|
FAQs
About MBS :
What
is Model-based Reasoning (MBR)?
Model
Based Reasoning is an inferring process using models abstracted
from the reality of a physical system. MBR is the symbolic
processing of an explicit representation of the internal
working of a system in order to predict, simulate and explain
the resultant behaviour of the system from the structure,
causality, functional and behaviour of its components.
What
are Model-based Systems?
Any
computer system developed by using Model Based Reasoning
techniques is broadly called a Model Based System. Computer
systems are often developed employing several techniques
to solve a realistic complex problem. A model based system
may consist of several models and techniques to cope with
a complex scheme.
Who
should use Model-based Systems and Qualitative Reasoning?
Model
Based Systems and Qualitative Reasoning can be used by engineers
who intend to design, analyse, simulate, diagnose, monitor
and maintain a physical technical system. These techniques
can significantly improve the efficiency of solving engineering
problems. This technique can also be applied to other application
domains such as education, ecology, biology and medicine.
What
are the main benefits of using Model-based Systems?
Model
Based Systems enable engineers to solve problems in a very
efficient manner. Models validated for a category of systems
can be used in many different scenarios. They can be reused
in the design, analysis, simulation, diagnosis and prediction
of a technical system. This ensures the quality and consistency
of a solution in carrying out the above tasks. In diagnostic
applications, model based systems play a vital role in identifying
the faults of a technical system and recommending a solution
for the rectification of faults. In design applications,
model based systems enable a designer to employ readily
available component models in the analysis and simulation
of the functional and behaviour performance of a system
design solution. This significantly improves the design
efficiency and quality of a system design.
What
are the application areas for Model-based Systems?
Model-based
Systems have been used in the following application domains:
Systems
Diagnosis:
Diagnostic tools for technical physical systems are used
to automate the diagnosis of such systems.
Systems
Functional/Behaviour Simulation:
Model-based function/behaviour prediction systems are
used to evaluate the behaviour of physical technical systems,
ecological systems and biological systems.
Product
Design and Synthesis:
Computer based design decision support systems such as
failure mode effects analysis systems.
Bond
Graphs are used to visually and qualitatively model a
product to facilitate the design solution generation process.
Environmental
decision support systems using MBSs.
Planning:
Based on the results of a diagnosis system, planning systems
are used to supported or automated generation of repair
actions.
Education:
Model Based Reasoning techniques are used for realising
tutoring and training functions, and subject matter construction.
Other
Areas:
Model Based Systems have also been seen in other application
area, including, medicine, industrial processes/plants,
automotive, aeronautics, aerospace, physics, telecommunications
etc.
What
are the limitations of current Model-based Systems?
There
are several limitations associated with Model Based Systems
and the key limitations include model validation, model
reuse for a new system, degree of model accuracy, level
of model complexity.
What
are the current research areas within MBS?
Well,
depending on your interest and your application domain,
there are many research areas on Model-based Systems itself
and the research areas oriented towards their application.
The following is a short list which, by no means, is exhaustive:
- Model
abstraction and approximation;
- Model
construction;
- Temporal
and behaviour abstraction;
- Model
validation;
- Model
reuse;
Most
detailed application related research areas are described
in What are the application
areas for Model-based Systems? (Information
current at end of MONET Project funded period.)
FAQs
About MBD :
What
is Model-based Diagnosis?
Model-based
diagnosis is one of the main application tasks for model-based
reasoning (model-based systems). It consists in diagnosing
a (physical) system starting from a (qualitative) model
of the system itself. It can exploit models of the structure,
function and/or behavior (normal or faulty) of the device
(system) to be diagnosed.
What
are the advantages of Model-based Diagnosis?
The
model-based approach to diagnosis has several advantages
over the traditional heuristic approach:
-
While heuristic systems are based on empirical knowledge
elicited from experts (and thus these knowledge bases
tend to be subjective), model based systems rely on
objective models of the system to be diagnosed. Models
can be obtained from design or from first-principle
knowledge (e.g., basic laws of physics).
-
Models are re-usable, heuristic knowledge is not. Two
different forms of re-usability can be considered:
-
A model of a device can be used for different tasks
(e.g., diagnosis, monitoring, design, simulation,
...);
-
A model of a component (part) of a device can be
re-used in the models of all devices including that
component ("no-function-in-structure" principle).
