Robin_PHD/fmmd_concept/fmmd_concept.tex

\ifthenelse {\boolean{paper}}
{
\abstract{ 
This paper proposes a methodology for
creating failure mode models of safety critical systems, which
have a common notation
for mechanical, electronic and software domains and apply an
incremental and rigorous approach.

%% What I have done
%%
The Four main static failure mode analysis methodologies were examined and 
in the context of newer European safety standards, assessed.
Some of the defeciencies identified in these methodologies lead to
a wish list for a more ideal methodology.

%% What I have found
%%
From the wish list and considering some constraints determined from
the evaluation of the four established methodologies, a new
methodology is developed and proposed. The has been named Failure Mode Modular De-Composition (FMMD).

%% Sell it
%%
In addition to addressing the traditional weaknesses of
Fault Tree Analysis (FTA), Fault Mode Effects Analysis (FMEA), Failure Mode Effects Criticallity Analysis (FMECA)
and Failure Mode Effects and Diagnostic Analysis (FMEDA), FMMD provides the means to model multiple failure mode scenarios 
as specified in newer European Safety Standards \cite{en298}.
The proposed methodology is bottom-up and can guarantee to leave no component failure mode unhandled. 
It is also modular, meaning that the results of analysed components may be re-used in other projects.
}
}
{
 %%% CHAPTER INTO NEARLT THE SAME AS ABSTRACT

This chapter proposes a methodology for
creating failure mode models of safety critical systems, which
have a common notation
for mechanical, electronic and software domains and apply an
incremental and rigorous approach.

%% What I have done
%%
The Four main static failure mode analysis methodologies were examined and 
in the context of newer European safety standards, assessed.
Some of the defeciencies identified in these methodologies lead to
a wish list for a more ideal methodology.

%% What I have found
%%
From the wish list and considering some constraints determined from
the evaluation of the four established methodologies, a new
methodology is developed and proposed. The has been named Failure Mode Modular De-Composition (FMMD).

%% Sell it
%%
In addition to addressing the traditional weaknesses of
Fault Tree Analysis (FTA), Fault Mode Effects Analysis (FMEA), Failure Mode Effects Criticallity Analysis (FMECA)
and Failure Mode Effects and Diagnostic Analysis (FMEDA), FMMD provides the means to model multiple failure mode scenarios 
as specified in newer European Safety Standards \cite{en298}.
The proposed methodology is bottom-up and can guarantee to leave no component failure mode unhandled. 
It is also modular, meaning that the results of analysed components may be re-used in other projects.

}


 
\section{Current Static Failure Mode Methodologies}

There are four methodologies in common use for failure mode modelling.
These are FTA, FMEA, FMECA
and FMEDA (a form of statistical assessment).

These methodologies date from the 1940's onwards and have several draw backs and 
advantages that are discussed in the next section.
%In short
%FTA, due to its top down nature, can overlook error conditions. FMEA and the Statistical Methods
%lack precision in predicting failure modes at the SYSTEM level.


The Failure Mode Modular De-composition 
(FMMD) aims to address the 
weaknesses in these methodoligies and to add
features such as the ability to analyse double
failure mode scenarios, and to allow modular re-use
of analysis.

%FMMD is an incremental bottom up FMEA process. 
The FMMD
methodology presented here provides a more detailed and analytical
modelling system which will create a more complete and detailed hierarchical failure mode model from which
the data models from FTA, FMEA, FMECA and FMEDA (the statistical approach) can be 
derived if required. An FMMD model is therefore a super set of all these models.
It also applies rigorous checking in all the analysis stages
ensuring that all component failure modes must be considered in the model.

%
This methodology has been named Failure Mode Modular De-composition (FMMD)
because it de-composes a SYSTEM into a hierarchy of modules or {\dc}s.
This
\ifthenelse {\boolean{paper}}
{
paper
}
{
chapter
}
presents the design considerations that determined
the FMMD methodology.
It first briefly reviews the four traditional
static failure mode analysis methodologies and 
lists their known weaknesses. A wish list is then drawn up
addressing these weaknesses and adding some extra requirements.
Using this wish list the philosophy for the new methodology
is built up.
%
FMMD works by working from the bottom up, taking small groups
of components, {\fgs}, and then analysing how they can fail.
This analysis is performed using FMEA from a micro rather than a macro perspective.
Thus instead of looking at component failure modes and determining how
they {\em may} cause  a failure at SYSTEM level, we are looking at how
they {\em will} affect the {\fg}.
When we know the failure modes of a {\fg} we can treat it as a `black box'
or {\dc}. With {\dc}s we can build {\fgs} 
at higher levels of analysis, until we have a complete
hierarchy representing the failure behaviour of the SYSTEM.
%
Because all the failure modes of all the components
are held in a computer program, we can determine if the model is complete
(i.e. all component failure modes have been included in the model).


