Knowledge Management within a global automotive company Factors affecting Knowledge Dissemination from quality improvement projects Master of Science Thesis in Quality and Operations Management JOEL LAESTADIUS CARL SJÖLUND Department of Technology Management and Economics Division of Operations Management CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden, 2014 Report No. E2013:122 i Knowledge Management within a global automotive company Factors affecting Knowledge Dissemination from quality improvement projects JOEL LAESTADIUS CARL SJÖLUND © Joel Laestadius & Carl Sjölund, 2014 Report No. E2013: 122 Department of Technology Management and Economics Chalmers University of Technology SE-412 96 Göteborg Sweden Telephone + 46 (0)31-772 1000 Printed by Chalmers Reproservice Göteborg, Sweden 2014 ii iii Abstract Knowledge as a corporate asset has been highlighted in research during the last decades. For Volvo Group Trucks, a Knowledge Management initiative has currently been rolled out in order to improve the reuse of knowledge from previous projects and enhance the product quality. A central part of Knowledge Management is Knowledge Dissemination, i.e. knowledge sharing or knowledge transfer. The Knowledge Dissemination is affected by different factors with different influence on the Knowledge Dissemination success depending on the context. The purpose of this study is to analyze Knowledge Dissemination from product quality improvement projects in a global automotive company with the intent to identify influencing factors on an intra-organizational level. The case company is Volvo Group Trucks. This case study had a qualitative approach where interviews, mainly with concerned key individuals in product quality improvement projects, were conducted. Three types of projects were studied and a total of 23 semi-structured personal interviews were performed. A literature review has also been carried out to aid the analysis. In order to sort and structure identified factors, a research framework for analyzing Knowledge Dissemination has been used. The framework address the knowledge actors, both source and recipient, the context in which the dissemination takes place, the media used to disseminate the knowledge content, and the activity of the actors two-way interaction. The study concluded that the most relevant recipient for knowledge from product quality improvement projects was the R&D organization and that this Knowledge Dissemination can be analyzed in a short-term and long-term perspective. The short-term perspective represents the immediate dissemination of knowledge after identifying the problem root cause and taking the decision to implement a solution. The long-term perspective reflects Knowledge Dissemination issues with taking preventive actions to avoid reoccurrence in the future. In the short-term perspective, Knowledge Dissemination is mainly dominated by means of meetings and personal interaction. Individuals from the recipient organization taking part in the projects enables Knowledge Dissemination to their specific departments and so does the trust between the actors. However, the main problems are factors related to responsibilities, feedback, resources and how the project process supports prevention of reoccurring quality issues. In the long-term perspective a database for storing knowledge at R&D is central for the Knowledge Dissemination. Main inhibitors for Knowledge Dissemination in the long-term perspective are factors related to the usage of the database and how to reuse stored knowledge, as well as management support and resources available for using the database. In order to enhance short-term Knowledge Dissemination, and ensure its success, the recommendations concerns emphasis on a balanced product quality improvement project process, through implementation of KPIs together with a standardized evaluation of entire projects together with a clearly communicated process end and clear responsibilities for knowledge activities. In the long-term perspective, the knowledge database must be given management support and resources for reuse of knowledge. Also, it cannot be ambiguous how knowledge regarding quality issues is stored and reused, therefore measurements and education for using it is recommended. Keywords: Knowledge Management, Knowledge Dissemination, knowledge transfer, knowledge sharing, automotive industry, quality improvement projects, influencing factors iv v Acknowledgements The researchers are truly thankful for the support from all the people that have made this study possible. During the progress we have met a lot of people who have shown an impressive engagement and we would like to send our gratitude to the following persons: First, a great thanks to our supervisors Robert Toll and Dan Paulin, who have inspired us with professionalism and engagement during the entire project. Thank you for your reflections, for believing in us, for challenging us, and for being patient. All Volvo Group Truck employees who have participated in the study. Thanks for your time and your honest contribution to the study. Big thanks also to the entire Quality and Customer Satisfaction Strategy department who have shown interest in our work and treated us as part of the team from the very first day. Thank you Karin Strömberg, our project sponsor, for giving us the opportunity to perform our thesis at Volvo Group Trucks, and thank you Pia Hammarstedt for your support and facilitating the project. Thanks Amer Catic and Ludvig Lindlöf for your valuable input to the study and for sharing your expertise in the Knowledge Management field. Anton Dannetun and Van Do, our opponents during our seminar, your reflections have been of great value for us. We appreciate your effort. And thank you, all other Volvo Group Truck employees, foremost Ivan Ivanovic, Cecilia Pedersén, Charlotta Uddh, Philippe Mauguin, Guillaume Renart, Stefan Andersson and Jenny Erneman, that have shared ideas with us and showed interest in our research. Lastly, we want to thank our wonderful friends and families. You have contributed to our study more than you can realize. vi Table of contents 1 Introduction� 1 1.1 Background� 1 1.2 Problem analysis� 3 1.3 Purpose� 3 1.4 Research questions� 4 1.5 Delimitations� 4 2 Methodology� 5 2.1 Research strategy� 5 2.2 Research design� 5 2.3 Research methodology� 6 2.4 Limitations of methodology� 9 2.5 Research ethics� 10 2.6 Trustworthiness� 11 3 Theory� 13 3.1 Conceptualizing and defining knowledge� 13 3.2 Knowledge dimensions and conversions� 14 3.3 Views on knowledge� 15 3.4 Knowledge Management � 16 3.5 Knowledge Dissemination� 18 3.6 Research framework� 19 4 The studied area� 20 4.1 Types of Quality Journals� 20 4.2 Addressed quality problem� 20 4.3 The Quality Journal process� 21 4.4 Quality Journal team members� 21 4.5 Quality Journal process tool Argus� 23 4.6 Market Ready meeting� 23 4.7 Quality Journal White Books� 24 4.8 The Design Verification Guidelines database� 24 5 Results� 25 5.1 General results� 25 5.2 Quality Journal step 1D to 5D� 26 5.3 Quality Journal step 6D� 27 5.4 Quality Journal step 7D to 8D� 29 5.5 Tools used for Knowledge Dissemination� 31 5.6 After the Quality Journal� 34 6 Analysis� 36 6.1 Knowledge creation� 36 6.2 Short-term Knowledge Dissemination and capture� 37 6.3 Long-term Knowledge Dissemination and reuse� 43 vii 7 Discussion� 49 7.1 Successful Knowledge Dissemination� 49 7.2 Design Verification Guidelines database effectiveness� 49 7.3 Technical knowledge and process technical knowledge� 50 7.4 Knowledge Management strategy� 50 7.5 Knowledge Dissemination to other recipients� 50 7.6 Push or pull knowledge� 50 7.7 Methodology� 51 7.8 Trustworthiness� 51 8 Conclusions� 53 8.1 Recipients of disseminated knowledge� 53 8.2 Factors affecting the Knowledge Dissemination� 53 8.3 Contributions to the research field� 56 9 Recommendations � 57 9.1 Knowledge Dissemination in the short-term perspective� 57 9.2 Knowledge Dissemination in the long-term perspective� 57 9.3 Actions to take� 58 9.4 Perform a complementary study� 58 10 Further research� 59 10.1 Further investigate knowledge reuse� 59 10.2 Quantify the impact of the identified factors� 59 10.3 Identify the correlation between factors� 59 10.4 Investigate the time lag factor� 60 Appendix� 65 Appendix 1 - Project plan� 66 Appendix 2 - Interview guide� 67 viii DVG GCR GTO GTT KD KM KPI PMQJ PQL QCS QJ QJWB R&D VGT Design Verification Guideline Global Component Responsible Volvo Group Trucks Operations Volvo Group Trucks Technology Knowledge Dissemination Knowledge Management Key Performance Indicator Project Manager Quality Journal Project Quality Leader Quality and Customer Satisfaction Quality Journal Quality Journal White Book Research and Development Volvo Group Trucks List of abbreviations 1 1 Introduction In this part, a background description of the studied case company is introduced. Furthermore, there is an analysis of the overall problem, the study’s purpose and research questions are stated and so are the delimitations of the study. 1.1 Background This section contains a description of Knowledge Management (KM) applicability for companies. The case company is described generally and in detail regarding the focus of this study. 1.1.1 Knowledge Management in companies Since the 1990’s knowledge has been recognized as a corporate asset in the global economy (Davenport & Prusak 2000, Hansen et al. 1999). The conclusion is that knowledge must be managed as carefully as any other corporate asset. Burton-Jones (2003) argue that knowledge is a firm’s biggest asset and others argue that knowledge will shape the future society (Drucker 2001, OECD 1996). According to Nonaka and Takeuchi (1995, p.74): The most critical element of corporate strategy is to conceptualize a vision about what kind of knowledge should be developed and to operationalize it into a management system for implementation. In contrast to tangible corporate assets, knowledge assets increase as they are being used which gives the usage of knowledge a sustainable advantage (Davenport & Prusak, 2000). Another advantage of knowledge is that knowledge is hard for competitors to replicate. Unfortunately, it is also hard for companies to replicate knowledge internally (Szulanski, 1996). Consequently, in order to benefit from knowledge it has to be managed effectively. 1.1.2 Case company Volvo Group Trucks (VGT) is owned by Volvo Group and is a parent organization to Volvo Group Trucks Technology (GTT), Volvo Group Sales & Marketing and Volvo Group Trucks Operations (GTO). Most of Volvo Group’s technical development takes place at GTT and it is the research center for the entire Volvo Group. GTT is a global organization, with approximately 10.000 employees, engineering products and services on the automotive market for commercial vehicles (Volvo Group 2013a). The operations of GTT involve a set of activities, from planning and developing, to deliver complete trucks and also support the aftermarket (Volvo Group 2013b). GTT’s responsibilities are divided into seven areas, Advanced Technology and Research, Complete Vehicle, Powertrain Engineering, Product Planning, Project and Range Management, Purchasing and Vehicle Engineering (Volvo Group 2013b). Volvo Group as well as VGT and GTT have three core values: Quality, Safety and Environmental care. These core values are the areas where Volvo Group wants to be world leaders. The area of 2 quality emphasizes a customer centric view towards reliable goods and services and VGT aims to be the best in customer satisfaction. (Volvo Group 2013c) 1.1.3 The RnD30 improvement project at Volvo Group Trucks Technology In order to sustain a competitive advantage GTT is running an extensive project called RnD30. This project aims to reduce costs and lead times in the Research and Development (R&D) processes by eliminate non-value adding activities. As part of this project, a KM initiative is incorporated. The KM initiative is focused on lessons learned and using knowledge from previous experiences. It is in this area this study have focused, see Figure 1. Measures have already been taken in order to implement KM strategies and tools, many of them at R&D. Figure 1 RnD30 Group