Evaluation of Process Risks in Industry 4.0 Environment.
Svingerova, Michaela ; Melichar, Martin
Evaluation of Process Risks in Industry 4.0 Environment.
1. Introduction
In today's industrial production, process risk assessment by
means of PFMEA is one of the cornerstones of control processes in most
companies. The earlier automotive standard version, ISO TS 16949, merely
recommended FMEA as an appropriate tool for companies involved in the
automotive supply chain to master their preventive measures. Its
revision and transformation into IATF 16949 and linking to an equally
updated version of ISO 9001:2015, the role of risk management was
greatly strengthened. IATF-certified companies are required to identify
and assess the risks of all company processes on a regular basis as part
of their internal management system. Now, the FMEA or PFMEA tool is not
just recommended but expressly required by this standard.
The history of FMEA goes back to the 1940s when the U.S. Army
sought a technique for eliminating errors in the machines and equipment
in use. 1960s saw the the first non-military use of this method when
NASA (National Aeronautics and Space Administration) began to use FMEA
to identify potential risks in Apollo and Gemini programmes. The truly
civilian era of risk assessment began 10 years later when the method was
adopted by the automotive industry, particularly the Ford company in an
attempt to address the low quality of its Ford Pinto model.
PFMEA is a team-based approach to potential risk assessment where
the multidisciplinary team is usually led by the engineer who is
responsible for the product design. Active involvement of
representatives of all areas concerned is expected, including assembly,
production, design, testing and others [7]. It is advisable to begin
risk mapping with a flow chart of the process in order to map it clearly
and identify its boundaries. The customer is typically the end user but,
depending on the definition, it may comprise the downstream operation or
subsequent assembly operation.
Once the process and its boundaries have been clearly defined,
brainstorming, Ishikawa diagram or other tools are used to identify as
many risks as possible which may threaten the process[2;6]. These risks
are then evaluated according to a consistent method from three aspects:
severity of consequences in respect of the customer, occurrence of
consequences in respect of the customer and the chance of detection.
Each of these aspects is assigned a score on a scale from 1 to 10 as
indicated in the tables below.
After the assessment, risk priority numbers (RPN) are determined
for individual risks by simple multiplication of values for all three
aspects. Risks with RPN values higher than an internal or contractual
limit and those with the greatest severity are addressed on a priority
basis as part of prevention measures.
2. Experimental Procedure
2.1. Real case study
This paper describes an assessment of automation of production
using real-world data from a company which operates in the automotive
industry. The data comes from serial production of canned catalysts. Two
production lines were compared. Both of them are currently used in
production. The one which has not been automated yet has been in
operation for several years with minor improvements gradually
introduced. one year ago, an automated line was installed next to
it[2;3;4]. In this new line, most operations which used to be performed
by a worker/operator are done by a robot. Even this automated line,
though, requires the operator to carry out basic infeed and outfeed
tasks. The automated line is monitored continuously and its processes
are being refined. Such preparations may eventually lead to full
automation.
2.2. Finished product
The finished products of both lines are canned catalysts. They
consist of the monolithic catalyst, a support mat and a steel tube. It
should be pointed out that for the purposes of this paper, the canned
catalyst is considered the final product, although this only applies to
the production lines described here. The end use requires its
incorporation into the exhaust system of a vehicle.
The monolithic catalyst is the single most important and most
expensive part of the catalytic converter. It is a ceramic cylinder
which contains elements such as platinum, rhodium and palladium. It is a
very brittle part. In the catalytic converter, its role is to reduce the
amount of harmful products in the exhaust gases. This is provided by its
sophisticated inner structure which is not identical in all monolithic
catalysts. Every car manufacturer and every car model has its specific
monolithic catalyst structure and size. The figure below shows various
cell structures as seen from the top or bottom side of the monolithic
catalyst[1;5;7].