%OK need to describe the need for it
\section{The need for a new failure mode modelling methodology}

%%- There are dificulties with bot up methodologies,
%%- and this is in part due to the fact that accidents 
%%- are always unforseen and unexpected.

%%- what do we have ENV factors, component failure modes.

%%- how difficult is it to take a single component failure mode and
%%- then from that determine how it will react with other components
%%- and how it will be affected 

\subsection{General Comments on bottom-up and top down approaches}

\paragraph{A general defeciency in top-down systems analysis}
With a top down approach the investigator has to determine
a set of undesirable outcomes or accidents.
As most accidents are unexpected and the causes unforseen \cite{safeware} 
it is fair to say that a top down approach is not guaranteed to
predict all possible undesirable outcomes.
It also can miss known component failure modes, by
simply not de-composing down to the base component failure mode level of detail.

\paragraph{A general problem with bottom-up}
With the bottom up techniques we have all the known component failure modes
and the freedom to determine how each of these may affect the SYSTEM.
We do have a real prolem though in determining how
the failure mode of one component will affect another working component
to cause an undesirable state. Because of the number of components
our one failure mode may interact with is large, 
we cannot consider them all and human judgement is used to 
decide which interactions are important.

Let N be the number of components in our system, and K be the average number of component failure modes
(ways in which the component can fail). The total number of base component failure modes
is $N \times K$. To examine the effect that one failure mode has on all the other components
will be $(N-1) \times N \times K$, in effect a set cross product.


Complicate this further with applied states or environmental conditions
and another order of cross product of complexity is added.
We may have a piece of self checking circuitry for instance that
has two states, normal and testing mode commanded by a logic line.
Or we may have a mechanical device that has a different 
failure mode behaviour for say, different ambient pressures or temperatures.

If $E$ is the number of applied states or environmental conditions to consider
in a system, the job of the bottom-up analyst is complicated by a cross product factor again 
$(N-1) \times N \times K \times E$.
If we put some typical very small embedded system numbers\footnote{these figures would
be typical of a very simple temperature controller, with a micro-controller sensor and heater circuit} into this, say $N=100$, $K=2.5$ and $E=10$
we have $99 \times 100 \times 2.5 \times 10 = 247500 $.
To look in detail at a quarter of a million test cases is obviously impractical.

If we were to consider multiple simultaneous failure modes,
we have yet another complication cross product.

For instance for looking at double simultaneous failure modes, 
the equation reads $(N-2) \times (N-1) \times N \times K \times E$.

The bottom-up methodologies FMEA, FMECA and FMEDA take single failure modes and link them
to SYSTEM level failure modes. Because of the astronomical number of possible interactions,
some valid ones are in danger of being missed, we can term this analysis a `leap of faith' from the 
component failure mode to the SYSTEM level.



\paragraph{Ideal static failure mode methodology}
An ideal static failure mode methodology would build a failure mode model
from which the traditional four models could be derived.
It would address the short-comings in the other methodologies, and
would have a user friendly interface, with a visual (rather than mathematical/formal) syntax with icons
to represent the results of analysis phases.
%
%There are four static analysis failure mode methodologies in common use.
%Each has its advantages and drawbacks, and each is suited for 
%a different phase in the product life cycle.
The four methodologies in current use are discussed briefly below.

\subsection { FTA }

This, like all top~down methodologies introduces the very serious problem
of missing component failure modes \cite{faa}[Ch.9].
%, or modelling at
%a too high level of failure mode abstraction.
FTA was invented for use on the minuteman nuclear defence missile
systems in the early 1960s and was not designed as a rigorous
fault/failure mode methodology. 
It was designed to look for disasterous top level hazards and 
determine how they could be caused.
It is more like a structure to
be applied when discussing the safety of a system, with a top down hierarchical
notation using logic symbols, that guides the analysis. 
This methodology was designed for
experienced engineers sitting around a large diagram and discussing the safety aspects.
Also the nature of a large rocket with red wire, and remote detonation
failsafes meant that the objective was to iron out common failures
not to rigorously detect all possible failures.
Consequently it was not designed to guarantee to cover all component failure modes, 
and has no rigorous in-built safeguards to ensure coverage of all possible
system level outcomes.