Trucks Technology Volvo Group Trucks Knowledge Management Prevent Reoccurrence Master Thesis The study related to the activities at VGT. 1.1.4 Costs of poor quality Quality is an important contributor to an organization’s profit, prosperity and success (Bergman & Klefsjö 2010), see Figure 2. As a core value, quality is highly important for VGT. Lacking quality will generate costs that account for internal and/or external failure costs (Bergman & Klefsjö 2010). An example of internal failure costs is costs for scrap or rework, whereas external failure costs would be when a defective product reaches a customer, causes dissatisfaction, bad will and even warranty claims (Bergman & Klefsjö 2010). Therefore, external failures are serious matters for VGT. If a defective product is causing a significant problem for a customer, VGT has to respond and a quality improvement project is initiated. Figure 2 Cost of poor quality Internal failure costs External failure costs Costs of poor quality. (Bergman & Klefsjö 2010, p.68) 3 1.1.5 Quality Journals Quality Journal (QJ) is VGT’s name for a prioritized product or component quality improvement project initiated due to feedback from the market. An eight-step problem solving process, in this report referred to as the QJ process, defines the execution of a QJ. The QJ process is generic for VGT but there are slight differences among tools and activities in the process between different parts of the organization, e.g. between Powertrain Engineering and Vehicle Engineering. The aim is to solve the issue reported from the market as quick as possible in order to prevent other customers to be affected. The QJs are considered processes for stopping the effects of quality problems and thereafter identifying the problem root cause and finding a permanent solution for eliminating it. Quality and Customer Satisfaction department (QCS) collects market input, initiates cross- functional QJ team and leads the QJ along the QJ process. There are three different types of QJs that can be initiated. In chapter 4 a detailed description will be given to QJ types, the QJ process and the QJ setting. 1.2 Problem analysis A brief problem analysis is provided in order to present the problem area at VGT that gives shape to the purpose of the study and explains the study’s research questions. 1.2.1 Reoccurring quality issues QJs are reactions to quality issues that have been experienced by customers. The problems have different impact on the market, and some are more crucial than others. The issues have to be prioritized and addressed. Solving the problems and helping the customers is important. Another important part is to avoid reoccurrence of the quality issues. Despite the success of developing a solution from a number of QJ projects, VGT has experienced that some quality issues have reoccurred. This fact has made QCS interested in the KM initiative, as means for understanding how QJ knowledge can be managed in order to prevent having the same problem more than once. There is an overall target to minimize quality issues experienced by the customers, but also to reuse the knowledge obtained from a QJ as means for improving product quality. VGT have acknowledged the need to better understand how knowledge from QJs is utilized. 1.2.2 Knowledge Dissemination Knowledge Dissemination (KD) is “a collective term encompassing both knowledge transfer and knowledge sharing” (Paulin 2013, p.2). KD is a crucial knowledge activity for VGT in order to understand the prerequisites for managing knowledge obtained from QJs. Since the problem solving process for QJs are generic within VGT it is of great value to grasp and understand what influence effective KD from QJs. Knowing this could give a direction in how knowledge can be managed more effective, with the ultimate goal of enhanced product quality. 1.3 Purpose The purpose of this study is to analyze Knowledge Dissemination from product quality improvement projects in a global automotive company on an intra-organizational level. The goal is to identify factors that have an impact on the Knowledge Dissemination. 4 1.4 Research questions One main research question has been developed to set the scope of this study. The main research question reflects the purpose and is examined at VGT, representing a global company in the automotive industry. The formulation of the main research question: • What factors affects the Knowledge Dissemination from market initiated product quality improvement projects at Volvo Group Trucks? In order to understand the KD and identify the factors, the participating actors are crucial to identify. Assuming that the product quality improvement project team obtains knowledge about the quality issue they solved and has an intention to disseminate it, the appropriate recipients of knowledge have to be known. Finding the right recipients in a big organization might be complex and consequently worth investigating. Once the knowledge recipients are identified, it is possible to investigate the KD, and identify factors that hinder and facilitate the dissemination. Identifying theses factors are important in order to understand and be able to improve the KD. Based on this reasoning, the main research question leads to two underlying research questions. They will aid in answering the main research question. The two underlying research questions are: • RQ1: Who are the recipients of the knowledge disseminated from market initiated product quality improvement projects at Volvo Group Trucks? • RQ2: What factors affect the Knowledge Dissemination between market initiated product quality improvement projects and identified recipients at Volvo Group Trucks? 1.5 Delimitations VGT is a company with presence around the world but this study is limited to QJs only run by QCS at GTT on the Gothenburg site. All data collection was also performed at the Gothenburg site. Furthermore, possible KD to suppliers from QJs is not included in the scope of this study. These delimitations were made to ease the data collection. The study is focused on identifying factors affecting the KD and limited effort is put into ranking or benchmarking the factors against each other. Moreover, the factors impact on KD is studied and not the factors influence on each other. Also limited effort was put into analyzing the difference in factors or their influence between the three different types of QJs. These delimitations were set because of the limited time available for the study. Identifying and prioritize the factors would require more data, both qualitative and quantitative. Furthermore this study neither assess the solutions developed by the QJs nor if the correct problem root cause was found. The quality improvement processes are sometimes different between the various VGT Brands e.g. Volvo and Renault. Therefore, only processes linked the Volvo truck brand is included in the study. Moreover, only projects aimed at quality issues regarding Powertrain Engineering are investigated. This delimitation was made since the process used for QJs aimed at Powertrain Engineering are expected to be implemented for QJs also aimed at Vehicle Engineering. The process used towards Powertrain Engineering is also considered more mature, according to GTT. 5 2 Methodology In this chapter considerations regarding the researchers’ scientific perspective and the design of the study are explained. Insights are given to the study’s ethical considerations and how the quality of this research is assessed. Furthermore, used methods and tools are presented together with the study’s line of action. 2.1 Research strategy The research strategy of this study has been based on a hermeneutic approach. As a qualitative study the emphasis was on words and understanding human behavior rather than testing a hypothesis with quantitative measures (Bryman & Bell 2011). Limited effort was therefore put into gathering quantitative data. 2.1.1 From data to theories Regarding the role of theory an inductive approach was primarily chosen. This implies that data first are gathered and thereafter theories are developed. This approach is in line with research questions of an understanding nature. However, there are some influences of deductive elements since the research was based on KM theory and further data collection was carried out to confirm formulated theories (Bryman & Bell 2011). 2.1.2 The difference between studying people and objects An epistemological position of interpretivism has been used regarding what is considered acceptable knowledge. Bryman and Bell (2011) writes that interpretivism implies that there is a difference between people and objects of the natural sciences, arguing that the subjective meaning of social action have to be considered. If assumed there are only subjective truths about social interactions measures to improve objectivity can be taken. 2.1.3 The social context is affected by the people in it The chosen ontological orientation was based on constructionism, which asserts that social actors continually accomplish and influence social phenomena and their meanings. According to Bryman and Bell (2011) this means that since the people acting within a social context actually decide what is truth at every point in time, it is impossible to draw definitive conclusions about social reality. 2.2 Research design This study was performed as a one-time single organization case study at VGT between September 2013 and January 2014. The case study research design is acknowledge as common and widely used in business research (Eisenhardt & Graebner 2007). The choice of research design was based on its alignment with the overall purpose of the study. Performing a case study and keeping the research questions narrow made it possible to deeply explore the studied area. 6 2.2.1 Level of analysis Within this study the primary level of the analysis have been on a group level, i.e. intra-organizational level (Bryman & Bell 2011). This means the study’s focus have not been on individuals but on different types of groupings e.g. the QJ team and the involved departments. 2.2.2 Data collection and analysis based on Grounded theory The relationship between data and analysis in this study have been inspired by the Grounded theory. Data was hence used to develop new theory, and data collection and analysis was performed simultaneously and iteratively. The collected data was coded into concepts and labeled for the analysis. (Bryman & Bell 2011) 2.3 Research methodology In this section the study’s line of action will be explained in detail. The study has been planned based on Bryman and Bell’s (2011) work phases for qualitative research. The involved phases are shown in Figure 3. Figure 3 M 1. General research questions 2. Selecting relevant sites and subjects 3. Collection of relevant data 4. Interpretation of data 5. Conceptual and theoretical work 6 Writing up findings/conclusions 5b. Collection of further data 5a. Tighter specification of the research question(s) ain phases of qualitative research. (Bryman & Bell 2011, p.390) This study was started with a pre-study aimed at tighten the specification of the initial research question and break it down into manageable sub questions. In the pre-study step 1 to 5a were performed. The main study was started after the initial loop, with the specific research questions more defined, and continued through step 1 to 6. A representation of the general time plan and research phases are shown in Appendix 1. 