-
In model-based systems it is more natural to deal with
complex aspects of diagnostic problem solving such as
multiple faults, time-dependent behavior.
-
The amount of time needed to build and tune models is
lower than the amount of time needed for heuristic rules.
-
More accurate explanations of the results can be provided
to the user.
.
What
are the disadvantages of Model-based Diagnosis?
-
Model-based reasoning is more complex from the conceptual
and computational point of view.
-
Building models is a complex activity. Modeling involves
making choices, building abstractions, ... Thus, different
models can be built for a device, according to the modeling
choices and assumptions that are made. Different models
can be adequate for different tasks.
Are there applications of Model-based Diagnosis?
There
are several applications in areas such as: electronics,
aerospace, automotives, telecommunications, power plants
and power transmission, industrial processes, ...
FAQs
About QR :
What
are Qualitative Models?
Qualitative
Models aim to capture the fundamental behaviour of a system
or mechanism into a computer model, while suppressing much
of the detail. These models are defined by abstract, imprecise,
vague, and incomplete (qualitative) expressions/terminology.
Methods such as abstraction and approximation are often
used to build models based on qualitative rather numerical
aspects of a system.
What
is Qualitative Reasoning?
Qualitative
Reasoning is an inferring technique using qualitative models
to derive new knowledge of, and gain insight into a physical
technical system in a specific subject area. It can generate
an approximate solution to a given engineering problem.
It applies production systems in a qualitative manner and
uses other knowledge based methods such as constraint networks,
ontology etc during its reasoning process. This method is
an approach closer to a human being's thinking and reasoning.
What
are the main benefits of using QR?
In many
engineering problems, engineers don't normally have all
detailed information regarding the problem they are trying
to solve. This makes it impossible to construct a detailed
quantitative model to design, analyse, simulate, monitor
or diagnose a physical system. One of the main benefits
of QR is that it is possible to use an approach which provides
an approximate solution to a given problem when all necessary
detailed precise information is unavailable. Other benefits
include high solution generation efficiency due to its simple
model representation and ease of use as it uses human natural
language like expressions which can provide a more descriptive
solution.
Why
has Qualitative Reasoning been used in engineering problem
solving?
The
motivation for qualitative reasoning arose predominantly
from research on engineering problem solving, which sought
techniques for automating engineering practice for a variety
of important tasks. It quickly became clear that, if we
want to capture the skills of an engineer or technician
we must do far more than build bigger simulators or equation
solvers- the computer must somehow embody the common sense
of these experts. While these traditional tools are crucial,
much of an engineer's energy is devoted to problem formulation
- deciding when and how these tools are applied -and interpretation
- identifying the significant features of the analysis results
and evaluating their impact with respect to the task being
performed. Thus the heart of the qualitative reasoning enterprise
is to develop computational theories of the core skills
underlying engineers, scientists, and just plain folk's
ability to hypothesize, test, predict, create, optimize,
diagnose and debug physical mechanisms.
Taken
from: B. C. Williams, J. de Kleer: Qualitative reasoning
about physical systems: a return to roots. Artificial
Intelligence 51 (1991) 1-9
What
are the limitations of current QR techniques?
There
are several limitations associated with Qualitative Reasoning
and the key ones include difficulty in model construction,
and limited reasoning capability with low efficiency. The
construction of a qualitative model is, in itself, difficult
and requires careful consideration about the set of qualitative
descriptions of a problem.
Currently
the reasoning capability of qualitative techniques is
limited, the efficiency of these techniques are usually
low and they can take a while to reach a solution.
What
are the current application areas for Qualitative Reasoning?
Qualitative
Reasoning has been used in the following application domains:
Systems
Diagnosis:
Functional modelling using QR is a very important aspect
of systems diagnosis. An effective and efficient qualitative
model will enable a fast and correct behaviour diagnosis/prediction
of a physical system.
Systems
Functional/Behaviour Simulation:
QR can be used to coordinate the employment of multi-level
model simulation for physical systems.
QR
can be used during simulation to detect causality within
a system which highlights the behaviours from a product
development point of view and behaviour problems of the
system from diagnostic point of view.
Product
Design and Synthesis:
During a products conceptual design and synthesis, qualitative
reasoning can be of great use as typically at this stage
there exists incomplete, imprecise and vague information
about the product design requirements, the behaviour of
a current design solution, and the function the product
is intended to perform.