\subsubsection{ FTA weaknesses }
\begin{itemize}
\item Possibility to miss component failure modes
\item Possibility to miss environmetal affects.
\item No possibility to model base component level double failure modes.
\end{itemize}

\subsection { FMEA }

\label{pfmea}
This is an early static analysis methodology, and concentrates
on SYSTEM level errors which have been investigated.
The investigation will typically point to a particular failure
of a component. 
The methodology is now applied to find the significance of the failure.
Its is based on a simple equation where $S$ ranks the severity (or cost \cite{bfmea}) of the identified SYSTEM failure,
$O$ its occurance, and $D$ giving the failures detectability. Muliplying these
together, 
gives a risk probability number (RPN), given by $RPN = S \times O \times D$.
This gives in effect
a prioritised `todo list', with higher the $RPN$ values being the most urgent.


\subsubsection{ FMEA weaknesses }
\begin{itemize}
\item Possibility to miss the effects of failure modes at SYSTEM level.
\item Possibility to miss environemtal affects.
\item No possibility to model base component level double failure modes.
\end{itemize}

\paragraph{note.} FMEA is sometimes used in its literal sense, that is to say
failure Mode effects Analysis, simply looking at a systems internal failure
modes and determing what may happen as a result.
FMEA described in this section (\ref{pfmea}) is sometimes called `production FMEA'.

\subsection{FMECA}

Failure mode, effects, and criticality analysis (FMECDA) extends FMEA.
This is a bottom up methodology, which takes component failure modes
and traces them to the SYSTEM level failures. 
%
Reliability data for components is used to predict the 
failure statistics in the design stage.
A openly published source for the reliability of generic
electronic components was published by the DOD
in 1991 (MIL HDK 1991 \cite{mil1991}) and is a typical 
source for MTFF data.
%
It can do this using probability \footnote{for a given component failure mode there will be a $\beta$ value, the
probability that the component failure mode will cause a given SYSTEM failure}.
%
This lacks precision, or in other words, determinability prediction accuracy \cite{fafmea}, 
as often the component failure mode cannot be proven to cause a SYSTEM level failure, but
assigned a probability $\beta$ fator by the design engineer.
%Also, it can miss combinations of failure modes that will cause SYSTEM level errors.
%
The results, as with FMEA are an $RPN$ number determining the significance of the SYSTEM fault.

%%-WIKI- Failure mode, effects, and criticality analysis (FMECA) is an extension of failure mode and effects analysis (FMEA).
%%-WIKI- FMEA is a a bottom-up, inductive analytical method which may be performed at either the functional or 
%%-WIKI- piece-part level. FMECA extends FMEA by including a criticality analysis, which is used to chart the 
%%-WIKI- probability  of failure modes against the severity of their consequences. The result highlights failure modes with relatively high probability 
%%-WIKI- and severity of consequences, allowing remedial effort to be directed where it will produce the greatest value. 
%%-WIKI- FMECA tends to be preferred over FMEA in space and North Atlantic Treaty Organization (NATO) military applications, 
%%-WIKI- while various forms of FMEA predominate in other industries.


\subsubsection{ FMECA weaknesses }
\begin{itemize}
\item Possibility to miss the effects of failure modes at SYSTEM level.
\item Possibility to miss environmental affects.
\item No possibility to model base component level double failure modes.
\end{itemize}


\subsection { FMEDA or Statistical Analyis }

Failure Modes, Effects, and Diagnostic Analysis (FMEDA).

This is a process that takes all the components in a system,
and from the failure modes of those components, the investigating engineer
must tie them to possible SYSTEM level events/failure modes.
This technique 
evaluates a products statistical level of safety
taking into account its self-diagnostic ability.
The calculations and procedures for FMEDA are
described in EN61508 Part 2 Appendix C \cite{en61508}[Part 2 App C].
The following gives an outline of the procedure.


\subsubsection{Two statistical perspectives}
he Statistical Analysis method is used from two perspectives,
Probability of Failure on Demand (PFD), and Probability of Failure
in continuous Operation, Failure in Time (FIT).
\paragraph{Failure in Time (FIT)}.

Continuous operation is measured in failures per billion ($10^9$) hours of operation.
For a continuously running nuclear powerstation
we would be interested in its operational FIT values. 

\paragraph{Probability of Failure on Demand (PFD)}.
For instance with the anti-lock system on a automobile braking
system, we would be interested in PFD.
That is to say the ratio of it failing 
to succeeding on demand.

\subsubsection{The FMEDA Analysis Process}

\paragraph{Determine SYSTEM level failures from base components}
The first stage is to apply FMEA to the SYSTEM.
%
Each component is analysed in terms of how its failure
would affact the system.%
Failure rates of individual components in the SYSTEM
are calculated based on component type and
environmental conditions.
%
Statistical data exists for most component types \cite{mil1992}.
%
This phase is typically implemented on a spreadsheet. Along with a components
type, placing in the system, part number, environmental stress factors etc.
%will be a determination of whether the component failing will lead to a `safe'
%or `unsafe' condition.