2.3.1 Pre-study After agreeing upon the main objectives for the study together with supervisors at GTT and Chalmers, the pre-study begun. The purpose of the pre-study was to increase the researchers KM understanding, familiarize with VGT and develop the research questions. A literature study on KM theory was started and continued throughout the whole pre-study in parallel with the other tasks. The choice of literature included books, articles and reports. Literature was mainly searched for at the Chalmers library, including its databases, and Google Scholar. Other literature was found via reference lists in read literature and by personal recommendations. Commonly used search words, also in combination with each other, were: Knowledge Management, knowledge transfer, knowledge sharing, product development, automotive industry, quality, project, barriers, factors. To understand the VGT organization structure and the main processes the supervisor at GTT was 7 consulted, company presentation material was read and the researchers participated in online learning courses for new employees. Together with the supervisor at GTT, meetings were held with employees from QCS, R&D and other departments to furthermore increase the understanding of the processes in focus. These meetings also gave new insights regarding the main idea of the study. Throughout the pre-study the researchers also discussed research strategy, research methodology, research questions and the main idea of the study with GTT and Chalmers supervisors, a KM specialist at GTT and a PhD at Chalmers with knowledge within the academic field. 2.3.2 Main study With the researchers improved KM understanding, knowledge about VGT, information about problems, possible value for VGT and input on general feasibility it was possible to specify the research questions and initialize the main-study. This section will in detail describe the process of the main study and the methods used. The main study was divided into two phases. The first phase included planning the data collection, deciding on which reference projects to study and finding relevant interviewees. In the second phase the data collection was conducted, the data was analyzed and the analysis was verified. 2.3.2.1 Choice of reference project To conduct the study, one reference project was picked from each of the three different types of QJs at VGT, see chapter 4 for an explanation of the different project types. The reason for choosing a reference project from each type of QJ project was to get a more complete understanding of QJ projects and increase the likelihood for general conclusions regarding the QJ process. Looking into three reference projects, the time was sufficient to gather rich data and study these projects thoroughly. Also by focusing on three specific projects it was possible to compare data from different employees involved in the same project and thereby get a more complete picture of each project. The three studied QJs were labeled QJA, QJB, QJC based on their type. A requirement for the chosen reference QJs was that they had concluded on the QJ created knowledge and disseminated most of it. Therefore, all studied projects have at least reached the 6D step of the QJ process, see chapter 4 for an explanation of the project process. It was also ensured, that the reference projects had reached the 6D step during 2013 to increase the likelihood of interviewees being available at the Gothenburg site and having their project fresh in mind. The choice to look at projects where KD had occurred also made it possible to interview sources and recipients regarding the same KD. Three QJs fulfilling these requirements, as well as being considered as representative by QCS managers, were selected for the study. 2.3.2.2 Identifying interviewees When the choice of reference QJs was made, suitable interviewees were identified and selected. The subjects for interviews were chosen considering their role, responsibility and involvement in the project. The different roles involved are explained in chapter 4. A condition was also that the individuals were present during the KD. Interviews were conducted with individuals on both the source and recipient end of the KD. In the QJA, six persons in total were interviewed. Two of them represented QCS, the QJA-holder and a PMQJ. Three interviewees were working at R&D, the engineering leader and two design engineers. There was also one interviewee from GTO. Eight interviews were conducted regarding the QJB. From QCS, two PMQJs were interviewed. Three employees from R&D were interviewed: an R&D auditor, a GCR, and a design engineer. From Aftermarket, a Product Quality Manager took part in the interview. Furthermore, a Purchasing employee was interviewed as well as a safety investigator. In the QJC, interviews were held with two individuals from QCS, a PMQJ and a PQL. 8 From R&D, three interviews were held with two different design engineers and one engineering leader. There was also one interview conducted with a Purchasing employee. To increase the likelihood of developing universally conforming recommendations, data was also gathered from individuals involved in QJs other than the reference ones to enable a comparison of the results and validate the findings. In total three interviews of this kind were conducted, all of interviews were conducted with PQLs from QCS. 2.3.2.3 Data collection and interview methodology Interviews are a well-established method for qualitative data acquisition. The access to subjects was assumed due to the study being anchored at a high level within VGT. Using qualitative interviews, the goal was to get deep understanding of subjects’ experiences and subjective opinions. Personal interviews have been chosen to comfort subjects in voicing their thoughts, securing rich and trustworthy data. An additional advantage with personal interviews is the possibility of also gathering data from social cues (Opdenakker 2006). Social cues could be expressed in e.g. the body language and the voice of the interviewee and it can be valuable input if the interviewee is seen as a subject (Opdenakker 2006). When performing personal interviews it is also essential to ask probing questions to go beneath surface appearances since the subject probably interpret things differently than the interviewer first thinks (Bryman & Bell 2011). However, the risks associated with personal interviews are that the interviewer could influence the interviewee (Opdenakker 2006). This was mitigated through acknowledging the risk and using a written interview guide during all interviews. The interview guide was also used in order to explain e.g. the study’s purpose in an equal way for all interviewees before the interviews begun, see Appendix 2. The interview questions was structured with respect to the research framework, presented in chapter 3. Before the main data collection begun, a test interview was performed to evaluate the interview structure and the questions. Adjustments were done to better tailor the questions to the interviewees. The data collection was started by interviewing subjects from the knowledge source, i.e. the QJ. The researchers’ intention was to let QJ team individuals state appropriate knowledge recipients and then assess the KD from the recipient’s perspective. The decision to start with the source was made since the knowledge is created during the QJ when the recipient might not be known yet. Furthermore, since the QCS department initiates the QJs it was easy to get in contact with QJ team members. All interviews were semi-structured and followed the interview guide. Semi-structured interviews were chosen because it gives an opportunity to deviate from the structured questions if an interesting topic is brought up. Therefore, the questions were seldom asked in the exact same order and questions were sometimes left out to make room for investigating topics of particular interest. In total, 23 personal interviews were carried out and data was collected in new interviews until the researchers perceived a convergence in the data. 21 of them were conducted face-to-face, one was conducted via telephone and one via videoconference. The reason for having two non face-to-face interviews was due to practical issues as the interviewees were not present at the Gothenburg site at the time for the interview. Two of the interviews were ended premature since the obtained data was not conforming of the expected extent and quality. The interviews had an average duration of 1.5 hours. During the interview session one researcher took notes and the other conducted the actual interview. All interviews were recorded and transcribed to ensure to capture as much detailed data as possible. According to Bryman and Bell (2011) writing transcripts takes approximately six longer than the actual interview. Apart from the interviews, the researchers have spent a considerable time in the building where the QCS interviewees do their daily work. This implies that informal conversations with interviewees 9 have taken place, discussions have been overheard and other observations, with or without relevance, have been made. Also, the researchers attended two main QJ meetings in order to observe how they were practically conducted. These methods have increased the researchers understanding of the studied context and probably to some extent influenced the data collected. However, none of the methods have generated any documentation used as data, although it is likely that the researchers’ presence could have an impact on the analysis and the conclusions made. 2.3.2.4 Analysis As described earlier, the data collection and analysis was done in tandem in accordance with the Grounded theory. Data was collected through interviews and after each interview, time was used to reflect on the interview by revisiting notes and the semi finished transcript. During these reflective sessions concepts for potentially significant factors and theory was compiled and labeled. In the beginning of the data collection these concepts was vague but as the interviews continued concepts was refined and confirmed by more data. New concepts were also continuously added as the study progressed. Both potential and verified concepts were kept in a continuously updated document. A similar coding process occurred after all interviews had been conducted and all transcripts had been finalized. The transcripts was printed and cut into pieces while being coded into concepts. The concepts were developed by gathering corresponding data together and label the concepts with post-IT notes. These concepts were compared to the already made concepts and higher-level categories could be developed. In the end of the analysis, a number of defined categories had emerged with concepts and data supporting them. Paulin’s (2013) research model has been used as the research framework of this study in order to aid the understanding of KD, design interviews and structure the analysis. The framework was chosen since it was considered comprehensive and concrete. Also, it had been developed from other similar models of KD and it has been used for analyzing KD in a similar company setting (product realizing multinational corporations) as this study. Even though the researchers’ supervisor from Chalmers has developed the framework, the influence of this choice was considered to have no significant impact on the study. The framework was used to structure the analysis and give perspective on how to label factors. Regardless of the choice of research framework, the collected data would probably reveal the same factors, but from another perspective. The research framework is explained in chapter 3. 2.3.2.5 Validation of the analysis In order to validate the analysis two unstructured interviews with personnel possessing general insight of the studied processes and groups were conducted. This was an additional effort made to confirm the credibility since the main study’s interviews had already continued until the collected data converged. 2.4 Limitations of methodology As a qualitative one-time case study focused on a social context the findings cannot be truly objective. It is impossible to study the social world and decide what is the absolute truth since the social world is constantly in change. Also, the researchers own values will influence the interpretation of the social scene to some extent. (Bryman & Bell 2011) Furthermore, the interviewers’ interview skill will affect the performance of interviews as a method. For example, if subjects feel uncomfortable with the situation the data they contribute with might be sparse and unreliable. Another drawback when using interviews as the data collection method 10 is that only a limited number of interviews are feasible since they require a lot of resources and a lot of time from both researchers and subjects. The risk is to only interpret the social scene through the eyes of a few of the individuals in that scene. The sample of interview subjects is therefore important. Efforts have been made to ensure that the interviewed sample spanned over all different roles participating in the studied QJs. 2.5 Research ethics In business research there are four main issues for considering ethics: harm to participants, lack of informed consent, invasion of privacy and deception (Diener & Crandall 1978 in Bryman & Bell 2011). In this study all issues were addressed and treated as important. This section describes how the study took the four ethical considerations into account. 