming orders (Riezebos, 2010). In practice, to avoid excessive waiting, which can influence the throughput time, occurrences such as quality defects, shortage of supply or due date changes can be resolved by using “safety cards”. The safety cards contain identical information as the original POLCA cards, with some visual labeling that indicates the distinction. In case the delay is due to the order, the POLCA card is replaced with a safety card, to allow next orders waiting to be processed without delay, while the issue in question is resolved (Suri, 2003).

As noted above, POLCA is a combination of two material control systems (push and pull) that attempts to make the “best of both worlds”. The push element is represented with the release authorizations and the release lists that ensure that the orders are released at the right moment. The pull element is represented through the visual signaling of the POLCA cards which indicate the capacity availability and ensure that an order can be further processed. Such combination makes this material control system suitable for high-variety, low-volume production environments, where processing and routing variety create challenges for workload balancing (Vandaele, Van Nieuwenhuyse, Claerhout, & Cremmery, 2008).

3.3.Evidence from the company

A common tool for determining whether POLCA is applicable for a specific organization is the POLCA Scan, developed at the University of Groningen (Riezebos, 2010). It is important to distinguish between the suitability of POLCA for an organization (which cannot be assessed by this tool) and applicability and readiness for POLCA. To determine the latter, the Scan is a prompt tool that consists of five stages, that are performed consequently: POLCA Objectives, POLCA Cells, POLCA Lead Times, POLCA Loops, and POLCA Effectiveness. The scanning process starts with identifying the problems that POLCA can address as well as objectives and targets. It continues further into analyzing the main components of QRM, the POLCA cellular structure, the lead times and the control loops. Lastly, it evaluates the effectiveness of the proposed system and suggests additional improvements (Riezebos, 2010).

(22)

not determine achievement of all objectives, due to the early stages of implementation (Riezebos, 2010).

The company has been experiencing steady improvement over the years, with stable lead times, zero inventories, and well versed use of cells and POLCA cards, adopted by the entire company. The successful implementation of the system has compelled the organization to rethink their strategy and consider additional improvements in the lead times. Such decision was further motivated by acquisition of new operations, which operated in market with even more competitive lead times (1,5 – 2 weeks).

3.4.Research Design and Conceptual Model

The discussion above together with the management questions posed in Chapter 2 of this thesis, leads to formulation of the problem statement as:

An investigation of the lead time reduction through assessment of the implementation of the system as an approach for improvement in the POLCA system implemented at Bosch Hinges and Metal.

More specifically, this problem statement leads to the following research question:

How can the implementation of POLCA at Bosch Hinges and Metal be further improved to facilitate reduction of the current lead times?

The analysis in this research can be conducted by attending to the following conceptual model. The starting point will be diagnosing activity during which an audit of the POLCA system will be executed. This audit will be made reflecting on the POLCA Quick Scan . In order to accommodate the situation at the company, where the system has been already implemented and fully operational an adjusted scan, titled “audit” will be performed. As such, the audit will consist of four distinctive steps: Evaluation of Cells, Evaluation of Lead Time, Evaluation of Loops and Overall Assessment of the System Effectiveness. Following, an analysis of the overall system effectiveness will be performed. Based on the results from the POLCA Audit, the last step of the Audit will be crucial in determining the factors that influence the lead time and will direct the areas for improvements. It is expected that the designed audit will reveal potential for further examining specific aspects of the system. Consequently, this will lead to a system redesign that will address the problem areas. This in end, will result in answer to the proposed research question.

(23)

Figure 5. Conceptual Model

3.5.Methodology

The proposed research is structured as a qualitative study with inductive and discovery-oriented approach (Pereira Heath & Tynan, 2010). As such this research will take the standing point of the interpretivist paradigm where social reality is perceived from the view point of the individual perceptions and the goal is to get an “insider’s perspective” of the topic of study. Consequently, this research relied on qualitative methods and techniques that yielded descriptive data. In addition to that, the analysis implemented several methods in parallel (triangulation) to obtain greater confidence of the findings and compensate for the weaknesses of specific methods (Pereira Heath & Tynan, 2010).

Descriptive data was obtained using the following methods:

Literature review Observations In-depth interviews Audit of the system Analysis of historical data

The data was accessed during a six month internship at the company premises, where the above mentioned methods were applied. Additionally, for the analysis of the historical data, output from the company’s monitoring software was used. The data sources used and the methodology applied have enabled collection of very rich data that very accurately reflects the current situation at the organization. Due to the vastness of data, not all could be included in this thesis, but can be made available upon request.