\paragraph{Overall SYSTEM failure rate.}
Product failure rate is the sum of all component
failure rates. This is the sum of safe and unsafe
failures.

\paragraph{Self Diagnostics}
We  next evaluate the SYSTEMS’s self-diagnostic ability.

Each component’s failure modes and failure rate are now available.
Failure modes are now classified  as safe or dangerous.
This is done by taking a component failure mode and determining
how it will react with any other components in the SYSTEM and taking a decision
based on hueristics.
Detectable failure probabilities are labelled `$\lambda_D$' (for
dangerous) and  `$\lambda_S$' (for safe) \cite{EN61508}.

\paragraph{Determine Detectable and Undetecable Failures}
Each safe and dangerous failure mode is now
determined as detectable or un-detectable by the SYSTEMS’s 
self checking features.
%
This gives us four failure mode classifications:
Safe-Detected (SD), Safe-Undetected (SU), Dangerous-Detected (DD) or Dangerous-Undetected (DU),
and the failure rate of each classification 
is represented by lambda variables
($\lambda_{SD}$, $\lambda_{SU}$, $\lambda_{DD}$, $\lambda_{DU}$).

Because some failure modes may  not be discovered theoretically during the static
analysis, the
% admission of how daft it is to take a component failure mode on its own
% and guess how it will affect an ENTIRE complex SYSTEM 
next step  is to investigate using an actual working SYSTEM. 

Failures are deliberately caused (by physical intervetion), and any new SYSTEM level 
failures are added to the model.
Hueristics and MTTF failure rate for the components
are used to calculate probabilities for these new failure modes 
according to their saefty and detectability classifications (i.e.
$\lambda_{SD}$, $\lambda_{SU}$, $\lambda_{DD}$, $\lambda_{DU}$).
These new failures are added to the model.
%SD, SU, DD, DU.

With these classifications, and statistics for each component
we can now calculate statistics for the diagnostic coverage (how good at `self checking' the system is)
and its safe failure fraction (how many of its failures are self detected or safe compred to
all failures possible).

The calculations for these are described below.

\paragraph{Diagnostic Coverage.}
The diagnostic coverage is simply the ratio
of the dangerous detected probabilities
against the probability of all dangerous failures,
and is normally expressed as a percentage.

$$ DiagnosticCoverage = \Sigma\lambda_{DD} / \Sigma\lambda_D $$

The diagnostic coverage for safe failures is given as

$$ SF = \frac{\Sigma\lambda_SD}{\Sigma\lambda_S} $$


\paragraph{Safe Failure Fraction.}
A key concept in  FMEDA is Safe Failure Fraction (SFF).
This is the ratio of safe  and dangerous detected failures
against the safe and dangerous failure probabilities. 
Again this is usually expressed as a percentage.

$$ SFF = \big( \Sigma\lambda_S + \Sigma\lambda_{DD} \big) / \big( \Sigma\lambda_S + \Sigma\lambda_D \big) $$

%This is the ratio of 
%Step 4 Calculate SFF, SIL and PFD
%The SIL level of the product is finally determined from the Safe Failure Fraction (SFF) and the Probability of Failure on Demand (PFD). The following formulas are used.
%SFF = (lSD + lSU + lDD) / (lSD + lSU + lDD + lDU)
%PFD = (lDU)(Proof Test Interval)/2 + (lDD)(Down Time or Repair Time)

% Often a given component failure mode there will be a $\beta$ value, the
% probability that the component failure mode will cause a given SYSTEM failure.

%\paragraph{Risk Mitigation}
%
%The component may be have its risk factor 
%reduced by the checking interval (or $\tau$ time between self checking procedures).
%
%Ultimately this technique calculates a risk factor for each component.
%The risk factors of all the components are summed and 
%%give a value for the `safety level' for the equipment in a given environment.





\paragraph{Classification into Safety Integrity Levels (SIL).}
There are four SIL levels, from 1 to 4 with 4 being the highest safety level.
In addition to probablistic risk factors, the 
diagnostic coverage and SFF
have threshold bands beoming stricter for each level.
Software techniques and constraints are 
also become stricter for each SIL level.

FMEDA uses MTFF and other statistical models to determine the probability of
failures occurring, and provide an adaquate risk level.
%
%A component failure mode, given its MTTF
%the probability of detecting the fault and its safety relevant validation time $\tau$,
%contributes a simple risk factor that is summed
%in to give a final risk result. 
%
Thus a statistical
model can be implemented on a spreadsheet, where each component
has a calculated risk, a fault detection time (if any), an estimated risk importance
and other factors such as de-rating and environmental stress.
With one component failure mode per row, 
all the statistical factors for SIL rating can be produced.