2.5.1 Harm to participants This study can guarantee that no physical harm affected the participants or anybody else. However, the biggest risk of harm to participants in this study was psychological. One source of harm would be if the participants got connected to critical statements about VGT, their job, managers or colleagues, leading to negative consequences for them. To address the risk of connecting critical statements to the participants, the highest possible extent of anonymity was strived for. However, it cannot be guaranteed that no one will know who participated in the research. It is clearly difficult to interview subjects at their work, ensuring that no one else knows they are taking part in the study. On the other hand, the participants was guaranteed that no statement would be possible be trace to a single individual. Documented answers were separated from the participants name or any other information that could reveal their identity. In order to do this, the documentation was identified with a number that had nothing to do with the participant instead of the participant’s name. Recorded data was also held confidential to all but the researchers. Another source of psychological harm would be if the questions recall bad memories or experiences, or if the questions make the participant feel inadequate or in any sense uneasy. Whether the questions evokes an uneasy feeling or recalls bad memories is a subjective issue that can’t be addressed in a standardized way. To avoid it to the largest possible extent the interview questions were revised after a test interview with a non-participating individual. The participants were also told in advance that they could stop the interview if they wanted, without explanation. They could also skip questions, without explaining their reason. Also, since the study regards quality issues, which might be a sensitive topic for some participants, the alternatives of stopping the interview or skipping questions were necessary to offer. It was also emphasized that the study was not about finding scapegoats and individual mistakes, but rather understand the circumstances. However, no interview was stopped by the interviewees nor was any questions skipped in the study. 2.5.2 Lack of informed consent The researchers should thoroughly explain the study to potential participants in order to get their consent of participation, according to Bryman and Bell (2011). In this study the participants were informed about: the purpose and outline of the study, who was the supervisor, the researchers academic background, how the collected data would be used in order to meet the purpose and answer the research questions. They were also informed about their right to stop the interview and skip questions, and that collected data would be treated anonymously and nothing they said would be traceable back to themselves. The estimated time needed was stated in the interview invitation but it was repeated in the introduction to the interview. All this information was provided before the interview started. Since it is hard to judge how much information is sufficient for the participants to make a sound decision of consent (Bryman & Bell 2011), it was also time available 11 for the participants before and after the interview to ask questions. The recorder was started after the informative introduction and only after approval by the participants. It was communicated when the recorder was being used. 2.5.3 Invasion of privacy Even when the participants have a good understanding of the research and had agreed to take part in the study, it is still important to protect their privacy. Participants were reminded that they at any time could revoke their participation and choose to not answer further questions. They were also aware of how data was being documented and how it would be used in the research. All participants also had the opportunity to get back to the researchers if they wanted to comment upon something or add something in regard to their interview afterwards. The interview questions regarded KD, the QJ process and the research framework exclusively. No other topic, e.g. personal matters, was addressed during the interviews. 2.5.4 Deception The purpose of the study should be clear and honest and performed with integrity, representing the research as something it is not would be deception (Bryman & Bell, 2003). The researchers’ role was therefore clearly stated and agreed upon together with VGT at the start of the study. The study has also emphasized its true purpose in all means of data collection. The participants have been informed about how data will be used and they were also invited to take part of the outcome and final findings of the study. 2.6 Trustworthiness Trustworthiness has been chosen as the criteria to assess the quality of this qualitative study. This is a common way to assess qualitative studies, which acknowledge that there are no absolute truths about the social world (Guba & Lincoln 1994 in Bryman & Bell 2011). Furthermore, Dubois and Gadde (2013) state that criteria originally developed for quantitative research (construct validity, internal validity, external validity and reliability) are not suitable for evaluating deep-probing qualitative studies. The four aspects of trustworthiness are (Bryman & Bell 2011, p.395): • Credibility - How believable are the findings? • Transferability - Do the findings apply to other contexts? • Dependability - Are the findings likely to apply at other times? • Confirmability - Has the investigator allowed his or her values to intrude to a high degree? There are some common critique against qualitative research methods stating that they are too subjective, difficult to replicate, hard to generalize and lacks transparency (Bryman & Bell 2011). This critique is relevant also in regard to this study. Beneath the aspects of trustworthiness are described and performed counter actions are explained. 2.6.1 Credibility Credibility is a measure on how well the study’s findings correlate with reality. To ensure credibility the researchers tried to follow good practice. For example, if feeling unclear of the meaning of an interview statement the interviewee was asked to explain the statement again. Also both researchers attended the majority of the interviews so the interpretation of the data could be discussed. Triangulation made sure that the study’s findings were crosschecked by a different source of data by performing two additional interviews with new interviewees. The study’s findings were also presented during a presentation for personnel working within the studied context and for participants of the study followed by discussions and questions. It is however questionable if the 12 studied subjects can be used to validate the research findings without bias. The researchers were aware of this issue at the time of the presentation and could act accordingly. 2.6.2 Transferability Transferability is a about the study’s findings generalizability to other contexts. Since this is a one-organizational case study at VGT and the study is based on a fairly small amount of personal interviews the findings will likely be limited to the studied context. The transferability to other contexts is therefore questionable for the case study, which is a common issue (Bryman & Bell 2011). To maximize the transferability of the study, the studied context was described in detail to help others assess if the findings do apply to their context. 2.6.3 Dependability Whether a study is possible to replicate at another time and still generate the same findings is the main aspect of dependability. Due to its subjectiveness this can be hard for a qualitative study. Naturally it will be difficult for someone to perform this study in the same way, with the same subjects and at the same company. Subjects are anonymous and of a fairly limited number meaning a single changed individual can have a rather large impact on the findings. Also the researchers themselves are not totally objective in their interpretations. To improve the dependability of this study records was written during the study of the reasoning and the process of all conducted phases. The researchers have also tried to give a rich description of the study’s process in this report. 2.6.4 Confirmability Confirmability regards if the researchers have been as objective as possible, even if complete objectivity is impossible, when conducting the study and when formulating the findings. To ensure that the researchers own values affected the study as little as possible almost all data collection was performed by both researchers present, turns was taken interviewing and taking notes. Also, full transcriptions were written. The researchers in collaboration also carried out all data analysis, in order to mitigate the chance for only interpreting a piece data from one perspective. Throughout the study the researchers tried to maintain as objective as possible, e.g. being aware of possible attempts to influence the study by VGT individuals engaged in internal politics or by VGT individuals possibly with a lot at stake depending on the study’s findings. The researchers did not embraced thoughts or suggestions from others regarding the study’s findings without having clear support in the collected data. 13 3 Theory The theoretical background used in the study is presented in this part. The concept of knowledge is introduced and defined. General insight into KM considerations and the research area are given together with the researchers’ view on KM and KD. Lastly, the research framework used in the study is presented. 3.1 Conceptualizing and defining knowledge In order to understand the meaning and complexity of knowledge, one can compare it with the related notions, data and information, which are more familiar and more comprehensive terms. This way to conceptualize knowledge is common (Paulin 2013). Data, information and knowledge are not interchangeable concepts but by acknowledging the difference in meaning of them, they can be related and their transformation into each other can be understood (Davenport & Prusak 2000). The following definitions are based upon the work of Davenport and Prusak (2000). Data “is a set of discrete, objective facts about events” (Davenport & Prusak 2000, p.2). It has no inherent meaning and provides neither judgment nor interpretation. However, data can be processed and constitute the raw material of decision-making and it is possible to store. Although data is possible to transform into information, there is a risk of gathering too much. Information is described as a message with the intention to “shape the receiver”. As a contrast to data, information has meaning, in other words, data is transformed into information by adding meaning. This transformation happens when the source provides a context to data, categorize the data, calculate it, correct it and/or condense it. The intention of informing is to have an effect on the receiver’s judgment and behavior. Knowledge can be created from information in a similar way as meaning added to data becomes information. This happens through comparison and connection of information but also by elaborating on consequences of information and having conversations about it. The value of knowledge, compared to data and information, is the closeness to taking action. Knowledge is more complex than the previous concepts and the following definition by Davenport and Prusak (2000, p.5) illustrate the complexity of its meaning: Knowledge is a fluid mix of framed experience, values, contextual information, and expert insight that provides a framework for evaluating and incorporating new experiences and information. It originates and is applied in the minds of the owners. In organizations, it often becomes embedded not only in documents or repositories but also in organizational routines, processes, practices, and norms. 