are we communicating? reflection on knowledge management approaches in semiconductor industry by: -...

24
Are We Communicating? Reflection on Knowledge Management Approaches in Semiconductor Industry By: Stephen Dun-Hou Tsai, Professor, National Sun Yat-Sen University - Ching Fang Lee, Assistant Professor, Shih Chien University - Mansour Amjadi, PhD student, National Sun Yat-Sen University - Hong-Quei Chiang, PhD, Director of Taiwan Tobacco an Wine Corp.

Upload: jasper-lambert

Post on 02-Jan-2016

215 views

Category:

Documents


0 download

TRANSCRIPT

Are We Communicating?

Reflection on Knowledge Management

Approaches in Semiconductor Industry

By:

- Stephen Dun-Hou Tsai, Professor, National Sun Yat-Sen University

- Ching Fang Lee, Assistant Professor, Shih Chien University

- Mansour Amjadi, PhD student, National Sun Yat-Sen University

- Hong-Quei Chiang, PhD, Director of Taiwan Tobacco and Wine Corp.

Contact person: [email protected]

August 11, 2009

2

Prologue

The Complexity of Semiconductor Industry

3

Prologue (2)

The typical trend:Integration of Knowledge

Management (KM) and

Information Technology (IT)

for competitive advantages

(Grant, 1996; Kogut

& Zander, 1992)

– Example: Xerox Corporation and the development of “Eureka”

(Bobrow & Whalen, 2002; Davenport & Prusak, 1997; Nonaka, 1994;

McDermott, 1999).

4

Prologue (3)

• Conventional Approach to KM:

– Introducing advanced information system,

and

– Encouraging members to codify a systematic

documentation of their experience (Alavi & Leidner, 2001)

• The conventional approach resembles Resource-Based View (RBV)

(Wernerfelt, 1984; Barney, 1986)

5

Prologue (4)

• According to RBV:

– Firms have different collections of physical

and intangible knowledge and capabilities

which RBV calls ‘resources’

– Competitive advantage attributed to the ownership of

valuable resources

(Wernerfelt, 1984; Barney,

1986)

6

Prologue (5)

• Contemporary Approaches to KM:

– Knowledge can only emerge or created during

real-time interactions and work practices

(Kellogg et al., 2006; Tsoukas, 1996).

– Knowledge is an ongoing product of practice (Orlikowski, 2002)

– Knowledge and practice are not independent entities

(Orlikowski, 2002, 2007; Wenger, 1998)

7

Research Questions– What are the challenges and difficulties implementing the

conventional approach in troubleshooting?

– What is the nature of organizational knowledge?

8

Research Method

• Grounded theory techniques

(Eisenhardt, 1989; Glaser & Strauss,

1967;Strauss & Corbin, 1998).

• Comparative analysis of two companies in Taiwan

– ChipMaker and ChipTest (the names are pseudonyms)

• The two companies selected are leading in semiconductor industry with some similarities

9

Research Method (2)

Data Collection and Analysis: – Adopt a process-tracing methodology

• (Langley, 1999)

– Qualitative ethnographic approach • (Golden-Biddle & Locke, 1993)

– Observation of engineers’ practice:• Troubleshooting activities, installation procedure, routine maintenance

– Interviews: KM team members, managers, and engineers

– Research data included: ChipMaker: 83 interviews with 63 participants

ChipTest: 26 interviews with 12 participants

10

Semiconductor Fab and Packaging Process Sequence of production processes: Intellectual Property, IC design,

IC fabrication and IC testing.

IC fabrication process: diffusion, photo, etching and designing thin film

➊ Diffusion:

➋ Photo:

➌ Etching:

➍ Thin film:

Repeat Similar

Cycles

11

Knowledge Management Models:

Two Case Studies

– Case #1 : Object-Based KM (OBKM)

model at ChipMaker

– Case #2: Community-Based KM (CBKM)

model at ChipTest

12

Case #1 : Object-Based KM (OBKM)

model at ChipMaker• Introduction of Best Knowledge and Practice (BKP) model:

– ChipMaker: A global supplier of semiconductor

equipment and technological services.

