“knowledge management caters to the critical issues of organizational adaptation, survival and competence in face of increasingly discontinuous environmental change. Essentially, it embodies organizational processes that seek synergistic combination of
-data and information processing capacity of information technologies,
-and the creative and innovative capacity of human beings”
This is a strategic view of Knowledge Management that considers the synergy between technological and behavioral issues
as necessary for survival in “wicked environments”.
The need for synergy of technological and human capabilities is based on the distinction between the “old world of business” and the “new world of business” Malhotra
Old world of business VS New world of business
Within this view, the old world of business is characterized by predictable environments in which focus is on prediction and optimization based efficiencies.
This is the world of competence based on “information” as the strategic asset
and the emphasis is on controlling the behavior of organizational agents toward fulfillment of pre-specified organizational goals and objectives.
Information and control systems are used in this world for achieving the alignment of the organizational actors with pre-defined “best practices”.
The assumption is that such best practices retain their effectiveness over time.
In contrast, high levels of uncertainty and inability to predict the future characterize the new world of business.
Use of the information and control systems and compliance with the pre-defined goals, objectives and best practices
may not necessarily achieve long-term organizational competence.
This is the world of “re-everything”, which challenges the assumptions underlying the “accepted way of doing things”.
This world needs the capability to understand the problems afresh given the changing environmental conditions.
The focus is not only on finding the right answers but also on finding the right questions.
This world is contrasted from the “old world” by its emphasis on “doing the right thing” rather than “doing things right”
In our thinking, Knowledge Management is a framework within which the organization views all its processes as knowledge process.
In this view, all business processes involve creation, dissemination, renewal and application of knowledge toward organizational sustenance and survival.
This concept embodies a transition from the recently popular concept of “informational value chain” to a “knowledge value chain”.
What is the difference?
Informational value chain VS knowledge value chain
The information value chain, considers technological systems as key components guiding the organization’s business processes,
while treating humans as relative passive processors that implement “best practices” archived in information databases.
In contrasts, the knowledge value chain treats human systems as key component that engage in continuous assessment of information archived in the technological system.
In this view, the human actors do not implement best practices without active inquiry.
Human actors engage in an active process of sense making to continuously assess the effectiveness of best practices
The underlying premise is that the best practices of yesterday may not be taken for granted as best practices of today or tomorrow.
Hence double loop learning, unlearning and relearning processes need to be design into the organizational business processes
In the previous definition, Knowledge Management embodies organizational processes that seek synergistic combination of capacities of information technologies and human beings.
So, if we want to adapt tools and organization in order to realize such synergy, we have to associate the knowledge capitalization with the use of expert system and the concurrent engineering with the use of integrated design.
Formalization of knowledge
Knowledge is the internal representation that people can do in their minds, when they acquire new information.
The same information is not consider with the same value for anybody, as the understanding depends strongly of what has been learnt before.
Information is transformed in knowledge with a cognitive process and depend of the context (time, location, environment) this information is taking into account.
In this way, you can formalized your knowledge and transcript it in information, but you only can capitalize out of your brain information.
Knowledge cannot be directly shared. Only information can be.
So we normally can use the term of formalization of knowledge, but we must also normally use the capitalization, sharing and re-contextualization of information.
Functional analysis id a method to define the needs the product has to answer. These specifications highlight the services the product has to give back.
For the functional analysis, we have first to define the frontier between the product itself and the external environment of the product.
The specifications concern the external functional analysis.
The design process will use the internal functional analysis
The specifications need five steps to be completely defined.
The first one is the specifications of the needs, what the user can expect from the product.
The second step has to define the product life cycle., not only the different phases from specifications to recycling, but also if the product is used in different situation, in different cycles.
For each of these cycles, the next step define the external environment of the product, why this environment concerns the product. This permits to express the interaction and adaptation service functions and to validate these functions defining the goals and causes attached to them.
Next, for each of the service functions is defined the characteristics of the external environment and the performances we expect for this function, giving a value criteria.
At the end, we need to identify the different participants who are concerned by the different situations of the product life cycle to formulate the external and internal constraints.