The sizes of monolithic catalysts differ as well, with diameters of
approximately 15 cm and heights around 15 cm in passenger vehicles. The
typical appearance of a monolithic catalyst is shown in the figure
below.
The support mat in the canned catalyst is made of glass wool. It
provides both thermal and noise insulation and protects the brittle
monolithic catalyst from damage. Shapes of the support mat can differ,
as illustrated below, but the insulation effect must remain.
To give an idea of the glass wool texture, the diagram below shows
the classification of fibres of which the insulation is composed.
The last part of the canned catalyst is its outer shell. It is in
fact an iron tube. The first step involves rolling up sheet metal to a
cylinder shape and laser welding it together in the machine. The
resulting tube is cut by laser to the desired lengths to be used as
"cans" for the catalytic converter. This part is shown in the
figure below, on the left. The article in the centre is the monolithic
catalyst and the one on the right is the rolled-up support mat. The
figure below them shows a 3D rendering of the completed canned catalyst.
2.3. Production process
The sequence for making the canned catalyst is obvious from the
parts used. The process flow below indicates the individual steps.
The automated line will be considered first. The boxes to the right
of and below the dotted rectangle in the process flow chart indicate the
operations which are done by human operator even in the automated line.
They include feeding individual parts to the machine, final inspection,
and transfer to downstream lines for final processing. Inside the dotted
rectangle, there are the operations which are performed by the machine.
The photograph (Fig. 9) below the flow chart shows the
"Kirschenhofer Maschinen" automated canning line.
In the conventional canning line, all the operations indicated in
the flow chart are carried out by a human operator in several work
stations with production machines and respective control computers [8].
Today, this conventional line is only used occasionally, on a customer
request, typically for making small batches of canned catalysts. Due to
confidentiality restrictions, the photograph of this conventional
line--referred to as "shrinking line"--cannot be shown.
3. FMEA analysis
3.1. Manual operator
3.2. Automated process
3.3. Summary
An FMEA analysis based on a previous SPC analysis of process
stability has clearly shown that substitution of manual operations with
an automated process considerably reduced RPN values. The reduction was
in the range from 33% to 84%. The only area where the risk priority
number has not changed was the packaging of the part which remained a
manual operation even after process automation. The most interesting
item in the analysis is the number of critical processes, i.e. the
potentially critical ones.
Like its many automotive counterparts, the company has defined its
internal RPN acceptance level as 100. As the analysis shows,
appropriately configured automation eliminated seven critical processes
which are of major importance to the end user.
4. Conclusions
Research team of university of West Bohemia had the goal to analyse
influence of risk management connected with replacement human work.
As mentioned in the introduction, today's automotive industry
relies in its mandatory risk assessment activities exclusively on the
FMEA method which is required by the IATF standard. The method was also
used for experimental analysis of a production process, i.e. a
substitution of a manual assembly workplace with a robotic cell, an
upgrade which can undoubtedly be classified under Industry 4.0. Effects
of a specific robotisation step were evaluated in collaboration with
researchers from the University of West Bohemia/Regional Technological
Institute.
Both processes were evaluated by a multidisciplinary team
comprising members from all areas of the production process. Using
brainstorming and data from real-world production (which were also
utilized in the previous contribution [9]), risks were identified for
each process. Each of the risks was classified by the FMEA team from
three perspectives: severity/occurrence/detection. Since supply chain
experience suggests that setting internal limit RPN values is not
effective, this calculation was only use for statistical evaluation and
for comparing both groups of results.
The results clearly show that despite the high initial investment
costs for introducing an Industry 4.0 workplace, this investment pays
off, at least from the risk management perspective. One general finding
is that although the robotic cell comprises elements for active
detection of errors and integrates a number of features of poka-yoke
philosophy, the outcomes are virtually incomparable. RPN values become
lower by an order of magnitude, on account of a drop in the detection
and occurrence aspects.