3.6.Thesis outline

(24)

their main issues and challenges. The remainder of the thesis is organized in two more sections. The next chapter elaborates the diagnostic stage, where the analysis of the system was performed. For the purposes of the analysis a particular methodology has been developed. This methodology was labeled as POLCA Audit (based on the principles of the POLCA Scan) and it contains four distinctive steps. For purposes of clarity and consistency all steps have been included into single chapter (Chapter 4).

(25)

A n a ly si s o f cu rr e n t si tu a ti o n – P O L C A A u d it 25

4. Analysis of current situation – POLCA Audit

4.1.Introduction to the POLCA Audit as a diagnostic tool and analysis of

the steps

Practical application of POLCA advises a four phase implementation procedure, consisting of: pre-POLCA assessment, design of the POLCA system, launch of the POLCA implementation and post-implementation evaluation. Given that the company in question has more than three years of experience with implemented POLCA material control system, for the purpose of this research only the fourth phase (post-implementation evaluation) is of interest. The goal of this phase is to track and measure the quantitative and qualitative benefits of the implementation. Assessing the computable benefits of the implementation involves metrics such as lead time of products, throughput of cells, delivery reliability and WIP levels at different points in the system. These metrics can be obtained through measures of delivery performance of cells, POLCA card audit or similar. Furthermore, implementation of POLCA is accompanied by qualitative benefits, observed through productivity improvement, worker satisfaction, customer satisfaction, communication improvement, team work improvements, etc (Suri & Krishnamurthy, 2003). Although there are measures that allow quantifying such qualitative improvements, this research will focus only on the measurement of the quantitative benefits of POLCA implementation.

Building up on the post-implementation evaluation, an audit tool has been designed to gauge the current state of the POLCA system implemented at the company. This audit has been designed reflecting on the POLCA Quick Scan proposed by Riezebos (2010), for designing POLCA systems. For the original application, the POLCA Scan was designed to determine the suitability of POLCA for specific organization or shop-floor and their readiness for implementation. The Scan consists of five stages that are executed in sequence, as following: POLCA Objectives, POLCA Cells, POLCA Lead Times, POLCA Loops, POLCA Effectiveness (Riezebos, 2010). Four stages of the Scan are of particular interest for this research: the cellular structure of the system, the control loops of the cards that circulate, the lead times, which reflect the planning structure of the system and the overall system effectiveness.

(26)

consequent steps that will be performed in the audit and provides an outline of the specific activities that each step will entail.

Figure 6. POLCA Audit Steps (adapted from Riezebos, 2010)

4.2.Step 1 of POLCA Audit – Analysis of the cells

4.2.1. Criteria for evaluating the cells

The concept of cell, as a structural entity on the manufacturing shop floor and building element in cellular manufacturing has been receiving attention by both theoreticians and practitioners. Consequently, there is wide variety of definitions available that capture the essence of a manufacturing cell (Wemmerlӧv & Hyer, 1987; Wemmerlӧv & Hyer, 1989; Suri, 1998; Irani, 1999; Riezebos, 2001; Suri, 2009). To summarize those definitions, a cell can be seen as cluster of different machines and workers dedicated to operating them, located in specifically designed area of the shop floor, that is responsible for producing more than one product and performs more then one operation.

The starting point of the analysis of the cells were the questions “How should a QRM cell be

designed?” and “How should a QRM cell appear?”. Reflecting on the available literature in

cellular manufacturing, several different dimensions were developed that could serve as criteria to evaluate the cells. These dimensions are:

• Resource perspective • Proximity perspective • Operations perspective • Product flow perspective

(27)

as “consisting of a set of resources (worker, equipment of information)”, (Riezebos, 2001), or as “set of dissimilar machines…operated by multi-skilled workers” (Suri, 1998) or even more aggregate as “set of dedicated, collocated, multu-functional resources” (Suri, 2009). Consequently, this criterion will examine the resources that are dedicated to cells, with emphasis on the workers and the machines.