\subsubsection{FMEDA and failure outcome prediction accuracy.}
This suffers from the same problems of 
lack of component failure mode outcome prediction accuracy, as FMEA in section \ref{pfmea}.
%
This is because the analyst has to decide how particular components failing will impact on the SYSTEM or top level.
This involves a `leap of faith'. For instance, a resistor failing in a sensor circuit
may be part of a critical monitoring function. 
The analyst is now put in a position
where he probably should assign a dangerous failure classification to it.  
%
There is no analysis
of how that resistor would/could  affect the components close to it, but because the circuitry
it is part of critical section it will most likely
be linked to a dangerous system level failure in an FMEDA study.
%
%%- IS THIS TRUE IS THERE A BETA FACTOR IN FMEDA????
%%-
%A $\beta$ factor, the hueristically defined probability
%of the failure causing the system fault may be applied.
%
But because there is no detailed analysis of the failure mode behaviour
of the component in its local environment
but traceable directly to the SYSTEM level, it becomes more
guess work than science.
%
With FMEDA, there is no rigorous cause and effect analysis for the failure modes
and how they interact on the micro scale (the components adjacent to them in terms of functionality). 
Unintended side effects that lead to failure can be missed. 
Also component failure modes that are not
dangerous, may be wrongly assigned as dangerous simply because they exist in a critical
section of the product.

% some critical component failure 
%modes, but we can only guess, in most cases what the safety case outcome
%will be if it occurs.

This leads to the practise of having components within a SYSTEM partitioned into different 
safety level zones as recomended in EN61508\cite{en61508}. This is a vague way of determining
safety, as it can miss unexpected effects due to `unexpected' component interaction.

The Statistical Analysis methodology is the core philosophy
of the Safety Integrity Levels (SIL) ebodied in EN61508 \cite{en61508}
and its international analog standard IOC5108.



\subsubsection{ FMEDA weaknesses }
\begin{itemize}
\item Possibility to miss the effects of failure modes at SYSTEM level.
\item Statistical nature allows a proportion of undetected failures  for given S.I.L. level.
\item Allows a small proportion of `undetectable' error conditions.
\item No possibility to model base component level double failure modes.
\end{itemize}
%AND then how we can solve all there problems

\section{A wish list for a failure mode methodolgy}
\begin{itemize}
\item All component failure modes must be considered in the model.
\item It should be easy to integrate mechanical, electronic and software models \cite{sccs}[pp.287].
\item It should be re-usable, in that commonly used modules can be re-used in other designs/projects.
\item It should have a formal basis, that is to say, it should be able to produce mathematical proofs
for its results, such as system level error causation trees, reliability and safety statistics.
\item It should be easy to use, ideally using a graphical syntax (as oppossed to a formal mathematical one).
\item From the top down, the failure mode model should follow a logical de-composition of the functionality
to smaller and smaller functional modules \cite{maikowski}.
\item Multiple failure modes may be modelled from the base component level up.
\end{itemize}


\section{Design of a new static failure mode based methodology}

\paragraph{New methodology must be bottom-up}
In order to ensure that all component failure modes have been covered
the methodology will have to work from the bottom-up
and start with the component failure modes.
%
\paragraph{Natural Fault Finding is top down}
The traditional fault finding, or natural fault finding
is to work from the top down. 
%
On encountering a 
fault, the symptom is first observed at the top or
SYSTEM level. By de-composing the functionality of the faulty system and testing
we can further de-compose the system until we find the
faulty base level component.
De-composition of electrical circuits is formalised and explored
in \cite{maikowski}. This top down technique de-composes by functionality.
Simpler and simpler functional blocks are discovered as we delve
further into the way the system works and is built.


\paragraph{Need for a `bottom-up' system de-composition}
There is an apparent conflict here. The natural way to 
de-compose a system is from the top down.
%
If we do this though, we do not naturally include
all failure modes in the modules determined as we
de-compose downwards.
%
What is required here is to mimic this top-down de-composition
with a bottom up technique.
By doing that, we can take all base component failure modes
and ensure they are included in the model.

By taking components that form {\fg}s from the bottom up
and then taking those to form higher level
{\fg}s we can get a close approximation of the de-composition process from the bottom up.
The philosophy of top down de-compositon is very similar.
Top down de-compositon applies functional 
de-composition, because it seeks to break the system down
into manageable and separately testable entities.
A second justification for this is that the design process for a product requires both top down and bottom-up
thinking. To analyse a system from the bottom-up is a useful
design validatio process in itself \cite{sommerville}.

\paragraph{Design Decision: Methodology must be bottom-up.}
In order to ensure that all component failure modes are handled,
this methodology must start at the bottom, with base component failure modes.
In this way automated checking can be applied to all component failure modes
to ensure none have been inadvertently excluded from the process.