– Incorporating four key elements: • technology, strategy, personnel and process

into a centralized KM system

– Established a knowledge repository database: “BKP”

13

Case #1 : Object-Based KM (OBKM)

model at ChipMaker (2)

• Implementation of BKP:

– 1) Recording work experience routinely

– 2) Follow a systematic and structured format

– 3) New knowledge logged into the BKP system

– 4) Approved knowledge in BKP system was rewarded

• Significant adjustments:– “knowledge exchange protocol” to improve communication

– Development of “lessons learned” concept

– Hired professional editors to make BKP more readable

14

Case #1 : Object-Based KM (OBKM)

model at ChipMaker (3)

• Paradox Consequences:– Successful quantitative indices of KM performance:

• 1= Increase in the number of knowledge sharing submissions

• 2= Reduction in time spent to identify valuable and workable knowledge

• 3= Reduction on installation time of new equipments

• 4= Increase of utilizing BKP and learning from the database

• Contradicting Qualitative Comments:“It is impossible to write down my experience on-site in detail. I don’t have

that luxury to sit down and spend time writing my best practice tips. To address a troubleshooting problem is like to save a life in an emergency”.

15

Case #2: Community-Based KM

(CBKM) model at ChipTest• Introduction of Technical Knowledge Community (TKC) model:

– Company: ChipTest: A global semiconductor manufacturer in Taiwan

– TKC goal:• a) Accelerating the output & increasing the speed of R&D capabilities • b) Shortening the time of learning curve • c) Improving troubleshooting abilities

– Creation of “Technical Knowledge Community” (TKC), based on engineers’ job functions and departments

16

Case #2: Community-Based KM

(CBKM) model at ChipTest (2)

• Implementation of TKC:

– Four basic principles

• Time-pacing

• Flexibility

• Direct participation

• Evaluation

17

Case #2: Community-Based KM

(CBKM) model at ChipTest (3)

• Outcome: Shifting from passive to active participation

– Senior engineers active participation, and sharing their own

experiences

– Participants’ new discovery and innovative ways for resolving

work-related challenges and troubleshooting

– Putting into practice the tips discussed in TKC and became

convinced of the usefulness

18

Case #2: Community-Based KM

(CBKM) model at ChipTest (4)

• Some positive comments:– “I can call help from knowledge communities at the moment when I

encounter tough technical problems. After discussion I have some alternative feedbacks which I couldn’t imagine before”.

– “While I was preparing, I had to reflect retrospectively on the whole process of how I had tackled the difficult problems. It triggered me to rethink the reasoning logic. The direct feedbacks from participants not only gave me invaluable tips, but also revealed my professional myopia”.

– “The TKC is like a platform of social networks. Though the knowledge content is important, the timely professional idea from others is even more attractive”.

19

Comparison of the two models

ChipMaker Case: BKP Model ChipTest Case: TKC Model

Pre-understanding Knowledge-as-object Knowledge-as-interaction

Key elements

Information System: Best Knowledge Practice (BKP)

Accumulate and transfer of knowledge directly

Top down Control BKP system

Technical Knowledge Communities (TKC)

Interaction and emergence

Four principles to enhance bottom up emergence

Reflection

Key to failure:Knowledge consists of multiple

cognitions

Particular setting and situation enacts a social network

Knowledge emerges right after

impromptu action

Key to success: Community accommodates connection

and collaboration Community shapes collective identity

Emerging knowledge is accumulated by metaphorical memory

20

Discussion and Implications• In BKP (case #1), it is the individuals who learn and

create knowledge. Organizational knowledge

accumulates to construct a company-wide capability for

troubleshooting by a workable IT system.

• In this object-based model, organizational knowledge

comes from coding the tacit knowledge located in

individual heads and translating that tacit knowledge

into explicit forms available to the organization.

21

Discussion and Implications(2)

• In TKC (case #2), systems, database, and written artifacts are considered as records that can only become knowledge when people use them as tools or boundary objects in their process of gesturing and responding to each other

• In TKC, new knowledge is not a given; it does not exist anywhere in the network in any form other than a potential, the form of which is to be unfolded by the experience of relating between agents at the local level.

22

Epilogue (1)

Lessons learned:

– Management’s role as ‘scene-setter’, not as a ‘scriptwriter’• Initiating some simple rules

– Management’s role as ‘community builder’

• Focusing on relationships and communities

23

Epilogue (2)

• Reflection:

– Management’s role as ‘facilitator’

• New voices: welcoming a wide diversity of views

• Constructive dialogues: Encouraging communication across

previously isolated knowledge sets

• New perspectives: being open to new ways of seeing /doing

things

• New passions: fostering a sense of shared destiny in the

organization

24

Thank you for your attention