Knowledge Management and Delft Cluster

Document Sample
Knowledge Management and Delft Cluster Powered By Docstoc
					Knowledge Management and Delft Cluster
Roland K Price, Trekker, Theme 7 Delft Cluster

"To the small part of ignorance that we arrange and classify we give the name
knowledge" Peter, L J, The Peter Prescription: How to make things go right. Bantam
NY 1972

1. Introduction
It is difficult to identify the nature and purpose of knowledge management for an
organisation such as Delft Cluster. This is not surprising for several reasons:
 Knowledge management is a comparatively new subject that has only recently
     emerged as a serious area of activity in its own right
 It is primarily the commercial business sector that has focused on knowledge
     management because it has identified the need to manage existing knowledge
     assets in a better way if companies are to survive
 Consequently, much of the literature on knowledge management focuses on re-
     engineering businesses to improve performance
 There is comparatively little written about knowledge management in knowledge
     generating (scientific/engineering research) organisations, in that it is usually
     assumed that they already know how to do it.
The fact is that knowledge management is as important, if not more important, in
knowledge generating organisations than it is in commercial companies. In
particular, Delft Cluster, as a research organisation, needs to be able to target
knowledge generation accurately and efficiently so that best value for
(ICES/stakeholders/own) money is obtained, as well as disseminating the generated
knowledge effectively to stakeholders. Knowledge management can bring
significant improvements in both of these areas, as well as producing additional
benefits for Delft Cluster and its member organisations.

This note enlarges on the proposal submitted as Chapter 7 in the Delft Cluster
Onderzoeksprogramma 1999-2002 (July 1999)

2. Knowledge Management defined
There are a large number of different definitions proposed for knowledge
management. A few of these are as follows:

       A systematic and organised attempt to use knowledge within an organisation
       to transform its ability to store and use knowledge to improve performance
       (see KPMG

       An organisation creating, maintaining and using its previous experience to
       inform its future behaviour (see University of Durham

       The identification and analysis of available and required knowledge, and the
       subsequent planning and control of actions to develop knowledge assets so as
       to fulfil organisational objectives (see Ann Mackintosh, University of
       A system for managing the gathering, organising, refining, analysing and
       dissemination of knowledge in all its forms within an organisation. It
       supports organisational functions while addressing the needs of the individual
       within a purposeful context (see Jackson,

       Knowledge management caters for the critical issues of organisational
       adaptation, survival, and competence in the face of increasingly
       discontinuous environmental change. Essentially, it embodies organisational
       processes that seek synergistic combination of data and information-
       processing capacity of information technologies, and the creation and
       innovative capacity of human beings (see Daryl Morey:

These definitions are similar while emphasising different aspects. The following
definition for Delft Cluster draws on several of these definitions:

       Knowledge management is the strategic and systematic acquisition,
       application and dissemination of existing and generated knowledge by Delft
       Cluster to ensure its productivity, adaptation and competence in the face of
       increasingly discontinuous environmental change. It centres on
       organisational and business processes that provide a means of sharing both
       explicit and tacit knowledge assets within an environment formed out of a
       synergistic combination of information and communication technologies and
       the creative and innovative capacities of human beings. The environment
       fosters individual and organisational learning.

The length of this definition is deliberate so as to include a number of different
aspects that are generally perceived as being important for knowledge management,
and will have to be considered for Delft Cluster.

3. Emergence of knowledge management
But why should knowledge management become important now when we have been
handling knowledge for millennia? What is this 'discontinuous environmental
change' that makes a structured approach to knowledge management so important?

The answer to these questions can be seen in the way we are moving in society from
an age when we were dependent on science and scientific rationalism, to what many
philosophers call the age of 'post-modernism'. A summary of these changes in
regard to problem solving is given in the table below (cf

We are moving away from having well-structured problems with analytical solutions
to multiple views of both the nature of the problems and their solutions. In
particular, we are having to accept that knowledge is dynamically and continuously
evolving, has tacit and explicit dimensions, and has contextual factors of subjectivity,
interpretation and meaning. This is evidently the case in areas such as medical
ethics, political systems and business management, whereas some might argue that
engineering is more traditional in its approach to problem solving. However, we are
seeing that the hard and fast boundaries surrounding civil engineering in particular
are being eroded. Today we have to be far more conscious of social, political,
ecological and economic value judgements. Society holds engineers accountable as
much as professionals in any other discipline. Besides the accountability to society
there is also a shift towards treating situations more holistically than previously. For
example, in coastal morphology there is greater reluctance to intervene with large
scale engineering works, and instead to have a form of 'continuous intervention'.
Such an approach is being advocated in a number of other different areas, such as
river basin and urban management. Engineering cannot avoid the trends in culture
and society.

