A review of the importance of assessing risk and how to measure it using a FMEA tool.

1. Section & Lesson #:
Pre-Requisite Lessons:
Complex Tools + Clear Teaching = Powerful Results
Risk Assessment with a FMEA Tool
Six Sigma-Improve – Lesson 7
A review of the importance of assessing risk and how to measure it using
a FMEA tool.
Introduction #11 – ABC Model
Six Sigma-Overview #02 – Risk Analysis
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted by any means
(electronic, mechanical, photographic, photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.

2. Review: The Role of Risk in our Decisions
o Remember the jelly bean example from the lesson on risk analysis?
• Question: How many red jelly beans are in the jar?
 Method 1: Empty out all of the jelly beans and manually count the red ones.
– Advantage: More Accurate vs. Disadvantage: More Time
 Method 2: Count the red ones in a small sample & multiply by the jar’s volume.
– Advantage: Less Time vs. Disadvantage: Less Accurate
o Which method of counting the jelly beans is best?
• Neither, because both are equally valid; the “best” method depends on RISK vs. REWARD.
• Risk is determined by the disadvantages for each method:
 Method 1: Is the reward worth the risk of taking “more time” to use this method?
– If the reward was $1M, then perhaps the risk of more time is worth ensuring we’re more accurate.
 Method 2: Is the reward worth the risk of being “less accurate” to use this method?
– If the reward was a t-shirt, then perhaps the risk of being less accurate is worth it taking less time.
o Remember the ABC Model?
• Risk is a critical part of what influences our beliefs, which influence our conclusions, actions, etc.
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted by any means (electronic, mechanical, photographic,
photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.
2
BELIEF SCALE
Consequences

3. Why is risk so important?
o Risk affects every decision we make!
• Would you drive your car if you knew there was a high risk it would break down?
• Would you eat at a restaurant that you knew had a high risk of food poisoning?
o Prudent business decisions should assess and measure risk.
• Risk has an inverse relationship with confidence.
• Data (i.e., proof, evidence, etc.) builds confidence.
• The lack of data (i.e., assumptions) creates risk.
• To reduce risk and build confidence, get data!
o 6 Sigma tools (like FMEA) help assess risk (and confidence) in business decisions.
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted by any means (electronic, mechanical, photographic,
photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.
3
The empty portion
(air) represents
assumptions (risk)
The filled portion
(water) represents
data (confidence)
The entire glass
represents our available
understanding for
making decisions
Just as adding
water displaces air
in the glass, adding
data displaces our
assumptions.

4. FMEA Tool Defined
o What is a FMEA tool?
• “FMEA” is a risk management tool that stands for “Failure Modes & Effects Analysis”.
• It tracks all possible failures in a process, their potential impact and a plan to mitigate them.
• The FMEA scores and prioritizes all possible failure modes to help mitigate potential risks.
o Below is the general layout of the FMEA tool and it’s key components defined:
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted by any means (electronic, mechanical, photographic,
photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.
4
Step
ID
WHEN:
Item/ Process Step
WHAT:
Potential Failure
Mode
Potential Effect(s) of
Failure
Severe
WHY:
Potential Cause(s) of
Failure
Occur
Current / Future Controls
Detect
RPN
Recommended Actions
(Response Plans: SOP's,
Containment, Control Plans)
Responsibility and
Target Date
Completion
Action Taken
Severe
Occur
Detect
RPN
Source file corrupt or
unavailable
DB won't get req'd data
feed
8 Error from mainframe 1 Users monitoring daily file
feed
1 8 0
System fails to import
data
System abends 5 Power or oper system
failure
1 Sys Admin monitors app;
process can be re-run
1 5 0
2 Identify new & cleared
credits
Cannot detect new or
cleared credits
System can't properly track
or report credits and
metrics
8 Orig system or fed data is
corrupt (e.g., invalid Pro #,
etc.)
2 Can only be detected by
doing a MSA to assure
accuracy; else it normally
5 80 0
Cannot identify mass re-
rated credits
Extra research may be
necessary for those credits
6 Criteria for identifying re-
rated credits may be
inaccurate or incomplete
5 Detection during research of
credits will help in modifying
the criteria for automatic
4 120 0
Identified credits are not
sent to Rating
Rating can't validate credit
status back to Cash App
7 System fails to transmit
identified credits to Rating
2 Sys will track file
transmission to ensure
delivery; Rating can also
2 28 0
Cannot accurately
identify overpayments
Extra research may be
necessary for those credits
6 Criteria for identifying
overpaid credits may be
inaccurate or incomplete
5 Detection during research of
credits will help in modifying
the criteria for automatic
4 120 0
Cannot access GetPaid
data
Collections can't confirm to
Cash App what credits to
4 GetPaid data unavailable
to connect to
2 Sys will identify any failure to
connect to GetPaid
1 8 0
Returned letters are not
audited
A credit may no longer be
valid, but refunded anyway
8 Clerk fails to audit the
credit before issuing the
5 Would be detected only by
doing random audits
7 280 0
Refund is not sent for a
valid credit
Payee doesn't receive the
refund they expect
6 Clerk fails to issue or mail
refund for a valid credit
2 Detect by matching what
refunds were paid per sys
with what funds were
3 36 0
5 Audit & send refund;
clear credit
1 Import open credits
3 Identify mass re-rate
credits; send to Rating
4 Identify overpmts;
compare to GetPaid
Failure Modes & Effects
Potential failures in each process
step and the expected effects &
severity from those failures.
Current Controls & Detection
Existing controls that help detect the
failures and their estimated
capability to detect those failures.
Mitigation Plan
The actions and assigned
owner(s) for mitigating the
identified risks.
Process Steps
Detailed steps in the new or
existing process that are
affected by the improvements
RPN (Risk Priority Number)
A value measuring the severity, frequency,
and detection capability of each failure; the
RPN is used to rank and prioritize all failures
Potential Causes
Potential reasons why a failure
occurs and the estimated
frequency of occurrence.