Important aspect in the process of designing (or evaluating) cells is the proximity of the resources that are allocated to the cell. This is a complex dimension that involves issues such as limitations of physical space, size of machines, safety requirements for placement of machines, etc. These issues can be resolved by mathematical models for designing the layout of cell, available in practice (Irani, 1999). Given that the cells at Bosch Hinges and Metal have already been designed, this research will not focus on the above mentioned aspects. Rather, it will evaluate cells to determine whether the machines and workers dedicated to each cell are in close physical range. This niche has been chosen because closeness and collocation of resources are one the most prominent advanatages of applying cellular manufacturing (Wemmerlӧv & Hyer, 1989).

The third dimension evaluated is the operations perspective. This perspective more specifically refers to the following elements from the definitions: “capable of processing

several operations per work order” (Riezebos, 2011) and “complete a sequence of operations for jobs belonging to a Focused Target Market Segment” (Suri, 2009). In other words, cells

perform more than one operation per work order and in most of the cases once the order is finished being processed in the cell, it should not return. This dimension has been designed to evaluate the variety of processes that are contained in the cells. Set of different processes that can be performed into a cell is what distinguishes cellular from functional layout (Irani, 1999).

The last criterion that will be used to evaluate the cells is the product flow perspective. Within this perspective the dominant insight is that cells are designed to be “dedicated to

producing a family of products that need similar operations” (Suri, 1998). Similarly to the

operations perspective, the purpose of this dimension is to determine the focus of the cell with respect to variety of parts, products or product families that are allocated to a cell (Irani, 1999).

4.2.2. Results from cell evaluation

(28)

Table 1. List of QRM cell after the implementation of QRM and POLCA in 2007

No. Cell Name Original cell name Cell Color

1. Sheet metal preparation Uitslagen Orange

2. Piano hinges Piano Blue

3. Assembly and quality control Montage Red

4. Folding Vervolg Brown

5. Rolling Kralen Yellow

6. Heavy hinges Zware persen Purple

7. Finishing operations Borstelen Green

8. Expedition Expeditie Black

9. Auxiliary tools Hulp persen Grey

Given the changes that have occurred since the first implementation of QRM (i.e. the acquisition of additional company for light metal constructions), the situation presented in the table above has changed. At the beginning of this research, there were 13 cells (12+1) identified with the company. The list of the cells observed in July 2011 is shown in the Table 2 below.

Table 2. List of QRM cells as observed in July 2011

No. Cell Name Original cell name Cell Color

1. Sheet metal preparation Uitslagen Orange

2. Piano hinges Piano Blue

3. Assembly and quality control Montage Red 4. Wet processing Natte bewerkingen Pink

5. Folding Vervolg Brown

6. Rolling Kralen Yellow

7. Heavy hinges Zware persen Purple

8. Finishing operations Borstelen Green

9. Welding Lassen Lilac

10. Tube processing Buis verwerkingen Ochre

11. Expedition Expeditie Black

12. Auxiliary tools Hulp persen Grey

13. Buffer Uitslbuffer Orange

(29)

Figure 7. Cellular structure at Bosch Hinges and Metal (July 2011)

It is important to note that during the course of this research the company made additional changes in the cell structure on the shop-floor. However, of interest for this research was the situation that was found in July 2011 and thus is being examined in the audit.

(30)

Table 3. Summary of the results based on cell criteria

No. Cell Name Dedicated resources Proximity Operations Product Flows Workers Machines [Y/N]

1. Sheet metal _preparation 4 11 Y

Laser-cutting, punching, folding, milling, turning Special hinges Continuous hinges Heavy hinges,

Light metal constructions

2. Piano hinges 2 10 Y Folding, rolling,

carving, assembly Piano Hinges

3. Assembly and _{quality control} 2 11 Y Final assembly,

quality control

Special hinges Continuous hinges Heavy hinges,

4. Wet _processing 0 6 Y (partial)

De-burring, degreasing, polishing, drying, oiling Special hinges Continuous hinges Heavy hinges,

5. Folding 1 7 Y Folding, drilling,

carving Special hinges

6. Rolling 3 6 Y Pre-rolling, rolling, bending, carving, drilling Special hinges 7. Heavy hinges 3 12 Y Folding rolling, drilling, grinding, quality control Heavy hinges 8. Finishing _operations 1 6 Y Grinding, brushing, polishing, de-burring Special hinges Continuous hinges Heavy hinges,

9. Welding 2 7 Y

Welding, assembly, inspection

10. Tube _processing 2 18 Y

Sewing, bending, drilling, milling, grinding, de-burring

11. Expedition 1 4 Y Outsourcing

12. Auxiliary tools N/A 7 Y Pressing, drilling,

rolling, cutting

13. Buffer 0 0 N/A

Buffer for jobs from sheet metal processing to rest of the factory

4.2.3. Analysis and discussion of the results

(31)

cell size of 15 machines (Wemmerlӧv & Hyer, 1989). Given such findings, the number of machines per cell conforms to situations observed in industry practices.