                                 Age of science (information     Post-modernism
                                 management)                     (knowledge management)
Process of understanding         Prescriptive - descriptive      Creative
                                 Problem - answer                Shifting solutions in flux
Aims and end result              There exists a 'right' answer   Ownership/participation
                                 Definitive information          Solution that works now
Task breakdown                   Individualistic                 Social
                                 Access via username             Gained by interaction
Nature of problems               Tame problems                   'Wicked' problems
solved                           Condition - response            Approaches and concepts
Method of working                Gather, analyse, formulate,     Shared understanding
                                 Test and predict
Approach                         Reductionism                    Holism
                                 Sections, fields and tags       Cultural perspective


5. The human dimension of knowledge management
Delft Cluster is an organisation whose members have a large range of appropriate
knowledge assets. Each member organisation is, in fact, a knowledge enterprise in
which its knowledge assets are used to support its core business activity, whether
consultancy, research or education in various branches of civil engineering.

The most valuable assets of any one of the Delft Cluster organisations are its staff. It
is the human individual who is the creator of knowledge, not any sophisticated ICT
facility. This important statement, as obvious as it may seem, is the basis of the
theory of knowledge management developed by Nonaka and Takeuchi (1995). They
present an in-depth study of knowledge management in different organisations,
especially in Japan, and come to the conclusion that knowledge 'is created through
the social interaction between tacit knowledge and explicit knowledge'.

Tacit knowledge is that which is internal to human beings, taking the form of beliefs
and norms, understanding, skills and practices that may be difficult to articulate. On
the other hand, explicit knowledge is essentially something that has already been
made external to human beings, such as in multi-media documents, modelling
systems, information systems, best management practices, and so on. According to
Nonaka and Takeuchi knowledge is created (and transferred) in four different modes:
                     Socialisation:       Externalisation:
                     Tacit to Tacit       Tacit to Explicit
                     Internalisation:     Combination:
                     Explicit to Tacit    Explicit to Explicit

Generally, the transfer of knowledge takes place in the four ways described in the
table above. Each of these transfers has it's own particular difficulty. The
socialisation transfer is very dependent on the ability of human beings not only to
convey knowledge but for them both to appreciate their own and the other person(s)
context for giving or receiving the knowledge. This 'meta-knowledge' is also needed
in the other three modes of transfer. The heart of the problem of knowledge transfer
in all four modes is how to foster a shared experience; that is, how to share
knowledge at the individual and corporate levels.

Nonaka and Takeuchi address this issue by looking at different organisational
activities that facilitate the creation of knowledge. By examining the experience of
different organisations they come up with five important activities:
 Intention - an organisation's aspiration to its goals
 Autonomy - at individual and group level, so far as circumstances permit
 Fluctuation and creative chaos - continuous process of questioning and
    reconsidering existing premises by individuals
 Redundancy - intentional overlapping of information
 Requisite variety - rapid access to the variety and complexity of information in
    the specific environment

These activities function best within the context of an organisational knowledge-
creation process with five steps:
1. Sharing tacit knowledge- within an entity such as a self-organising team
2. Creating concepts - making tacit knowledge explicit through reasoning
3. Justifying concepts - determining the worthwhile-ness of concepts
4. Building an archetype - converting justified concepts into something tangible
5. Cross-levering knowledge - benefiting from archetypes in other areas

Such a process requires three categories of people to be effective: knowledge
practitioners, knowledge engineers and knowledge officers. These people function
best within what Nonaka and Takeuchi call a 'hypertext', 'middle-up-down'
organisation; that is, an organisation with cross-links between three structures or
 Hierarchy: for the acquisition, accumulation and exploitation of explicit
 Task forces: for creating new knowledge from existing tacit knowledge
 Knowledge base: to contains the corporate vision, organisational culture and
    technology, with a focus on storing and reinterpreting both tacit and explicit