5. When and How to Build the FMEA
o When is the FMEA tool used?
• It’s generally used when there is a significant process change that requires a risk assessment.
 It’s most often used in the Improve phase of the DMAIC methodology to evaluate any potential risks from
the proposed improvements.
 It can also be closely tied to the Control Plan created in the Control phase of DMAIC, therefore it may be
necessary to review or extend the FMEA after the improvements are implemented.
 It’s also used in the Design phase of the DMADV methodology before implementing a new process.
• The FMEA is generally built by a team of SMEs during a meeting (2 to 4 hours).
 The SMEs should include experts from any area affected by the improvements.
 Below are the general steps for leading the meeting when building the FMEA:
o Is the FMEA tool required for every project?
• No. Sometimes the Sponsor or customer is willing to assume inherent risks from the new
improvements and won’t require a risk assessment.
• When risks are unknown or if there is any hint of doubt in the expected success of a proposed
improvement, then the FMEA should be used to fully address and measure potential risks.
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted by any means (electronic, mechanical, photographic,
photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.
5
1. Identify list
of Process
Steps
2. Identify
Failure Modes
& Effects
3. Identify
Potential Root
Causes
4. Identify
Current
Controls
5. Calc RPN
and Define
Mitigation Plan
INPUT to Meeting PROCESS during Meeting OUTPUT from Meeting

6. FMEA Process Steps 1 & 2
o INPUT to Meeting
• #1 – Identify List of Process Steps
 These steps should be a part of the detailed process map that may have been created for mapping the
improvements. Each process step should be noted with its own Step ID #.
 It’s OK to limit the process steps to only those deemed as critical or within control of your team.
o PROCESS During Meeting
• Meeting Kick-Off (~20 min)
 The Sponsor/Champion should kick-off the meeting and return at the end to review the results.
 The list of process steps should be agreed upon by the team; allow for some clarification as needed.
• #2 – Identify Failure Modes & Effects (~30 – 40 min)
 For each process step, brainstorm one or more potential ways a failure may occur for that respective step.
 Rank the Severity on a scale of 1 (Lowest) to 10 (Highest) for each failure mode & effect.
– Below are suggestions for defining severity; these may be changed if the team agrees:
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted by any means (electronic, mechanical, photographic,
photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.
6
Minor (Rank 1) Low (Rank 2 - 3) Moderate (Rank 4 - 6) High (Rank 7 - 8) Very High (Rank 9 - 10)
Unreasonable to expect that
the minor nature of this
failure will have any
noticable effect on item or
system performance or
subsequent process or
assembly operation.
Customer will most likely not
be able to detect the failure.
Due to the nature of this failure,
the customer experiences only
slight annoyance. Customer will
probably notice slight
deterioration of the item or
system performance or a slight
inconvenience with a
subsequent process or assembly
operation, i.e. minor rework.
Failure causes some customer
dissatisfaction which may
include discomfort or
annoyance. Customer will
notice item or system
performance deterioration.
This may result in unscheduled
rework/repair and/or damage
to equipment.
High degree of customer
dissatisfaction due to the nature
of the failure, such as
inoperable item or system.
Failure does not involve safety
or government regulation. May
result in serious disruption to
subsequent processing or
assembly operations and/or
require major rework.
Failure affects safety or
involves noncompliance
to government
regulations. May
endanger machine or
assembly operator (9
with warning, 10 without
warning)