\paragraph{Problem with functional group hierarchy}
A hierarchy of functional grouping, leading to a system model
still leaves us with the problem of the number of component failure modes.
The base components will typically have several failure modes each.
%
Given a typical embedded system may have hundreds of components
This means that we have to tie base component failure modes
to SYSTEM level errors. This is the `possibility to miss failure mode effects
at SYSTEM level' criticism of the FTA, FMEDA and FMECA methodologies.

\paragraph{Design Decision: Methodolgy must reduce and collate errors at each functional group stage.}
SYSTEMS typically have far fewer failure modes than the sum of their component failure modes.
SYSTEM level failures may be caused by a variety of component failure modes.
A SYSTEM level failure mode is an abstracted failure mode, in that
it is a symptom of some lower level failure or failures.
% ABSTRACTION
For instance a failed resistor in a sensor at a base component level is a specific 
failure mode. 
%
For example it could be called `RESISTOR 1 OPEN'. 
%
Now consider the  symptom in a functional group comprising the sensor channel that 
RESISTOR 1 is part of of `RESISTOR 1 OPEN'.
%
We might call it `READING~HIGH' failure perhaps. 
The Fault has become less detailed and more general. There may be other 
causes for a `READING~HIGH'. We can say that the failure
mode `READING~HIGH' is more abstract in terms of the STSEM, than `RESISTOR 1 OPEN'.
%
At a higher level still
this may be called `SENSOR CHANNEL 1' fault.
At a system level it may simply be a `SENSOR FAILURE'.
As we traverse up the fault tree the failure modes
become more abstract. 
%
At each functional group collection, there must be a process to collect
common symptoms and reduce the number of failure modes to handle.
This must be a process that incrementally reduces the number
of failure modes as the abstraction level reaches the SYSTEM level.

\paragraph{How to build a meaningful SYSTEM failure behaviour model.}
The next problem is how to we build a failure mode model
that converges to a finite set of SYSTEM level failure modes.
%
It would be better to analyse the failure mode behaviour of each
functional group, and determine the ways in which it, rather than its
components, can fail.
% 
By doing this, the natural process whereby symptoms of the {\fg},
which can potentially be caused by more then one 
component failure mode, become the target for reducing the number
of failure modes to handle as we traverse up the hierarchy.


\paragraph{Component failures and {\fg} failure symptoms.}
In other words we want to find out what the symptoms of the failures in the {\fg}s
are. 
The number of symptoms of failure should be equal to or
less than the number of component failure modes, simply because
often there are several potential causes of failure symptoms.
%
When we have the symptoms, we can start thinking of the {\fg} as a component in its own right.
%with a simplified and reduced set of failure symptoms.
%
We can now create a new {\dc}, where its failure modes
are the failure symptoms of the {\fg}.
In this way as we build the hierarchy, we naturally abstract the 
failure mode behaviour, but can check that all failure modes in 
the hierarchy have been considered and tied to causing symptoms.


\paragraph{Incremental Stages and \dcs .}
We can use incremental stages to build the hierarchy.
We can take small {\fg}s of components, where the {\fg}
is a small set of components that perform a simple
task.
%
This should be small enough to be able to consider all the failure
modes of its components.
%
We can consider these failure modes from the perspective
of the {\fg}. In other words, for each component failure mode in the {\fg},
we create a `test case' and decide how each failure affects the functional group.
%
With the results from the test cases we will now have the ways in which the 
{\fg} can fail.
%
%
We can refine this further, by grouping the common symptoms, or results that
are the same failure w.r.t. the {\fg}.
%
We can now treat the {\fg} as a component, and call it a {\dc}, in other words, a sub-system with a known set of failure modes.
%
We can now create a new/{\dc} and assign it these common symptoms
as its failure modes.
%
This {\dc} can be used to build higher level
{\fg}s, and this will naturally form a hierarchy.
This hierarchy can be extended until it encompasses 
an entire system. It can be considered complete when
all failure modes from all components are handled
and connectable to a SYSTEM level failure mode.

\paragraph{Directed Acyclic Graph (DAG).} 
The data structure produced from collecting functional groups
and deriving components will naturally form a DAG.
To ensure this we will have to ensure that
derived components cannot be included in {\fg}s
of a lower abstraction level.
%
%
By representing the failure mode model as a DAG, we
now have the capability to take SYSTEM level failure modes
and determine the possible combinations of component failure modes that
could have caused it.
In FTA terminology, a list of possible
causes for a SYSTEM level failure is known as a minimal cut set \cite{nasafta}.
%
%
If statistical models exist for the component failure modes
these failure causation trees (or minimal cut sets \cite{nucfta})
can be used to calculate Mean Time to Failure (MTTF) or Probability of Failure on demand (PFD) figures.
Contrast the analytical capability of FMMD with the 
methodologies where the component failure modes are linked
directly to SYSTEM failure modes with no analysis stages in between.