14 Nonaka (1994, p.15) defines knowledge as “justified true belief” and also distinguishes it from information. Nissen (2002) uses Davenport and Prusak’s (1998) conceptualization of knowledge in a “knowledge hierarchy triangle” and illustrate it with the dimensions actionability and abundance, see Figure 4. The triangle shows how the abundance level decreases when the lower hierarchy levels are transforming into higher hierarchy. The dimension of actionability illustrates how the knowledge is more supportive to an action than data and information. Figure 4 Knowledge Information Data Abundance Ac ti on ab ili ty The knowledge hierarchy. (Nissen 2002, p.253) As can be seen in Figure 4, the top of the triangle is missing. Davenport and Prusak (2000), discuss “higher-order concepts”, e.g. wisdom, but choose to exclude those from the conceptualization for practical reasons. This study will not explore higher-order concepts but instead, similarly as Davenport and Prusak (2000) and Nissen (2002), conclude that knowledge is essential in decision- making and for taking actions. 3.2 Knowledge dimensions and conversions Knowledge can generally be categorized into two different dimensions, namely explicit knowledge and tacit knowledge. Nonaka and Takeuchi (1995) states that explicit knowledge is a view of knowledge as something formal that can easily be expressed in words and numbers and shared via written documents and databases. Tacit knowledge is in contrast not easily visible and expressible, e.g. the know-how of a craftsman. Therefore, tacit knowledge is hard to verbalize and communicate to others as well as formalize and share in a systematic manner. Regarding tacit knowledge Polanyi (1966, p.4) states “we can know more than we can tell”. For tacit knowledge to be shared widely within an organization it needs to be converted into words and numbers. However, knowledge expressed this way will not fully capture the original knowledge. Whether all knowledge is either explicit or tacit is a controversy and not all authors agree on a clear division. For example Shin et al. (2001, p.337) states “the boundary between explicit and tacit knowledge, however, is not clear”. Explicit and tacit Knowledge can be converted in four ways, from tacit to explicit, explicit to explicit, explicit to tacit and from tacit to tacit knowledge. These four modes of knowledge conversion are socialization, externalization, combination and internalization, see Figure 5. Knowledge created by socialization is based on shared experience, e.g. via observation or interacting with customers, and can be shared mental models and technical skills e.g. how to perform an epic split or develop ideas for improvement. Externalization is a process in which tacit knowledge becomes explicit in the form of e.g. concepts, hypotheses or metaphors. An example would be generating a product concept by collective reflection. Combination involves combining and categorizing explicit knowledge through media like documents, databases, meetings, etc. to form new knowledge. 15 Middle managers breaking down corporate visions to make them more operationalized or to take product concepts and integrate them into a corporate vision are two examples. Internalization is when knowledge created in from of socialization, externalization and combination is internalized as tacit knowledge in individuals as mental models or technical skills. This is closely related to “learning by doing” and an example would be a team that have learned from previous experiences and changed how they conduct projects. (Nonaka & Takeuchi 1995) Figure 5 Socialization Externalization Tacit knowledge From To Explicit knowledge Internalization Combination Tacit knowledge Explicit knowledge Four modes of knowledge conversion. (Nonaka & Takeuchi 1995, p.62) Organizational knowledge is mainly created in the continuous interaction between tacit and explicit knowledge. This knowledge creation process is made up of the knowledge shifting between the four knowledge conversion modes like the spiral illustrated in Figure 6. Socialization is however, a limited form of knowledge creation since the knowledge needs to be explicit to be easily shared within the organization. So is also the combination process because merely compiling explicit knowledge in e.g. a document does not necessarily extend the organization’s knowledge. (Nonaka & Takeuchi 1995) Figure 6 Socialization Externalization Dialogue Learing by doing Internalization Combination Field (Team) Building Linking Explicit knowledge The knowledge spiral. (Nonaka & Takeuchi 1995, p.71) 3.3 Views on knowledge The views of knowledge in literature are diverse and somewhat ambidextrous. The different views upon knowledge are important to address when KM systems are developed (Alavi & Leidner 2001) and when KM performance is evaluated (Chang Lee et al. 2005). Alavi and Leidner (2001) present different perspectives of knowledge and their implication for KM and KM systems. Chang Lee et al. (2005) use an excerpt of Alavi and Leidner’s (2001) summary, which is shown in Table 1. 16 Table 1 Different perspectives of knowledge. (Chang Lee et al. 2005, p.470) Sveiby (2007) also elaborates on knowledge as an object and presents an opposing view, knowledge as a subjective contextual construct. The former view implies that “knowledge can be packed, stored and retrieved with relative ease” (Paulin 2013, p.5), but this view is criticized by many authors according to Sveiby (2007). Knowledge as a subjective social construct is based on Polanyi’s (1958) argument that “knowledge is constructed in a social context and that it cannot be separated from the individual and the context” (Sveiby 2007 p.1638). This view implies that the context of creating knowledge is where management should make an effort. Regardless the view on knowledge, the distinction between information and knowledge or tacit and explicit knowledge is still central according to Sveiby (2007). 3.4 Knowledge Management Knowledge Management is defined as “efficient handling of information and resources within a commercial organization” according to Oxford Dictionaries (2013). Several authors describes it as locating knowledge within a firm and use it where it is needed, in other words “know what you know” (Davenport & Prusak 2000, O’Dell & Grayson 1998, Burton-Jones 2003). Burton-Jones (2003) states that firms must enhance knowledge as a their biggest corporate asset. In order to do so, companies must be aware of their knowledge assets as well as closing the knowledge gaps. Closing the gaps can be facilitated with the use of knowledge maps, see Figure 7. Figure 7 Tacit Explicit Knowledge assets Knowledge gaps What we know we know What we don’t know we know What we know we don’t know What we don’t know we don’t know Mapping organizational knowledge. (Burton-Jones 2003, p.144) KM is often described in a set of events or activities. Nissen et al. (2000) call this set of knowledge activities the KM life cycle and made a comparison based on different authors, see Table 2. Implications for KM KM involves enhancing individual's learning and understanding through provision of information Key KM issue is building and managing knowl- edge stocks KM focus is on knowledge flows and the process of creation, sharing, and distributing knowledge KM focus is organized access to and retrieval of content KM is about building core competencies and understanding strategic knowhow Perspectives of knowledge State of mind - Knowledge is the state of knowing and understanding Object - Knowledge is an object to be stored and manipulated Process - Knowledge is a process of applying expertise Access to information - Knowledge is a condition of access to information Capability - Knowledge is the poten- tial to influence action 17 Similarly, Alavi and Leidner (2001) describe KM as a process of activities and argue that the labels of activities might differ as well as their number, but the underlying concept described is about the same. They also argue that there is a minimum of four basic activities namely: creation, storage/ retrieval, transfer and application. As have been mentioned previously, this study will mainly focus on the transfer/sharing (dissemination) phase within the KM research field. Table 2 KM life cycle models. (Nissen et al. 2000 p.30) In Nonaka (1994), the organizational knowledge creating process is described as an ontological dimension of knowledge that indicates the amplitude of knowledge created by individuals in the organization. The amplitude reaches through four levels of interaction: individual, group, organizational and inter-organizational, see Figure 8. Choo and Neto (2010) relate the interaction levels with different knowledge processes in their framework for enabling conditions for KM. The interaction level in focus in this study will be on group level, i.e. intra-organizational, team or department level. Figure 8 GroupIndividual Inter-organizationalOrganizational Interaction levels. Furthermore, KM practices can be divided into two different strategies, codification and personalization (Hansen et al. 1999). The codification strategy emphasis that knowledge is stored centrally, e.g. in a database, and made available to all employees. The strategy of personalization instead highlights knowledge to be communicated between individuals. Hansen et al. (1999) argues that one strategy should preferably dominate the other in a company. The “80-20 rule” is mentioned as good properties. One aspect central in the choice of strategy is how to use IT- systems. For instance, the IT system could be used to store explicit knowledge, or it could facilitate mapping the organizational knowledge. However, possessing technology does not per se imply the use of it or effective KM (Davenport & Prusak 2000) but it can be regarded as a facilitator (Goh 2002) or even an enabler for KM (O’Dell & Grayson 1998). Phase 2 Organize Map/bundle Organize - Organize Phase 1 Capture Create Create Generate Create Model Nissen Despres and Chauvel Gartner Group Davenport & Prusak Amalgamated Phase 3 Formalize Store Capture Codify Formalize Phase 4 Distribute Share/transfer Access Transfer Distribute Phase 5 Apply Reuse Use - Apply Phase 6 - Evolve - - Evolve 18 3.5 Knowledge Dissemination The terms knowledge transfer and knowledge sharing are more commonly used than Knowledge Dissemination, however the distinction between the notions are not clear and some authors use them synonymously (Paulin & Suneson 2012). The difference in ontological and epistemological perspective of knowledge, knowledge as an object or knowledge as a subjective contextual construct, might have an impact on the use of the notions knowledge transfer and knowledge sharing in the literature according to Paulin and Suneson (2012). In order to avoid taking a stand in the epistemological and ontological view upon knowledge and to be able to address a broad knowledge perspective, this study has used the encompassed notion Knowledge Dissemination. Disseminating knowledge with considered success is not the same thing as merely making it available (Davenport & Prusak 2000). Cummings and Teng (2003) use the definition of knowledge internalization for judging the success of KD. A recipient that fully internalize knowledge gets ownership of it, commitment to it and satisfaction with it, and when this happens knowledge can ultimately be used (Cummings & Teng 2003). Davenport and Prusak (2000) has a similar idea, they mean that dissemination of knowledge must involve absorption by recipient in order to even be considered KD. They further argue that KD is not adding value to the organizations if it does not lead to use, change, development etc. These ideas of successful KD correlate to the hierarchy triangle, the capability perspective of knowledge, and the KM processes earlier presented. KD is part of KM theories but it is also an important element in the theories of learning organizations (Garvin 1993, Goh 