There is little doubt that Nonaka and Takeuchi have much to say that is important for
knowledge management generally. In particular they stress the significance of the
individual within knowledge management. This has to be highlighted in the
knowledge management plans for Delft Cluster, such that the organisational work
culture and human resources development are as important for knowledge
management as any ICT functionality introduced to facilitate it. What is not clear
however, is the appropriateness of their theories for a new civil engineering research
organisation such as Delft Cluster. There appear to be obvious parallels that can be
the basis for a possible system. But there are also other areas that are not considered
by Nonaka and Takeuchi, namely, aspects of human resource development such as
exploiting the capacity of human beings for knowledge creativity (such as through
Neuro Linguistic Programming; see ), and
recognising that knowledge is encapsulated explicitly in human artifacts that are
themselves used by researchers and practising civil engineers to generate new
knowledge. Such artifacts include modelling systems, databases, information
systems and so on. This introduces the significance of ICT in providing access to
such artifacts and in facilitating communications, collaborative working, document
and content management and learning, both individually and corporately.

5. Knowledge management in Delft Cluster
Central to any knowledge management in Delft Cluster therefore is the recognition
that it begins with the people who make up the organisation, and in particular with
the researchers who are responsible for the seven themes. Management particularly
has a role in facilitating the integration of people with whatever ICT facilities that
may eventually be introduced. These will have the objectives of improving the
innovative capabilities of researchers in terms of the efficiency and effectiveness of
their research, and disseminating acquired and generated knowledge to Dutch
industry and further afield.

In the first instance we can follow van der Speck and de Hoog (quoted on and ask questions about five areas of activity
that Delft Cluster needs to consider in order to handle its knowledge assets
 Identifying existing knowledge assets
         Where are the assets located (within and external to Delft Cluster)? What are
         their contents? How are they used? What form do they take? How accessible
         are they? (What is the 'organisational memory' of the Delft Cluster? see Wiig
 Generating (creating) new knowledge assets that add value
         What knowledge needs to be generated? How can additional knowledge be
         acquired from elsewhere? How can new knowledge be created? How may
         this be done efficiently? How will the new knowledge be stored, used, and
         made accessible? How is acquired knowledge synthesised with existing
 Analysing how the total knowledge assets can add value
         What are the opportunities for using the assets? What are the effects of their
         use? What obstacles exist to their use? How can tacit knowledge assets be
         incorporated and shared effectively? What is the increased value to Delft
 Using the knowledge assets to provide added value:
         What plans should be made to use the knowledge? How will the plans be
         implemented? How do we monitor the use of the assets?
 Reviewing the use of the knowledge assets to ensure added value:
       What is the added value produced by the assets? How are they best
       maintained? What new opportunities are created?

We pursue answers to these questions by considering some particular issues.

External sources:                                                    Tacit
      WWW                                                          knowledge
    TUD Library
  Human networking

                                      DC Staff

  Insight -
                          Internal knowledge base

  Knowledge bank:                                 Legacy knowledge:
  Knowledge objects (text, multi-nedia)           Old databases
  Metadata (metrics, voting, classification)      Non-integrated databases
  Systems (modelling, GIS, MIS,                   Old modelling systems, GIS etc
  visualisation, monitoring, learning, etc)

  Natural selection of knowledge
   - highly/poorly rated knowledge

                      Individual and organisational learning culture

Knowledge management architecture vision (based on Morey (1998)

   5.1. Knowledge mapping
   Without a good understanding of the content of the existing knowledge assets it
   is very difficult to determine how to use such knowledge efficiently and what
   additional knowledge is required. One of the important initial tasks of Delft
   Cluster is to map its existing assets and to ensure that the map is continuously
   maintained. This is a virtual space where different types of knowledge can be
   brought together in an organised manner. The knowledge map forms a bridge
   between informal networks and the formal capitalisation of knowledge assets. It
   is not a repository of knowledge itself, but consists of pointers to knowledge
grouped around particular issues, and in particular to the actors (human or
computer-based) who hold the knowledge. (See nViews New Media Group,
Rogers Communication Centre, Ryerson Polytechnic University, and also University of Edinburgh This brings in the notion of its
maintenance by communities of (human) actors who share a common interest in
developing its scope. It is important to remember that it is human beings who
create, retain and develop knowledge. Such tacit knowledge is embedded in both
conscious and unconscious memory.

5.2. Knowledge categories
Knowledge comes, of course, in many forms, and its acquisition is generally
dependent on its form. Given the overall classification of tacit and explicit
knowledge then generally, explicit knowledge comes in such forms as
 Documents (multi-media)
 Best management practices
 Simulation models
 Monitoring systems
 Databases
 Geographic information systems
and so on. People acquire (tacit) knowledge from this range of sources, through
public and commercial suppliers, and from human experience and networking.