\paragraph{Design Decision: A functional group cannot contain {\dc}s at a higher abstraction level than its self}

We can implement this rule in two ways, firstly, by saying that any functional group
will take the `abstraction level + 1' of all components it includes.
Secondly we can say that no component may be derived from itself.

\paragraph{Natural Reduction in number of failure modes with abstraction level}
%
Because common symptoms are being collected, as we build the tree up-ward
the number of failure modes decreases (or exceptionally stays the same) at each level.
%
This decreasing of the number of failure modes is bourne out {\irl}.
Of the thousands of component failure modes in a typical product
there are generally only a handful of SYSTEM level failure modes
(or top level `symptoms' of underlying failures). 
%

\subsection{Outline of the FMMD process}
\label{fmmdproc}
FMMD builds {\fg}s of components from the bottom-up. 
The lowest level of components are termed base components.
These are the initial building blocks.
In Electronics these would be the individual
passive and active components on the parts~list.
In mechanics the the levers springs cogs etc.
Functional groups are collections of components
that work together to perform a simple function.
%
We can perform a failure mode effects analysis on each of the component failure
modes within the {\fg}. We can thus ensure that all component failure modes 
are covered. 
%
We can then treat the {\fg} as a `black box' or component in its own right.
We can now look at how the {\fg} can fail. 
%
Many of the component failure modes will
cause the same failure symptoms in the {\fg} failure behaviour.
We can collect these failures as common symptoms.
%
When we have our set of symptoms, we can now create
a {\dc}. The {\dc} will have as its set of failures
modes, the collected symptoms of the {\fg}.
%
Because we can now have {\dcs} we can use these to form
new {\fg}s and we can build a hierarchical `failure~mode' model of the SYSTEM.

The diagram in figure \ref{fig:fmmd_hierachy}, shows one stage
of the FMMD process. The resultant {\dc} may be used to
create higher level {\fg}s in later stages.
%%- Need diagram of hierarchy
%%-
%%-
\begin{figure}[h]
 \centering
 \includegraphics[width=200pt,bb=0 0 331 249,keepaspectratio=true]{./fmmd_concept/fmmd_hierarchy.jpg}
 % fmmd_hierarchy.jpg: 331x249 pixel, 72dpi, 11.68x8.78 cm, bb=0 0 331 249
 \caption{Example derived component created from the functional group comprised of components a,b,c}
 \label{fig:fmmd_hierarchy}
\end{figure}


% \begin{figure}[h]
%  \centering
%  \includegraphics[bb=0 0 331 249,keepaspectratio=true]{./fmmd_hierarchy.jpg}
%  % fmmd_hierarchy.jpg: 331x249 pixel, 72dpi, 11.68x8.78 cm, bb=0 0 331 249
%  \caption{Example derived component created from a functional group comprised of components a,b,c}
%  \label{fig:fmmd_hiarchy}
% \end{figure}
% 
% \vspace{20pt}
% NEED DIAGRAM OF HIERACY
% \vspace{20pt}


\subsection{Environmental Conditoions, Operational States and FMMD}

Any real world sub-system will exist in a variable environment and may have several modes of operation.
In order to find all possible failures, the sub-system must be analysed for each operational state
and environment condition that can affect it.
Two design decision are required here, which objects should we 
analyse the environment and operational states with respect to.
We could apply these conditions for analysis 
to the functional group, the components, or the derived
component.

\paragraph {Environmental Conditions and FMMD.}

Environmental conditions are external to the
{\fg} and are often things that the system has no direct control over.
Consider ambient temperature, pressure or even electrical interferrence levels.

Environmental conditions may affect different components in a {\fg}
in different ways.

For instance a system may be specified for
0 to 85oC operation, but some components
may show failure behaviour between 60 and 85
\footnote{Opto-islolators typically show marked performace decrease after
60oC whereas another common component, the resistor will be unaffected.}.
Environmental conditions will have an effect on the {\fg} and the {\dc}
but they will have specific effects on individual components.

\paragraph{Design Decision.}
Environmental constraints will be applied to components.
A component will hold a set of Environmental states that
affect it. 
Environmental conditions will apply SYSTEM wide,
but may only affect specific components.
%Some may not be required for consideration
%for the analysis of particular systems.

\paragraph {Operational States and FMMD}

Sub-systems may have specific operational states.
These could be a general health level such as 
normal operation, graceful degradation or lockout.
Or they could be self~checking sub-systems that are either in a normal or self~check state.