QR
can be used in qualitative constraints solving, vague
geometry/spatial reasoning and representation.
QR
can also be used to maintain the truth of multi-design
solution worlds.
Other
Areas:
Qualitative Reasoning has also been seen in other application
area, including, medicine, industrial processes/plants,
automotive, aeronautics, aerospace, physics, telecommunications
etc.
What
are the current research areas of QR?
Well,
depending on your interest and your application domain,
there are many research areas on Model Based Systems itself
and the research areas oriented towards their application.
The following is a short list which, by no means, is exhaustive:
- Qualitative
representation in application domains for the implementation
in computer systems;
- Mapping
between qualitative and quantitative representation
of problem definitions;
- New
approaches in qualitative representation;
- New
approaches in solving qualitative problems.
- Efficient
solution generation algorithms;
Most
detailed application related research areas are described
in What are the application
areas for Qualitative Reasoning? (Information
current at end of MONET Project funded period.)
Do
international organisations use MBS and QR?
Many
companies use MBS & QR techniques in their engineering problem
solving practice. These techniques have been successfully
used in their product development, within fault diagnosis,
behaviour simulation and system monitoring.
How
do I find more information about Model-based Systems and
Qualitative Reasoning?
There
are examples which describe how Model Based Systems and
Qualitative Reasoning are applied at the
MONET Technology page. You might like to visit these
pages to gain more knowledge about these technologies. These
systems should give you some ideas with the examples of
how they have been used successfully in the past. The two
tutorial examples are intended for introduction purposes
only so to find out more, please join MONET to get involved
in this important and evolving technology and enjoy all
the benefits the community offers.
FAQs
About MONET
What
was MONET?
MONET
was a Network of Excellence formed by Industrialists, Academics
and Researchers with a common goal of transfer of the technologies
of Model Based Systems & Qualitative Reasoning (MBS
& QR) into 'real world' applications in the Industrial
/ Public / Academic Sectors. MONET was funded by European
Community Research Framework Program 5 under the IST Program.
What
were the objectives of MONET?
The
MONET Network of Excellence provides a long-term framework
for technology transfer, research integration and co-operation
that will:
- Promote
technology transfer into industry
- Co-ordinate
European research in MBS & QR systems.
What
were the activities of MONET?
The
MONET Network of Excellence organised a number of activities
to fulfil its objectives. The activities included, but were
not limited to:
- Establishing
Web based information services to MBS and QR research
and application community.
- Aiding
the transfer of MBS & QR systems technology to industrial
applications.
- Organising
information concerning operational industrial applications
of MBS & QR systems.
- Organising
workshops, seminars and training courses to promote
MBS & QR technology.
- Providing
surveys on state of the art and identify research opportunities
and target applications.
- Increasing
the understanding and awareness of the key topics in
MBS & QR systems.
- Co-ordinating
European MBS & QR systems research and initiate
co-operative projects.
What
was the organisation of MONET?
MONET
was structured to be as efficient and smooth running as
possible. A full-time Network Co-ordinator and administrator
at Aberystwyth University of Wales dealt with day-to-day
running of the network and provided assistance in communications,
information and technical support for network members.
The
MONET network is composed of four Task Groups: Automotive,
Education and Training, Bio-Medical and Fault Detection
and Diagnosis (aka Bridge). The governing committee consists
of distinguished members of the European MBS & QR
community who are active in both the industrial and academic
sectors.
The
combination of committed staff, experienced industrialists
and eminent academics ensures that MONET provides a comprehensive
and professional resource for the network community.
What
were the benefits of joining the network?
It was
a simple, efficient and, best of all, FREE way to keep up
to date with this important area. The network is open to
potential end users as well as active researchers and so,
if you were interested in applying or just exploring MBS
& QR methods in your industry or field, you should join
MONET. Membership was FREE and provided access to considerable
expertise and information through the mutual interests of
the network community.
MONET
offers information, contacts, expertise, materials and
support. It maintains repositories of information such
as industrial case studies and state of the art reports,
and provided or support seminars and demonstrations. Channels
of communication were assisted in the form of newsletters,
events and electronic multimedia. Furthermore, MONET provided
financial support for workshops, meetings, technological
visits and scientific staff exchanges.
Who
should I contact to join the MONET Network?
The
Network is now no longer available to join, but you can
still Register to
use the MONET Information Resource.
|