2002). Learning in an organization is a requirement to continuously improve (Garvin 1993), and continuous improvement is vital in quality strategies (Bergman & Klefsjö 2010). Garvin (1993) also mentions “learning from past experiences” as an element in building a learning organization. This study has however only focused on KD in relation to KM. As for communication between individuals, KD can also be affected by different factors. Within the KM research field, KD has been widely explored. A lot of literature regarding influencing factors in KD has been written in recent years. Some examples are: Davenport and Prusak (2000), Cummings and Teng (2003), Cummings and Teng (2006), Riege (2005), Riege (2007), Søndergaard et al. (2007), Duan et al. (2010) and Wang and Noe (2010). Factors can have a negative impact on KD. In that case Davenport and Prusak (2000) call them “frictions” and argue that they can slow down or prevents KD activities. Other words for these negative factors are hinders, noise, barriers, obstacles and constraints. Factors having a positive influence on KD are for example called enablers and facilitators. Based on Cummings and Teng’s (2003) paper it is possible to view the factors’ influence on KD success as linear or curvilinear. Figure 9 illustrates different types of factors. Figure 9 Amount + linear - linear u-shaped n-shaped Po si ti ve in fl ue nc e Types of factors influencing KD. 19 3.6 Research framework The research framework used in this study is a research model developed by Paulin (2013). Paulin’s (2013) KD model synthesize Shannon and Weaver’s (1949) communications model for machines with Lindkvist’s (2001) linear communications model adapted for intra-organizational communication between R&D projects (also based on the model by Shannon and Weaver (1949)) and Cummings and Teng’s (2003) research framework for key factors affecting KD success on both domestic and international R&D partners. The research framework is based on seven components: Actors (including both Source and Recipient), Content, Media, Activity and Context. Paulin (2013) used the research model to group factors influencing KD. Figure 10 illustrates the framework. Figure 10 T Actor - source Actor - recipient Context Media Activity Content he research framework. (Paulin 2013, p.21) In the research framework the component Source address influencing factors on the source side of KD while Recipient is about the receiving side of the KD. The source of knowledge is predefined as a QJ in this study and the recipient is the department(s) or function(s) defined and addressed by the source. The component Content is about the disseminated knowledge itself, i.e. the message that will result in an action. The component Media is the element in which the KD takes place, for instance via personal interaction or through a database. The Activity component includes the activity of KD between the actors, e.g. how often they meet. Lastly, Context is the component in regard to the environment in which the KD takes place. Since the study investigate KD in product quality improvement projects, the context will reflect this setting. The double arrow in the research framework illustrates the two-way communication between the actors. This communication allows feedback to loop back and forth, making it possible for the source to tailor the knowledge content for a specific recipient. 20 4 The studied area This chapter describes processes, tools and other circumstances to give an understanding of the studied area at VGT. Explained are the different types of QJs, identified quality problems addressed in the studied QJs, the QJ process itself, roles involved in the cross functional QJ teams, tools used during the QJs as well as characteristics of important meetings. 4.1 Types of Quality Journals All QJs share the same purpose and follows the same process, but QJs can be more or less critical for VGT. Therefore, there are three different types of QJs with different levels of prioritization. The different QJs are called Quality Journals type C (QJC), Quality journals type B (QJB) and Quality journals type A (QJA) by the researchers. QJCs are the normal and most common type of QJs. QJBs are more critical than QJCs and can deal with possible safety concerns. In QJBs an upper level group will decide whether the QJB should continue to be treated like a QJB or be downgraded to a QJC. QJAs share the highest type of QJ priority together with QJBs and deals with the most complex and costly quality problems. The QJAs are initiated by an upper level steering committee and lead by a so-called QJA holder with full authority to obtain desired resources. A QJA can include a number of QJCs. Table 3 illustrates the different types of QJs. Table 3 QJ project characteristics and their priority. 4.2 Addressed quality problem The QJC was initiated to solve a problem caused by an earlier found problem of a technical nature. The supplier delivered a component with quality issues leading to a leak. When the component broke it was replaced with a new but similar component from another supplier. However, remaining liquid in the cabling from the old component could flow back leading to a newly installed component breaking as well. In the QJB the identified problem was a cover getting loose on a supplier delivered component. The root cause was found to be a specific state of interaction between physical, chemical and software factors. The QJA was started after a number of QJC projects and QJA projects had already been performed to solve the identified quality problem. However, the multiple root causes had not been fully understood. The problem area has therefore been known widely within GTT for a number of years. The main root cause is of an almost theoretical nature. QJ types QJA QJB QJC Priority Highest High/Highest High 21 4.3 The Quality Journal process All QJs regarding Powertrain Engineering components follows the same eight steps of the QJ process illustrated in Figure 11. Most steps of the current QJ process have been used by QJs for years but the 7D and 8D steps were implemented more recently. The eight steps of the QJ process are described below (Volvo Group Trucks Technology 2013a). Figure 11 Time Describe problem Assign QJ Containment action Investigate QJ Develop solution Implement solution Start 1D 2D 3D 4D 5D 7D 8D6D Follow up Follow up End MR-meeting Conclude knowledge Dissolve team Find root cause & solution Follow up & prevent reoccurrence The QJ process illustrated as steps with highlighted events. Describe problem - Investigate and document the problem, take a decision to open the QJ, appoint a PMQJ and a PQL. Assign QJ - Define the QJ team, hold a QJ kick-off, start the investigation step with identification and verification of the root cause, check former situations, secure resources for the investigation. Containment action - Decide if containment action is needed, implement the containment action and inform the market. Investigate QJ - Establish the root cause, propose solution (Business Case, Impact Analysis), verify root causes and identified solutions and create an action plan. Develop solution - Complete and release final development of solution including design changes, document and release production and aftermarket strategy. Implement solution - Implement solution in production and in aftermarket, conduct Market Ready meeting (MR-meeting), make service instructions ready and make parts available at warehouses. Follow up - Check the effectiveness of the implemented solution, e.g. by checking the fault frequency in the market after a time frame determined by the QJ team during the MR-meeting. Follow up - Prevent reoccurrence (check lessons learned). The 8D step is supposed to assure that quality issues do not reoccur and the QJ is now ready to be archived. 4.4 Quality Journal team members In a QJ, personnel from different departments participate depending on the characteristics of the identified quality issue. Some roles are however standard to gather at the QJ kick off, including personnel from: QCS, R&D, Purchasing, Aftermarket and GTO. Individuals that are part of the QJ team works for the QJ from their own department and on a daily basis usually deal with similar tasks and problems as they do in the QJ. The individuals in the QJ team also lead the QJ work within their department. Since QJs can be lengthy it is not certain that the same individuals will be part of the QJ team from start to end. The roles described in this section were part of the studied QJs. See Figure 12 for their usual involvement in QJs. 1D 2D 3D 4D 5D 6D 7D 8D 22 Figure 12 PQL PMQJ QJA Holder R&D Purchasing Aftermarket GTO Auditors Suppliers Time1D 2D 3D 4D 5D 7D 8D6D The participating roles mapped against the QJ process. Project Quality Leader (PQL) Before a QJ is opened a PQL working at QCS is responsible for evaluating market input, identify critical quality issues and initiate QJs. When the QJ is opened the PQL is a QJ team member but only participate in selected QJ meetings. A PMQJ leads the actual QJ work from 2D to 6D, while the PQL is monitoring the overall QJ progress. After 6D, the PQL is responsible for conducting the 7D and 8D step. Project Manager Quality Journal (PMQJ) The PMQJ leads the QJ work from 2D to 6D and have an overall responsibility for the QJ projects deliveries. The PMQJ has a central and important role within the QJ, organizing the team meetings and they are key for the team communication. The PMQJ also decides which specific departments should be part of the QJ team. Usually a PMQJ manage more than one QJ in parallel. QJA Holder In QJAs, the QJA Holder is the owner of the quality problem and has authority over all resources available for solving it. The QJA Holder is responsible for all QJA activities and is actively managing the QJA team from 1D to 8D. Compared to a PMQJ, the QJA Holder is in charge of all concerned resources and do not have to negotiate with department managers to access their personnel or other resources. R&D personnel Since QJs often deals with technical issues R&D most often have personnel in the QJ teams. Depending on the specific R&D department involved they will contribute with a GCR or a design engineer. The GCR decides who should participate since GCRs have the outermost responsibility for their components while design engineers are part of the GCR’s team. Whoever participates, the GCR will still take part in important QJ meetings and keep up to date with the QJ progress. In the QJ, R&D personnel contribute with technical knowledge regarding the quality problem and are responsible for designing technical solutions if necessary. Usually the GCR or the design engineer is supported by their department colleagues in their QJ work even if they are the actual QJ members. In the studied QJA also a so-called engineering leader from R&D was part of the QJ team. The engineering leader was in charge of the technical efforts made to find the root cause and developing the technical solution. The engineering leader organized technical meetings regularly with personnel from involved R&D departments. 23 Purchasing personnel Purchasing deals with supplier relationships, order changes and plan for introductions of new components in VGT manufacturing plants. However, purchasing personnel are more of a support function to the QJs and QCS as well as R&D personnel are more central. Purchasing individuals in QJs usually belongs to the operations part of the purchasing department. Aftermarket personnel Aftermarket personnel take part in QJs since they are the link to the affected customers. In QJs the Aftermarket personnel are not involved in the technical root cause analysis. Instead they take part in the QJs when appropriate. GTO personnel GTO personnel are regularly part of the QJ team. In QJs, GTO contributes especially with knowledge regarding manufacturing. However, in the studied QJs GTO are only involved in one QJ due to the nature of the specific quality issues and chosen solutions. QCS and R&D Auditors QCS and R&D auditors take part in QJs to help address and improve system and process issues. When they attend QJs, they usually only take part in the MR-meetings. Suppliers If a supplier has developed the component addressed in a QJ, the supplier together with R&D is supposed to analyze the technical root cause jointly. Suppliers therefore have an impact on the QJ progress and outcome. Sometimes there are differences in agendas between suppliers and VGT complicating the collaboration. It is important that the suppliers are synchronized with GTT and GTO in order to align their manufacturing with VGT’s when a design change of a supplier component is needed. 