7. FMEA Process Steps 3 & 4
o PROCESS During Meeting (Cont’d)
• #3 – Identify Potential Root Causes (~20 – 30 min)
 Define what the potential root causes are for each failure mode & effect.
 Rank the estimated occurrence on a scale of 1 (least frequent) to 10 (most frequent) for each root cause.
– Below are suggested ratios for defining these occurrences; these may be changed if the team agrees:
• #4 – Identify Current Controls (~20 – 30 min)
 Define whatever controls are already in place that may detect each of those failure occurrences.
– The existing controls can include controls managed by other people or processes outside the team.
 Rank the estimated detection rate on a scale of 1 (always detected) to 10 (never detected) for each failure
occurrence (i.e., the better the ability to detect the errors, the lower the score).
– Below are suggested rankings for defining the detection capability; these may be changed if the team agrees:
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted by any means (electronic, mechanical, photographic,
photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.
7
Remote (Rank 1) Very Low (Rank 2) Low (Rank 3 - 5) Moderate (Rank 6 - 7) High (Rank 8 - 9) Very High (Rank 10)
Failure unlikely. No
failures ever
assocaited with this
process or almost
identical processes
(1=1:1.5M)
Only isolated failures
associated with this
process or almost
identical processes
(2=1:150K)
Isolated failures
associated with similar
processes (3= 1:30K;
4=1:4500; 5=1:800)
This process has
occasional failures, but
not in major
proportions (6=1:150:
7=1:50)
This process or similar
processes have often
failed (8=1:9; 9=1:6)
Failure is almost
inevitable (10=>1:3)
Very High
(Rank 1 - 2)
High
(Rank 3 - 4)
Moderate
(Rank 5 - 6)
Low
(Rank 7 - 8)
Very low
(Rank 9)
Absolutely No
Detection (Rank 10)
Current controls almost certain to detect
the failure mode. Reliable detection
controls are known with similar processes.
Process automatically prevents further
processing.
Controls have a good
chance of detecting
failure mode, process
automatically detects
failure mode.
Controls may detect
the existence of a
failure mode.
Controls have a poor
chance of detecting
the existence of
failure mode
Controls probably
will not detect the
existence of failure
mode
Controls will not or can not
detect the existence of a
failure. No known controls
available to detect failure
mode.

8. FMEA Process Step 5
o OUTPUT From Meeting
• #5 – Calculate RPN and Define Mitigation Plan (~20 – 30 min)
 The Risk Priority Number (RPN) is automatically calculated as a multiplier of the 3 rankings defined in step
2 (severity), step 3 (frequency), and step 4 (detection capability).
– The RPN can range from 0 to 1000; despite that wide range, any process step with RPN over 250 should be considered a
potential high risk for failure.
– Since the scoring is subjective, review the RPNs for all process steps to determine what is a reasonable threshold for what
should be considered a high risk failure.
 Consider building a mitigation plan for the process steps having a high RPN.
– The mitigation plan isn’t required, but for very high risk process steps, it can be very helpful to define them in advance.
– As a team, define 1) the recommended actions in the event of a failure, and 2) the person responsible to perform the
action and estimated turnaround time to complete it.
– The remaining “Action Taken” and rankings are optional; for any process step having a very high risk of failure, it can help
to apply a ranking and RPN for those mitigation actions to assess their inherent risks.
o Now What?
• The FMEA can help unify the team around their improvements to implement and it can build their
confidence that all potential risks are identified with an appropriate mitigation plan.
• The FMEA should be retained on hand for the entire team and distributed to any stakeholder that
controls or is affected by the process improvements.
• If any process steps are very high risk (high RPN), then they should be validated by any other
stakeholder responsible for or knowledgeable of that respective process step.
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted by any means (electronic, mechanical, photographic,
photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.
8

9. Practical Application
o Identify an existing high-level process map used in your organization.
• Instead, you can build a process map for a simple routine process you follow such as driving to
school or work, ordering a meal at a restaurant, mowing the lawn, etc.
 Whatever process you use, ensure you only have 3 to 5 key process steps.
o Build a FMEA using the process defined above.
• Identify each critical process step.
• For each process step, identify all potential failure modes.
• For each failure mode, identify the potential effects from those failures and severity level.
• For each failure effect, identify all potential causes and the occurrence frequency level.
• For each failure cause, identify all controls that could detect the cause and its detection level.
• Calculate the Risk Priority Number (RPN) based on the severity, frequency, and detection levels.
• As needed, try filling out the remaining portion for building a mitigation plan.
o Based on the FMEA that was built, try answering the following questions:
• Which process steps had the most failure modes?
• Which failure modes had the highest RPN?
• Which process step had the highest combined RPN value based on all its failure modes?
• How do these results compare with the level of risk you originally expected for this process?
Copyright © 2011-2019 by Matthew J. Hansen. All Rights Reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted by any means (electronic, mechanical, photographic,
photocopying, recording or otherwise) without prior permission in writing by the author and/or publisher.
9