Operational states are conditions that apply to a functional group, not individual components.


\paragraph{Design Decision.}
Operational state  will be applied to {\fg}s.

\paragraph{UML Model of FMMD Analysis}

Draw a UML model showing the components and the functional group
with the ENV and OP\_STAT classes associated with them


\begin{figure}[h]
 \centering
 \includegraphics[width=400pt,bb=0 0 818 249,keepaspectratio=true]{./fmmd_concept/fmmd_env_op_uml.jpg}
 % fmmd_env_op_uml.jpg: 818x249 pixel, 72dpi, 28.86x8.78 cm, bb=0 0 818 249
 \caption{UML model of Environmental and Operational states w.r.t FMMD}
 \label{fig:env_op_uml}
\end{figure}



\subsection{Justification of wishlist}

By applying the methodology in section \ref{fmmdproc}, the wishlist can
now be evaluated for the proposed FMMD methodology.

\subsubsection{All component failure modes must be considered in the model.}
The proposed methodology will be bottom-up. 
This ensures that all component failure modes are handled.


\subsubsection{ It should be easy to integrate mechanical, electronic and software models.}
Because component failure modes are considered, we have a generic entity to model.
We can describe a mechanical, electrical or software component in terms of its failure modes. 
%
Because of this 
we can model and analyse integrated electro mechanical systems, controlled by computers,
using a common notation.

\subsubsection{ It should be re-usable, in that commonly used modules can be re-used in other designs/projects.}
The hierarchical nature, taking {\fg}s and deriving components from them, means that 
commonly used {\dcs} can be re-used in a design (for instance self checking digital inputs)
or even in other projects where the same {\dc} is used.



\subsubsection{ It should have a formal basis, that is to say, it should be able to produce mathematical proofs
for its results}
Because the failure mode of a SYSTEM is a hierarchy of {\fg}s and derived components
SYSTEM level failure modes are traceable back down the tree to
component level failure modes. This provides causation trees \cite{sccs} or, minimal cut sets
\footnote{Here minimal cut sets represent combinations of component failure modes that can result in s SYSTEM level failure.} 
for all SYSTEM failure modes.

\subsubsection{ It should be capable of producing reliability and danger evaluation statistics.}
The Minimal cuts sets for the SYSTEM level failures can have computed MTTF
and danger evaluation statistics sourced from the component failure mode statistics \cite {mil1991}.

\subsubsection{ It should be easy to use, ideally using a graphical syntax (as oppossed to a formal mathematical one).}
A modified form of constraint diagram (an extension of Euler diagrams) has been developed to support the FMMD methodology.
This uses Euler circles to represent failure modes, and spiders to collect symptoms, to
advance a {\fg} to a {\dc}.


\subsubsection{ From the top down the failure mode model should follow a logical de-composition of the functionality
to smaller and smaller functional modules \cite{maikowski}.}
The bottom-up approach fulfils the  logical de-composition requirement, because the {\fg}s
are built from components performing a given task. 


\subsubsection{ Multiple failure modes may be modelled from the base component level up}
By breaking the problem of failure mode analysis into small stages
and building a hierarchy, the problems associated with the cross products of
all failure modes within a system are reduced by an exponential order.



\subsection{Advantages of FMMD Methodology}

\begin{itemize}
\item It can be checked automatically that all component failure modes have been considered in the model.
\item Because we are modelling with failure modes the {\fgs} and {\dcs} these can be generic, i.e.  mechanical, electronic or software components.
\item The {\dcs} are re-usable, in that commonly used modules can be re-used in other designs/projects.
\item It will have a formal basis, that is to say, it is able to produce mathematical proofs
for its results (MTTF and the cause trees for SYSTEM level faults).
\item Overall reliability and danger evaluation statistics can be computed. 
By knowing all causation trees,
the statistical probabilities (from base component data) for all causes can be simply added.
\item A graphical representation based on Euler diagrams is used. Providing an interface that does not involve
formal mathematical notation. This is intended to be user friendly and to guide the user through the FMMD process
while applying automatic checks for unhandled conditions.
\item From the top down the failure mode model will follow a logical de-composition of the functionality; by
chosing {\fg}s and working bottom-up this hierarchical trait will occur as a natural consequence.
\item Undetectable or unhandled failure modes will be specifically flagged.
\item It is possible to model multiple failure modes.
\end{itemize}

\section{Conclusion}

This paper provides the background for the need for a new methodology for
static analysis that can span the mechanical electrical and software domains
using a common notation.
The author believes it addresses mant short comings in current static failure mode analysis methodologies.
\vspace{60pt}
\today