4.5 Quality Journal process tool Argus To follow the QJ process the PMQJ and PQL uses the Argus tool. Argus acts as a guide for all process steps and keeps track of requirements for passing project gates. Content must be added to Argus by the PMQJ and the PQL in order to pass the gates. This way Argus also functions as a database for storing information from the QJ. However, information used during the ongoing QJ is also stored in another online server environment accessible for the QJ team members. PMQJs and PQLs can furthermore use Argus to read about old QJ project when faced with similar quality issues. 4.6 Market Ready meeting The MR-meeting is a cross-functional meeting ending the 6D step. Participating is the PMQJ, the QJ team and a PQL. QCS and R&D auditors are also invited and usually take part. Sometimes GCRs attends if they are not already part of the QJ team. The PQMJ is in charge of the meeting and have a checklist for what to do. Other participants share their knowledge and give input. The MR-meeting is partly a method for sharing knowledge but also a forum for assessing if the criterias to reach market ready status as well as implementing a solution have been met, e.g. confirming the root cause. Mostly technical related knowledge is discussed but also process related knowledge. During the MR-meeting a decision is made on how to evaluate the success of the implemented solution and a discussion is held in order to improve the QJ process itself, e.g. teamwork and 24 efficiency. These lessons learned are documented in a so called a Quality Journal White Book which is developed or presented at the meeting. Also, during the MR-meeting a discussion on how to avoid the quality problem from reoccurring also takes place and it is decided if actions should be taken to address the newly developed knowledge. MR-meetings were implemented in the QJ process during 2012. 4.7 Quality Journal White Books The Quality Journal White Book (QJWB) is a tool for documenting lessons learned from QJs. It is used to describe positive and negative experiences about the QJ’s process and result, as well as ideas for improvements. The QJWB can also contain general reflections and a summary about the quality problem. The QJWB is mostly focused on the QCS owned QJ process and little content are aimed to other departments. Input to the QJWB is accepted from anyone on the QJ team but the PMQJ are responsible for making the QJWB. 4.8 The Design Verification Guidelines database The Design Verification Guidelines database (DVG) is GTT’s KM tool. It contains information about truck components and is mostly used by R&D. The DVG is used to manage design knowledge at the Product Attribute, System and Component levels. For each component there is a file in the DVG. DVG serves to guide design engineers all along the design process of the product and facilitates to share best practice within GTT. The Global Systems Responsibles (GSR) and GCRs are responsible to create and maintain the DVG. They have to capture and analyze knowledge in order to create standard engineering guidelines. The intention of the DVG is to update it as GTT learns, so the DVG becomes a key repository for engineering knowledge. These statements are based on internal GTT documents (Volvo Group Trucks Technology, 2013b). However, the use of the DVG is different among different R&D departments and so are the amount and type of the DVG content. Different R&D departments have furthermore been using the DVG for various times. 25 5 Results In this chapter the results from the data collection is presented. The main opinions and statements of the interviewees in regard to KD from the QJs are outlined. The chapter is structured according to the QJ process in the sections: General results, 1D to 5D step, 6D step, 7D to 8D step, Tools used for KD and After the Quality Journal. 5.1 General results In this section broad themes regarding KD from QJs are presented. Results linked to QJs in general are also described. 5.1.1 Reoccurring quality issues Some of the interviewees have experienced reoccurring quality problems from earlier QJ projects, implying that a new QJ have been initiated to solve the same problem. Two explanations are employee turnover and a not fully understood problem root causes when the QJ is ended. Another mentioned explanation is that e.g. former GCRs now part in R&D development projects do not read up on newly developed component knowledge since the former GCRs believe their old knowledge is still sufficient. During the interviews it was also stated that multiple QJs on similar issues sometimes are running simultaneously without knowing about each other. One interviewee said that in the context of a big organization it is not unusual that different people try to solve the same issues at different locations. This was the case for one of the studied QJs where the problem was being solved on the European market while it simultaneously existed in the US without the European organization of VGT knowing it. In this case, the communication issue was discovered by the component supplier, who delivered the addressed component to VGT in both Europe and the US. 5.1.2 Attitudes to Quality Journals The general attitude towards QJs expressed among interviewees is that they truly do not want the customer to suffer from quality related issues. An experienced engineer from R&D explained that it is not positive to have a QJ on one’s component. However, some components are naturally more exposed to quality issues than other. The reason may vary but it is usually possible to design a component so it can handle more than it needs to, but that will be a question about costs. One PQL have experienced that resistance may occur when discussions about opening a QJ starts. The reason is that QJs take resources from other projects and some R&D interviewees confirmed that they do not want QJs to take too much time from their daily work. However, after launching QJs they seem to receive full priority. A PMQJ also told that the priority of QJs is clear and the attitude towards QJs conforms to the priority anchored by top management. A QJA is supposed to have top priority but one interviewee have doubts whether QJAs always gets the appropriate priority. 26 Regarding knowledge creation, one R&D interviewee expressed that this is done at R&D and that knowledge is not disseminated from a temporary organization like a QJ. For the interviewee the QJC, QJB and QJA was merely a status implying that “it’s work as usual, but with higher priority”. Two other interviewees agreed and one explained that a QJ is a “forum for reporting, the real work is done outside the forum”. One R&D interviewee also claimed that there is no difference in how knowledge is handled at R&D based on the priority of the QJs. Another R&D interviewee further expressed that QJs are opened too often and ended too easily. 5.1.3 Key Performance Indicators on the Quality Journal process In the QJ process there are Key Performance Indicators (KPIs) on lead-time between step 1D and 6D. Many interviewees expressed skepticism to this and one PMQJ argued that it is contradicting that QJs are measured on time while R&D are judged more on quality (R&D projects are however also measured on lead time). Two interviewees both stated they have had experience from QJs being closed “just to meet the KPIs”. Other interviewees suggested that there should be more KPIs on improving the quality or solving the root cause and less on QJ lead-time. One interviewee said that KPIs focused on QJ lead-time are one of the biggest challenges in regard to KD. Regarding KPIs on step 7D and 8D, PQLs and other interviewees confirm that there are currently no KPIs for measuring KD. One PMQJ argued that preventive actions are not prioritized because there is no follow up on them for the PMQJ and another PMQJ explained that “you only get credit when you solve the problem on time”. 5.2 Quality Journal step 1D to 5D In this section the result reflect the first five steps of the QJ process focused on forming a team and finding a solution. The interviewees’ perception of the development of solution is described. 5.2.1 Developing a solution in cross functional teams The interviewees told that they consider VGT to be good at finding and addressing quality problems and argued that the QJs extra resources assist in doing this effectively. The early steps of QJs are described as intensive and involved individuals are available and alert. The communication is described as straight, fast and effective during the QJ. However, prior to the end of the QJ, solving the root cause is considered more important than reflecting on new learnings and KD from the QJ. The interviewees have an overall positive attitude to the constitution of the QJ teams and in both the studied QJs as well as in general, the cross functional team co-operation functions well according to interviewees. According to one interviewee, the collaboration between the different functions helps to close the knowledge gaps needed to solve the problem. However, in the studied QJA some interviewees stated that for one R&D department, with a component considered very relevant for the QJ, the participation and involvement was mostly missing. Also, even if the QJ teams are cross-functional, a big part of the QJ’s work to finding the root cause and developing a solution are done by R&D members of the QJ and other personnel at the specific R&D department. This means R&D usually have a central role within the QJ and that is the case for the studied QJs as well. 5.2.2 Quality Journal team meetings According to a number of interviewees, the frequent QJ team meetings facilitate dissemination of knowledge in a QJ. During the QJ meetings the work of each participating department is discussed as well as the progress made. The frequency of the QJ meetings depends on what step the QJ has reached. In two of the studied QJs the MR meeting was also used as a forum to decide whether or not the created knowledge was relevant to add into the DVG. 27 KD from QJs also occurs by informal methods. Short updates and briefings regarding QJs takes place during chats in the lunchroom and in corridors when QJ team members meet colleagues at random. In one of the studied projects, one interviewee said that KD was vastly facilitated due to the shared office space by two R&D component departments involved in the same QJ. Another interviewee expressed the difficulties to manage KD when two interdependent departments in a QJ are located on different floors in the same building. To some extent these issues are overcome by the cross functional QJ team meetings. 5.2.3 The Quality Journal process tool Argus Regarding the QJ process tool Argus the main user is the PMQJ, but other QJ team members can also use Argus to e.g. catch up on missed QJ meetings. However, a PMQJ interviewee told that, expect from PMQJs, others do not understand how Argus works and think the tool is complex. This is confirmed by R&D interviewees that mean Argus is less user-friendly than the DVG and one of the R&D interviewees stressed the need of education for using Argus. In a QJ, all members should have access to Argus according to a PMQJ interviewee. To get access an invite is required. R&D interviewees told that they do not always have access and they also argue Argus is mainly a tool for QCS personnel and the QJ process itself. One R&D interviewee points out that access to Argus is expensive and therefore it is common for R&D members to not have access after the 6D step. 5.3 Quality Journal step 6D In this section the results in regard to implementing the solution in the 6D step of the QJ process is described. The 6D step is the last step where the QJ team is intact and knowledge from the team progress is concluded and addressed. 