The number of workers on the factory shop floor is 21, organized between 12 cells, which results in average of 1,75 workers per cell. Of specific interest are particular cells which have none or only one worker dedicated to them. For example, at the wet processing cell it was observed that there is no formal dedication of worker to it. The handling of the machines in this cell is mostly done by the workers from the assembly cell on ah-hoc basis (when they are not performing jobs in the assembly cell). Furthermore, observations have also revealed that there is no clear communication about the responsibility and ownership of tasks within this cell. Such situation is conflicting with the principal design of QRM cells (Suri, 1998; Suri, 2009). Furthermore it can lead to longer waiting times in the buffer of this cell, increasing the total throughput time of this cell.

Next, the finishing operations cell and the folding cell, as observed, are being operated by only one worker. Such cell structure indicates a potential bottleneck, due to labor capacity limitations. At the folding cell, this occurrence is a skill issue – the operations require more complex set of skills which is possessed only by the worker operating this cell. At the finishing operations cell the situation is slightly different. In case of heavier workload the polishing operations are being outsourced to three different suppliers with whom the company has agreements. This is done on ad-hoc basis, while observing the production, to prevent this cell from behaving as a bottleneck.

Analysis of the proximity perspective revealed that all the machines are located in closeness to each other in demarcated areas that constitute the cells. Furthermore, the cell boundaries are clearly marked with the respective color of the cell, which allows greater visibility on the shop floor.

(32)

Interesting observations have occurred when analyzing the product flow perspective. The analysis revealed that serving a dedicated product flow holds true only for half of cells in the system. Namely, the sheet metal preparation, assembly and quality control, wet processing and finishing cell serve all the product flows produced in the factory. The structure of these cells is such that the machines used cannot be duplicated and added to the cells dedicated to the specific product flow. This situation requires examining additional perspective and looking into another classification of cells. Among the research on cellular manufacturing, there is classification that shows that cells can either be dedicated to fixed product family or dedicated to fixed combination of operations (Riezebos, 2001). In that sense, instead of the product flow perspective, we can conclude that the cells conform to this classification. The re-classification of the cells is presented in the Table 4 below.

Table 4. Re-classification of cells for product flow perspective

Dedicated to fixed product family Dedicated to fixed combination of operations

Piano hinges Sheet metal processing

Rolling Finishing operations

Folding Wet processing

Heavy hinges Assembly

Welding

Tube processing

Lastly, the existence of the buffer located between the sheet metal processing cell and the remainder of the cells points to a specific case that requires aggregated analysis of all the criteria. This buffer was created briefly after the addition of new operations to accommodate the changes in demand that have been placed on the sheet metal preparation cell. It was further justified with the decision to start producing for light metal construction few months prior to moving the machines in the factory. However, when we consider all four dimensions for the evaluating the cells, the buffer, cannot be described as cell. It can be rather seen as an additional order pool in the middle of production. Such view is problematic, as it will disrupt the flow of the jobs on the shop floor that is supposed to be regulated by the POLCA cards.

4.2.4. Conclusions

The addition of the new operations in the production (light metal constructions) has increased the number of cells in the system, compared to the initial POLCA implementation, by 40%. This increase has introduced additional complexity in the system with respect to material handling, capacity and workload balancing. The most prominent change can be observed in terms of capacity – the increase in number of cells has reduced the average number of workers per cell.

(33)

proximity perspective which will influence the material handling and the lead time. Given that reduction of lead time is the priority for the organization, a temporary trade-off can be made with physically separated machines that can belong to same cell, with clear definition of ownership of tasks and activities (Suri, 2009; Suri1998). For a more permanent solution, the layout can be evaluated by using some of the mathematical methods for cell design available (Irani, 1999).