5.3. Knowledge classification
Generally, people's knowledge can be classified in four ways:
 Process knowledge: recipes for doing things well
 Factual knowledge: residing in people's heads, and easiest to document
 Catalogue knowledge: knowing where to find knowledge when it is needed
 Cultural knowledge: knowing how to get things done (invisible rules and
Explicit knowledge can be classified in a variety of ways depending on its form.
The challenge is to categorise and archive knowledge in such a way that is
accurate in retrieval, effective (fit for purpose) and accessible. It is important to
recognise also the contextual attributes of knowledge. This points to the
implementation of knowledge 'banks' in which meta-knowledge is archived along
with the knowledge rather than knowledge 'bases', particularly as much
engineering knowledge is case-based.

5.4. Knowledge sharing and learning
As explained above, the sharing of knowledge within an organisation such as
Delft Cluster is crucial to its development, growth and prosperity. Precisely how
to develop a knowledge sharing environment is a matter not just of top-down
management but also bottom-up involvement and enthusiasm. Nonaka and
Takeuchi see it more as a middle-up-down management style, where initiatives
are taken primarily at the team level within constraints and ambitions set by
management. Some useful, practical ideas about knowledge sharing are provided
by Morey ( These have to be complemented by
particular management initiatives, such as those outlined by Nonaka and
   Takeuchi. What is clear is that the desired work culture has to be clearly
   articulated by management as a goal of Delft Cluster.

   Whether the process is tacit to tacit or explicit to tacit, there is a problem of the
   human receiver being able to assimilate knowledge for his/her purpose. Effective
   assimilation means that the person's knowledge is increased and affects his/her
   performance. In other words, learning has occurred. Delft Cluster is a learning
   organisation in so far as it experiences change because the effectiveness and
   efficiency of its knowledge management improves over time. Note that Delft
   Cluster should also be an 'unlearning' organisation; that is, it should be prepared
   to abandon knowledge, processes and cultural norms that are no longer helpful in
   achieving its ambitions.

6. Development of a knowledge management system
Chapter 7 in the Delft Cluster Onderzoeksprogramma 1999-2002 is centred about a
functional description of a knowledge management system. The description is
deficient because it does not address in much detail the underlying sociological
issues such as knowledge sharing, and topics such as knowledge representation and
use of meta-knowledge. Some attempt is made to include these below, but again
within a functional framework. This time the 'life-environment' of knowledge is
used as the basis of the framework. By 'life-environment' we are using a term that
has non-linear rather than linear connotations as with 'life-cycle'.

The life-environment of knowledge has the following structure:


    Archive & Access                                    Analyse & Apply

     Generate                    Communicate                    Review

 Classify & Synthesize                                  Assimilate & Learn


Consider these components of the life-environment in order.

   6.1. Acquire
   The key sources of knowledge for Delft Cluster are:
    Personnel:
          Staff: development: recruitment, education/training, experience, etc
          Networking: informal contacts, email, discussion groups, conferences, etc
    Documents:
      External published documents: library, publishers, etc
      Internal/external grey literature: reports, memos, etc
 Systems:
      Knowledge bases
      Information systems: GIS, MIS, DSS, etc
      Data mining: ANNs, genetic algorithms, fuzzy logic, chaos, etc
      Modelling: DIANA, Delft 3D, etc
      Monitoring: boreholes, river/ocean water levels, rainfall, remote sensing,
      Visualisation: infrastructure
The key issues for Delft Cluster however are:
 Capture/retention of staff knowledge and experience
 Sharing of knowledge
 Access to published documents and grey literature
 Access to systems
 Promotion of a learning culture

The capture of experience is part of the tacit-to-explicit mode of knowledge
creation above, and involves knowledge sharing. This is both a technological
and a sociological issue. Technologies have to be provided that enable
knowledge to be captured at the point of creation or saving (such as
ART*Enterprise, GrapeVINE, Knowledger Software). Correspondingly, sharing
knowledge has to be encouraged by rewards: enhanced prestige, recognition, and
other benefits for the individual, such as improved access to knowledge and
learning tools.