5.3.1 Acknowledged knowledge to address In the studied QJC the developed solution was addressed to R&D and Aftermarket. The knowledge obtained from the project concerned the importance of conducting failure mode and effects analyses before changing a component design. The earlier change of the design was approved to easily according to one interviewee. Other knowledge was about the need for VGT to secure that suppliers had reliable processes. The knowledge from the QJB project concerned how the problem originated and how software and hardware can be changed to avoid the problem. The knowledge was mostly connected to the design of the supplier owned component. In the QJA the team realized that previous problem solving efforts made on the same problem have had a wrong approach. The underlying problem was the same through all attempts to solve it and new design changes were necessary to grasp the real root cause in the end. This QJA has also generated a lot of spin-off knowledge, not related exclusively to the failing component. 5.3.2 Knowledge recipients In general, the knowledge created in QJs is addressed to the departments of: R&D, Purchasing, Aftermarket, GTO and QCS. Some interviewees argued that suppliers are recipients too. The main recipient department largely depends on the characteristics of the problem and the created knowledge. In the studied QJs the interviewees consider R&D the most important knowledge recipient and the DVG was mentioned frequently as a tool for storing and reusing knowledge. To some extent knowledge usually concerns the component involved in the QJ and therefore the recipient departments are usually part of the QJ. One interviewee from Purchasing consider Purchasing’s role in the QJ as secondary compared to R&Ds involvement and states that Purchasing is strongly supported by R&D. Therefore knowledge capture from a QJ is more important at R&D since the consequences of losing knowledge from a QJ at Purchasing is less critical. It was also 28 mentioned that Aftermarket, Purchasing and GTO commonly adjusts their work based on work at R&D. One interviewee stated that the QJ process facilitates finding a knowledge recipient since the QJ process defines who is responsible for different QJ activities. The knowledge recipient department can however sometimes be hard to find. One R&D interviewee means that when the problem and knowledge relates to multiple components functioning in a system, it is hard to identify a knowledge recipient if there is not a certain individual responsible for the whole system. Knowledge can be complex and another R&D interviewee figures that knowledge regarding R&D might be easier to address compared to knowledge regarding other departments. 5.3.3 Taking action on acknowledged knowledge The 8D step of the QJ process supports capturing new knowledge, and it is expected that lessons are learnt from QJs. The interviewees agreed upon the need for lessons learned but how well the QJ process supports this is not clear. Some interviewees see the need for a clearly stated way of disseminating knowledge from QJs and lack a way to receive feedback on actions taken for KD. One interviewee also thinks that responsibilities for what to do and how to do the KD should be clearer for both the QJ team and the recipient departments. It is unclear if the knowledge receiving departments has supportive processes for taking care of QJ created knowledge. This is especially unclear for other departments than R&D, where a well-defined process appears to be missing. However, this does not imply that knowledge is not taken care off. A number of interviewees say that they rely on the individual responsibility of QJ team members to share their knowledge with their departments and take responsibility for capturing knowledge. Regarding the responsibility for taking action on QJ created knowledge, a PMQJ stated that if decided actions are not performed, the responsible recipient department should take the blame and PMQJs should not follow up on actions addressed in the 6D step. Actions on acknowledge QJ created knowledge is performed during many steps of the QJs and sometimes the QJ demands DVG updates before the 8D step. PMQJs stated that it is not always necessary to wait until the 8D step to prevent reoccurrence, because if the QJ can validate the solution sufficiently during the QJ, e.g. at the MR-meeting, the DVG can be updated in the 6D step before the 7D step is executed. All interviewees that talked about the QJ enforcing a DVG update were positive. Also at the MR-meeting, content are sometimes added to the sections of Argus that should be completed in the 8D step since the content is already known. In regard to the studied QJs the interviewees stated that design changes of components and DVG updates, in some cases, was done to preserve the created knowledge. One R&D design engineer stated that “we don’t talk about it, we build it in in our designs”. The interviewees stated that preventing reoccurrence is most often associated with updating DVG and in two of the studied QJs updating the DVG were a requirement. In the QJA, further actions was also taken e.g. developing videos explaining the problem root cause, introducing new systems for diagnostics in workshops, implementing new methods at GTO and using the new solution in ongoing R&D projects. However, R&D project gates and KPIs sometimes make new product development projects incapable of receiving knowledge from ongoing QJ. In the QJA this caused problems since the R&D development project was inflexible to adjust their project process to use the new QJ created knowledge. The QJA had created new knowledge and wanted it to be utilized as fast as possible in an ongoing R&D project. The R&D project was however not willing to use this new knowledge since it would prolong the project and have a negative impact on project KPIs. Currently there are no guidelines for how to solve conflicting like this. Furthermore, the type of knowledge can have an impact on what actions are taken. One interviewee described two different types of knowledge content, technical related knowledge and process 29 technical related knowledge. The technical knowledge is addressed to the GCR of the relevant component, whereas the process technical knowledge is perceived as more difficult to manage. There are process owners to address but their processes concerns many stakeholders making this type of knowledge content more difficult to disseminate. Another circumstance that impacts actions taken on QJ created knowledge is the size of the VGT organization. With many employees and sites in a number of countries it takes time to change something globally after a decision is made. Also, in two of the studied QJs the created knowledge was considered sensitive by VGT and therefore communications was restricted in order to avoid information leaks. Due to this fact, knowledge was not disseminated openly and some communication channels were used sparse. 5.4 Quality Journal step 7D to 8D In this section the results in regard to the 7D and 8D steps of the QJ process, checking the solution effectiveness and prevent reoccurrence, is described. The parts reflect expectations, feedback, attitudes to KD, the interviewees’ perception of KD success and responsibilities. 5.4.1 Expectations on the 7D and 8D steps QCS interviewees, supported by others, told that the 7D and 8D steps was earlier performed poorly, but now the communication between the departments are improved and as a result knowledge is preserved to a larger extent. Also, according to QCS interviewees the 7D and 8D steps have been performed better since they became mandatory. However, the 7D and 8D steps are still regarded the weakest steps in the QJ process and how to practically prevent reoccurrence in the 8D step is not explicitly stated according to a PQL. Some of the PQLs also said that they do not know what the recipient departments do in order to preserve QJ created knowledge. 5.4.2 Feedback on addressing knowledge After the team has dissolved in the 6D step, there is less communication between the QJ and R&D. The feedback regarding preventive actions and the solutions performance on the market is described as sparse. The PMQJs are no longer in charge of the project, and one PMQJ said that it is the PQL’s responsibility to follow up the actions taken by R&D regarding preventing reoccurrence, as the PMQJs do not. However, R&D has the DVG KM tool, making it easier to follow up their preventive actions compared to actions at the other considered recipients e.g. Aftermarket or Purchasing. Interviewees say it is up to the PQL to check whether the DVG has been updated with the QJ created knowledge or not. At R&D, the GCR gets an automatic email sent from Argus at the end of the QJ. This email contains a description of the QJ and a prompt on who to contact if no content will be added to the DVG. One R&D interviewee said that they leave no feedback on any actions they take after the MR-meeting. What content to update the DVG with or when the update will take place is not discussed with the PQL. For R&D it is a standard process to evaluate if anything should be added to the DVG after all kinds of projects. The PQLs can check if the DVG was updated themselves, but they get no standard feedback from R&D about actual updates. One PQL interviewee stated that it would be good if they did know whether the DVG had been updated, but to read the DVG themselves is not an option. For an illustration showing the feedback communication between the QJ and the R&D organization see Figure 13. 30 Figure 13 R&D departmentQJ Feedback on DVG updates and actions for preventing reocurrance Email to GCRs with QJ description and prompt about DVG updates Feedback about the solution effectiveness on the market The feedback communication between the QJ and the R&D organization where an x indicates that the feedback channel is not formally anchored in the process. Regarding the feedback about the solution from the PQL to R&D the majority of the R&D interviewees are not fully aware of how well the implemented solution actually works in the market. Some of the engineers at R&D are not even sure when the QJ project is actually finished. One common assumption from R&D is that, if they do not hear anything about the problem the solution is probably working satisfactory. Also not all R&D interviewees are interested in getting feedback from the market and R&D employees that are interested will have to ask for feedback from QCS to get any information. An exception to this way of working can however be seen in the QJA were the solution effectiveness was monitored fortnightly in a cross functional meeting after the QJ end. 5.4.3 Attitudes to Knowledge Dissemination from Quality Journals According to the interviewees, there is a commitment at VGT to improve KM in general. However, from a number of interviews it was evident that VGT personnel have a stronger focus on problem solving than on problem prevention. The general focus of the QJ process is the first six steps and not the 7D and 8D steps. One R&D interviewee explained that “you solve something, and understand it. When it is done, it is more like, nice we made it, but there is no real effort in keeping it [the knowledge]”. Similarly, another R&D interviewee said “you will always get to know the problem. But for me it is more like, that was it, let’s move on”. Yet another interviewee told that they feel proud when they solved a complex problem, but afterwards regardless how proud they are, KD is simply assumed to take place. One interviewee states that “keeping knowledge is unfortunately not on top of our agenda, we could be better”. Also an R&D interviewee stated that “from the QJ, we hear nothing about dissemination of knowledge”. Moreover, a PMQJ said that in recent years KD have been more highlighted which per se is an improvement. A PQL also told: We have comments about it, but it is mostly on my side. It is my job. If we explore something new, or learn something we always try to transfer it. However, it is mostly the DVG we have in mind. Another PQL thinks alike and state that even if the 8D step is recognized as important,