Formalize knowledge about design
To formalize knowledge about design, we also need the use of models.
The design process uses two different model, the product model and the activity model.
Product model is the way to structure the decisions taken during the design process, when activity model is the way to run the design process.
We already saw in previous lectures that integrated design is based on concurrent engineering and asks the different actors concerned with the life cycle of the product to intervene during the design process using the just need concept, in order to achieve the emergence of the solution.
All the constraints and design decisions have to be stored and will form the data model. We propose to formulate this data model using three kind of elements that we will named: the components, the links and the relations.
But this data model cannot be dissociate to the knowledge used to build it
and will refer to a knowledge model.
We propose to formulate our knowledge model using two concepts: the features and the production rules.
The feature model is used to formulate factual knowledge, when the production rules are used for temporal knowledge.
We have to remind that each of the participant of the design process has a specific profession and so must have his specific view on the project.
For that, each participant has to use a set of features and production rules that are particular to their job, what we called vernacular features or vernacular production rules. They are vernacular because they only can be understood by their own profession.
Of course these people can also use some features and production rules that are common to different profession. These features and production rules are qualified as vehicular. They can have an interest for different actors of the design process and serve in discussions and negotiations.
Some of the vehicular features or production rules are so well known by anybody that we call them universal.
Geometry procures by example universal features. Anybody knows what is a cylinder or a plan. It is why the geometry has been used as collaborative concept. In the same reason, as anybody can use geometry, the corresponding market is very huge and this explain why we have today Computer Aided Design systems that are in fact Geometrical Computer Aided systems.
A feature is defined with
-a generic name and is relative to a context or a domain. The same name in two different contexts can give two different features.
-a list of characteristics that can be typed characteristics. We use for that the classical types as character, integer or float numbers, but also specific types we have predefined as type point, vector, line, straight line, surface, plane, and skin. The surfaces are the geometrical surfaces, mathematically defined. The skins are the representative of real surfaces, with the reference geometrical surface, but also with dimension allowances or defaults.
-and sometime some behavioral descriptors: these descriptors give some relations between the characteristics, from the concerned feature or from the concerned feature and some other one.You can see on the slide some example of features: A flat part is used in technological view, has a name (a string of characters) and a medium surface which has to be a plane.
A graphic representation of the model starts with the main component, here an electric motor.
The associated links to this component are attached to a vertical line below the component. Output shaft, power supply, attachment… are characteristics of an electric motor and are considered here to define the specification of the wanted product.
The relations give specific values for each characteristics represented in the links. The output shaft has to support a torque of 100mN with a speed of 1450 revolutions per minute, and the cost of the motor has to be lower that 120$.
Multi-view decomposition concept
With this concept, we can not only add a decomposition diagram for a component, but we want to add different decomposition of the same component, each of them being specific to a view or a domain. Her we can explain how is composed the rotor sheet, depending of the frame view, with the skin features, or the manufacturing view, with manufacturing features, or of the material view, using some material features.
The interest is to get the possibility to put some relations between characteristics of different view, keeping the reasons for the choices done.
The manufacturing of a part in sheet metal is normally done by punching. It is the case here to obtain the boring and this gives a quality 9 on the quality scale (quite rough).
The skin nick needs a better quality as it is the surface where the magnetic field is coupling.
If we use the punching, the plastic deformation of the sheet during the cutting gives an orientation of the metal grains that avoids to get a magnetic field parallel to the medium surface of the rotor sheet. Using laser cutting for this surface gives a better quality of the surface (quality 6), that reduce the possible allowance between the rotor and the stator but also keeps a perfect orientation of the grains.
The choice of the speed feed of the laser depends of the wanted properties of the metal, that is given by the metallurgists.
Artificial Intelligence is a technique in order to permit the computer not only to solve equations but also to reason as an intelligent actor in order to solve problems or to give diagnosis.
Prolog, Frames, Production Rules, are the new languages used for the description of the expert systems.
New specialists in cognition were engaged in order to question the old engineers, to extract their knowledge and to build virtual experts in different fields.