The TUD Library is seen as the gateway to all literature from commercial
publishers. Grey literature is however, a different matter. It involves the
individual, the team or (project) group to which the individual belongs, and the
corporate knowledge base(s) of Delft Cluster. The individual expert is viewed as
the primary source of grey literature, whether he/she creates it, or acquires it
from elsewhere. The expert provides value judgements (meta-knowledge) on
grey documents, and shares them with the group (which may include a proactive
individual, or knowledge editor, to facilitate the transfer). In turn the group
makes selected documents available to the corporate knowledge base (distributed
but co-ordinated by the TUD Library), again with appropriate meta-

6.2. Classify & Synthesise
The tacit knowledge of individuals may be classified in terms of process, factual,
catalogue and cultural (see above). These have their explicit counterparts. Each
is important to Delft Cluster. In particular, process knowledge is an important
classification for engineering because it includes best management practices
(BMPs) and other engineering procedures that encapsulate how to do things well.
How an individual does something well is difficult to make explicit, though
techniques do exist to try and capture some of this type of knowledge.
Factual knowledge can regarded as residing primarily in documents (if it resides
in people's heads then it is comparatively easy to extract and document). As
stated above, this knowledge (as with any other knowledge) is generated or
acquired with a context and meaning. In particular, meaning is facilitated by
having an enterprise vocabulary that everybody works with. Consequently, there
is a need to identify these, among other, attributes when classifying the
knowledge under different issue headings. The attributes form a sort of meta-
knowledge of the original knowledge.

Catalogue knowledge is about how to find knowledge when it is needed, and can
take the form of a person as well as links between explicit knowledge items.
This meta-knowledge will, of course, be important when searching for other

Finally cultural knowledge is about knowing how to get things done within an
organisation, and involves the invisible rules and norms that will develop within
Delft Cluster.

It is important to have a map of the knowledge in Delft Cluster as explained
above. Such a map has a number of important benefits, such as efficient and
effective searching, identification of knowledge clusters and 'holes', setting up
communication links, and so on.

The synthesis of knowledge is best done in the development of expert,
knowledge-based systems. Some of the market leaders in this area are dbProphet
(Trajecta), PowerPlay (Cognos), and Extract (Evolutionary Technologies)

6.3. Archive & Access
In view of the need to regard a knowledge item (or object) as having a number of
attributes (or meta-knowledge), such as context and assigned meaning, it will be
important to develop the notion of knowledge 'banks' (similar to 'term banks').
The indexing structure of these banks has to facilitate the retrieval of knowledge
in a manner that it is available, accurate, effective and accessible (cf Morey, Accuracy is improved by searching on meta-
data, such as keywords and classification. Effectiveness is assured by the quality
stamp that one or more experts give to the archived knowledge. Accessibility is
ensured by the use of open Web technologies and using powerful search engines.
Typical of advanced search engines are Excalibur RetrievalWare and
WebExpress, and Internet Spider

6.4. Generate
Here, it is accepted that the primary source of new knowledge is going to be the
human experts. What is important is that they are provided with the sort of
analysis tools they need, including access to the systems described above, and the
analysis tools below. These include tools for knowledge discovery and mining,
document content analysis, and so on.

6.5. Analyse & Apply
Analysis of knowledge can be done with a number of different tools for
linking/synthesising, concept mapping, and so on; see Gaines and Shaw
                                            Such an analysis can be seen as part of
the creative process or to assist in application.

The primary purpose of the acquired and generated knowledge is, however, to
improve the application to problems of infrastructure in densely populated deltaic
areas. At both the creative and application levels Delft Cluster experts are
involved in projects that address specific issues or problems. It is at this stage
that we need to recognise the importance of electronic communications for
collaborative working. The basic Delft Cluster intranet should include all the
facilities that enable a project to be conducted effectively. Here there is a need
not only for quick and effective emailing, voice and video conferencing (eg using
Microsoft NetMeeting), but also the ability to reference and work simultaneously
with documents and systems. Process modelling can facilitate project
management. Meta-knowledge about the project decision making process as
well as knowledge about sub-processes, including uncertainty in decision
making, need to be inherent to the knowledge transfer process. Electronic
trading tools will enable financial transactions associated with projects to be
effected. Lotus Notes (IBM) or Microsoft Exchange, among other similar
systems, could provide the basic framework for project execution and other
knowledge management tasks.

6.6. Disseminate
One of the primary goals of Delft Cluster is to disseminate knowledge to
stakeholders, to organisations in the Dutch infrastructure and water sectors, and
even more widely to organisations in other countries. The basic network
infrastructure for knowledge management will be designed for this purpose also.
The concept is that all of the facilities developed to promote knowledge creation
within Delft Cluster should form the basis of an interactive exchange of
knowledge with partners, stakeholders, and other organisations interested in
making use of the knowledge.

