Robustness
is a key fundamental element in APQP and FMEA product and service campaigns.
It is the ability of the engineering or development team to have traceable
product and service characteristics move through the product and service
development cycle to ensure the customer receives the product or service as
they desire. Customers would include end consumers and industry or government
regulatory bodies. Customers can also include internal downstream processes in
the development cycle - at a minimum they would receive the characteristics
requirements of the product or service from earlier sections or phases in the
product development process.
Robustness
occurs both vertically and horizontally in the product and service development
cycle. Vertically within the product, subsystem, component structure through
the cause failure mode relationship in the DFMEA and PFMEA (the cause of
failure at the system level becomes the failure mode at the subsystem level) and
horizontally as the different phases development are entered. Noise factors in
a Parameter Diagram as causes of failure in the DFMEA for example.
Cause
of Failure at one level is the failure mode at the level below.

Noises
identified in the P-Diagram become Causes in the DFMEA
The
focus of this article will be on the robustness of the effects column in the DFMEA and PFMEA. There are several areas of focus for robustness in the
product and service development process in particular the transition from DFMEA to PFMEA amongst others. We will save those for future articles. The
effects column in the 4th Edition FMEA is developed with the
Function, Failure Mode, Effects and Severity section in the FMEA and in the
new AIAG/VDA Handbook is in the Failure Analysis (Step 4) section.
Sound
robustness is one of the forerunners to developing and moving toward integrated Industry4.0 and Quality4.0 organisations. Connectivity, availability and
transparency of information, intelligent processing and interaction of data
along with the digitisation and integration of vertical and horizontal value
chains are intrinsically linked to the robustness level and organisation can
produce in its product and service development models and processes. A highly
robust product and service development process connects and integrates all its
tools (QFD/VOC, DFMEA, DVP&R, Characteristics Matrix, PFMEA, DFA/DFM, MSA,
PCP and SPC) allowing them to interact and transparently follow
characteristics, in particular critical and significant characteristics through
the process.
Robustness
is the basic building block into EIMS (Enterprise Integrated Management
Systems) and the data in the effects of the FMEA is one small block in many
that delivers ORA (Organisational Robustness Analysis). The product and
service development process (APQP) is an organisational process within EIMS
that requires prominent levels in an ORA. This is measured by the levels of
connectivity, interaction, transparency and intelligent processing which
requires standards of retrieval for the data to give maximum use to the
organisation. The effects column in both the DFMEA and PFMEA are one small
part in this.
In
some cases in the development of DFMEA and PFMEA it has been witnessed that
organisations are using the generic FMEA severity criteria as descriptors for
the effect. For example in the case of an engine losing power the descriptor
may read Loss of System Function with no mention of what is happening to the
system, neighbouring assemblies or the end customer. While this offers certain
efficiencies within the DFMEA and PFMEA in terms of time and rating the
severity it is done at the expense of robustness in the product and service
development system.
Poor
use of DFMEA Effects column that could lead to misallocation of severity and
loss of field data robustness
With
no descriptor on the effect of failure- neighbouring
systems are potentially denied vital information on potential input failure or
variance and hence the true level of the severity of failure may not be
recorded as it may escalate to a higher rating.
- escalation
to higher system events may not be determined potentially missing vital design
actions and misallocating the severity level.
- With
a misallocated severity level in the DFMEA the severity level in the PFMEA
may also be misallocated leading to insufficient process controls on a critical
characteristic in the design
- multiple
effects of failure that may exist for a particular failure mode are missed.
- While
facilitators strive for the highest effect of failure and may not necessarily
look for other effects once a high value is found there is value in capturing
all effects which becomes relevant in the next point. Lower scoring effects can
be determined and captured outside the FMEA sessions which should focus on
driving toward action.
- field
feedback/failures root cause analysis will be missing a vital robustness link
that potentially reduces or eliminates the need for lengthy root cause
analysis.
- Field
failures and feedback will be delivered to the organisation in a manner that is
the effect on the end customer/consumer and at times in an emotive manner. With
only the severity ratings as an effects descriptor the effect described by the
customer will not be able to matched in a database search, file search or any
search for that matter. The robustness link has been lost. The simple view of
then being to match the effect with a failure mode and then a potential cause
cannot be obtained and the organisation is left with the full requirement of
root cause analysis and no avenue to circumvent any potential investigation
based on what the organisation already knows from its FMEA activity.

Robust
use of the effects column in FMEA leads to better identification of severity,
capturing multiple effects of failure and allowing more descriptive searches
for field problems that can identify potential root causes in a more efficient
manner.
Placing
the actual effect on the product and the end customer/consumer in the effects
column of the DFMEA and PFMEA if not current practice in an organisation is a
starting point to delivering robustness in the product and service development
system and will lead the organisation towards the type of holistic thinking and
action required to be at Industry4.0 and Quality4.0 levels.