Similarly, the capacity of the folding cell, which is operated by only one worker, can be treated as a resource perspective issue. This can be addressed by merging this cell with the rolling cell. As such it will form a single cell that will be dedicated to special hinges and labeled accordingly. Such recommendation can be further justified by the fact that most of the orders that visit these cells follow the rolling-folding routing sequence. Furthermore, these cells perform similar operations and it will be easier in the future to cross-train workers from the rolling cell to operate the machines that belonged to the folding cell.

The proximity perspective analysis showed that cells conform to the design principles.

The operation perspective examination of the cells revealed an issue with the outsourcing cell. In addition to being currently a cell that performs only one operation (outsourcing), the establishment of it has increased the material handlings with the assembly cell. This has further implications on the control loops with respect to the moment when the POLCA cards are being returned to signal available capacity. A solution to overcome such occurrence is to assign responsibility of the outsourcing operations back to the assembly and quality control cell.

Within the product flow perspective, analysis showed that the current structure of the production and the nature of the operations does not allow complete conformance to the theory. However, taking into consideration classification of cells based on dedication to either product flow or combination of operations shows more fitting dimension to explain the cell formation.

As a separate case, the presence of the buffer-cell after the sheet metal preparation was analyzed. The analysis indicated that this cell serves as an additional order pool in production and it disrupts the flow of materials that should be regulated with the POLCA material control system.

(34)

Table 5. Proposed redesign of cellular structure

No. Cell Name Cell Color

1. Sheet metal preparation Orange

2. Piano hinges Blue

3. Special hinges Yellow

4. Heavy hinges Purple

5. Finishing operations Green 6. Assembly and quality control Red

7. Welding Lilac

8. Tube processing Ochre

Such proposal shows that the original number of cells can be maintained while at the same time accommodating the acquisition of new operations.

4.3.Step 2 of POLCA Audit – Analysis of cell performance

In production, each decision made, whether it is to start processing an order, to give priority to another order or to postpone operations influences the performance of the factory. Among the many available performance measurement tools, four criteria are common in manufacturing environments: resource utilization, inventory levels, lead times and delivery reliability (Land, 2004).

For this research and this step of the audit in particular, the lead time is the criteria that is of further interest. More specifically, this step of the POLCA Audit examines how the lead time has been modeled in the cellular system found in Bosch Hinges and Metal. Such approach was chosen based on one of the founding principles of QRM: understanding the concept of time. What this principle proposes is that attempts to reduce lead time should not focus only on reducing the processing time, but rather on the throughput time of products (Suri, 2009). Such approach allows considering the components of the throughput time and their contribution to the cell performance.

Because of the characteristic of the data available, the analysis performed in this step required slightly different methodology than the one presented in the research design chapter. Therefore, before presenting the findings a section is devoted to the data collection process, to annotate the specific conditions of this stage.

Excelling the game of POLCA cards – faster and better

Excelling the game of POLCA cards

– faster and better

Title Page

Excelling the game of POLCA cards – faster and better

Preface

Abstract

Table of Contents

List of figures

1. Setting the scene

1.1.Brief discussion of current trends in manufacturing

1.2.Quick Response Manufacturing as strategy for addressing the

current challenges in manufacturing

1.3.Topic choice decision and relevance

2. Bosch Hinges and Metal: context and organizational characteristics

2.1.Brief introduction of the company

2.2.Description of main product flows and main processes

2.3.Summary of improvement activities

2.4.Current problem areas and challenges

S

S

u

u

p

p

p

p

l

l

i

i

e

e

r

r

s

s

Summary of improvement activities

Current problem areas and challenges

C

2.5.Management questions

3. Theoretical foundation of the thesis

3.1.The building blocks of QRM: Cellular Manufacturing

3.2.Managing the cells: Material control in QRM

3.3.Evidence from the company

3.4.Research Design and Conceptual Model

3.5.Methodology

3.6.Thesis outline

4. Analysis of current situation – POLCA Audit

4.1.Introduction to the POLCA Audit as a diagnostic tool and analysis of

the steps

4.2.Step 1 of POLCA Audit – Analysis of the cells

4.2.1. Criteria for evaluating the cells

4.2.2. Results from cell evaluation

4.2.3. Analysis and discussion of the results

4.2.4. Conclusions

4.3.Step 2 of POLCA Audit – Analysis of cell performance

4.3.1. Data collection