An individual will be assisted to identify and locate the knowledge he or she
requires, whether stored in knowledge banks, encapsulated in modelling systems,
residing in databases, information or decision support systems, or dependent on
interaction with one or more experts. In addition, there will be the facility to set
up virtual teams of people, including staff from Delft Cluster and the client
organisation, to work interactively with all of the knowledge management
facilities on problem specification and solution.

6.7. Assimilate & Learn
As stressed above, the recipient of knowledge needs to appreciate the original
context of the knowledge and the implied meaning given by experts so that the
knowledge can be assimilated properly. This process of assimilation can be
regarded as a form of learning. Part of the knowledge management system
therefore addresses how the individual may best learn on-the-job and just-in-time
to improve performance. It is also concerned, however, with corporate learning
that comes through the research that is done.

6.8. Review
   One of the major problems with operating a knowledge management system is
   the constant need to maintain the knowledge base. Although there is the need for
   some servicing of the system, the content of the system, namely the knowledge,
   has to be maintained by the Delft Cluster experts. This should be a shared task,
   and is therefore best done using the concept of communities (synonymous with
   the Themes). These are groups of people sharing a common interest or practice,
   whose purpose is to share knowledge and experiences, help each other to learn,
   act as a support network, and as an informal centre of excellence. They would
   take on the task of ensuring that the knowledge banks and systems are up-to-date,
   quality assured and redundant knowledge or tools are removed.

   Another important aspect of review is to monitor and measure changes. Various
   organisations are experimenting with different indices to measure progress in
   knowledge management and in the different initiatives that are taken.

   6.9. Communicate
   Without effective communication with human beings the knowledge
   management system is unworkable. There is an obvious need for an efficient
   intranet with an effective man-machine interface. Much of the communication
   between the different elements of the knowledge platform (as it is called in the
   Delft Cluster proposal) should be handled using intelligent agent technology.

7. Organisation of knowledge management within Delft Cluster
The above description is an outline of some of the important issues that need to be
considered in developing an appropriate knowledge management system for Delft
Cluster. There are, similarly, organisational issues that provide the context for such a
system. These reflect the way in which knowledge management is stimulated and

The ambitions of knowledge management as stated in the proposal are
 To facilitate research (knowledge generation)
 To disseminate knowledge to stakeholders
 To broker knowledge to the infrastructure and water sectors world-wide
These will be achieved in the context of five areas of activity, namely
 Network infrastructure
 Collaborative working
 Document and content management
 Net-centric engineering
 Individual and organisational learning
It is interesting to note that KPMG, in their 1998 Knowledge Management Review,
comment that 'it will soon be commonplace for knowledge to be bought and sold in a
growing global market based on Internet technology and encompassing new roles
such as knowledge providers, knowledge brokers and highly-specialised knowledge
workers' ( Here brokerage is seen as being a focal point
for knowledge exchange, in which 'members' may be both suppliers and consumers.

The goals of Delft Cluster with respect to knowledge management could take the
form (see Boyd
 To make apparent knowledge creating and sharing activities so that all involved
   explicitly address knowledge management during the course of their research
   To plan and budget for knowledge management so that such activities are a vital
    component of the Delft Cluster ambitions
   To support knowledge management through the application of suitable
    information technologies
   To foster and promote knowledge creation and sharing as a critical and constant
    driver of Delft Cluster's strategic aims

In support of these goals it could be advisable for Delft Cluster management to bear
in mind the following points (following Malhotra
 View Delft Cluster as a human community capable of providing diverse
    meanings to information outputs generated by the knowledge management
 De-emphasise adherence to the way things are done within the separate member
    organisations so that Delft Cluster practices may be continually assessed from
    multiple perspectives for their alignment with the dynamically changing external
 Encourage diverse viewpoints by avoiding premature concensus
 Encourage proactive involvement (and development) of human imagination and
 Give explicit recognition to tacit knowledge and related human aspects
 Implement new, flexible technologies and systems that support and enable
    communities of practice (the Theme groups) and exchange of knowledge
    between such communities
 Ensure that the Delft Cluster information base is accessible by staff who are close
    to the action

Nonaka, I and Takeuchi, H (1995) The knowledge-creating company. Oxford
University Press p284
Davenport, T H and Prusak, L (1998) Working knowledge. Harvard Business
School Press, p199