The results achieved are in agreement with an earlier experiment,
in which this process was assessed from the perspective of capability
and stability. Next step in this analyse will be evaluation of risk
during mass production inspection.
DOI: 10.2507/28th.daaam.proceedings.142
5. Acknowledgements
This paper was created due to the project GA ZCU v Plzni:
SGS-2016-005 "Research and development for innovation in field of
Manufacturing processes--Technology of metal cutting II.
6. References
[1] Melichar M., Kubatova D., Kutlwaser J.: (2016). CMM measuring
cycle and human factor, Proceedings of the 27th DAAAM International
Symposium, ISBN 978-3-902734-08-2
[2] Sahno J., Shevtshenko E., Zahharov R.:(2015) Framework for
Continuous Improvement of Production Processes and Product Throughput,
Procedia Engineering, Volume 100, Pages 511-519, ISSN 1877-7058,
http://dx.doi.org/10.1016/j.proeng.2015.01.398.
[3] Hiltmann K.: Predicting Unknown Failures,(2015) Procedia
Engineering, Volume 131, Pages 840-849, ISSN 18777058,
http://dx.doi.org/10.1016/j.proeng.2015.12.392.
[4] Hak Soo Mok, Hyun Su Song, Deuk Jung Kim, Jin Eui Hong, Seung
Min Lee, Jung Tae Ahn: Determination of Failure Cause in
Remanufacturing,(2015) Procedia Engineering, Volume 100, Pages 14-23,
ISSN 1877-7058, http://dx.doi.org/10.1016/j.proeng.2015.01.337.
[5] Chandrajit P., Anand S.: Correlating Failure Mode Effect
Analysis (FMEA) & Overall Equipment Effectiveness (OEE), (2012)
Procedia Engineering, Volume 38, Pages 3482-3486, ISSN 1877-7058,
http://dx.doi.org/10.1016/j.proeng.2012.06.402.
[6] Xiaoli Zhang, Zhen He, Liangxing Shi, Process Quality Metrics
for Mechanical and Electrical Production Lin, (2011), Procedia
Engineering, Volume 24, Pages 6-11, ISSN 1877-7058,
http://dx.doi.org/10.1016/j.proeng.2011.11.2592.
(http://www.sciencedirect.com/science/article/pii/S1877705811054440)
[7] Vladimir Filaretov, Alexander Zuev, Anatoly Khvalchev,
Development of Approach to Automatic Machining of Composite Parts
without their Rigid Fixing by Means of Multilink Manipulators, (2014)
Procedia Engineering, Volume 69, Pages 4-12, ISSN 1877-7058,
http://dx.doi.org/10.1016/j.proeng.2014.02.196.
(http://www.sciencedirect.com/science/article/pii/S1877705814001982)
[8] Qi Guo, Kuangjie Sheng, Zheng Wang, Xilin Zhang, hengyi Yang,
Rui Miao, Research on Element Importance of Shafting Installation Based
on QFD and FMEA,(2017), Procedia Engineering, Volume 174, Pages 677-685,
ISSN 1877-7058, http://dx.doi.org/10.1016Zj.proeng.2017.01.205.
[9] Faurecia's internal sources
[10] Svingerova M., Melichar M., Evaluation of automated
production, (2017) Pro-tech-ma
Caption: Fig. 1. Flow chart [10]
Caption: Fig. 2. Example of cell structures [9]
Caption: Fig. 3. Monolithic catalyst Source: Internal company
documents
Caption: Fig. 4. Detail of isolation [9]
Caption: Fig. 5. Fibres classification [9]
Caption: Fig. 6. Canned catalyst [9]
Caption: Fig. 7. Making catalyst schema [9]
Caption: Fig. 8. Robotics work place [9]
Table 1. FMEA--severity [10]
Consequence Severity in respect of Score
the customer
Failure to meet safety Affects safe operation 10
of regulatory requirements without warning 9
Affects safe operation
after warning
Impaired or lost primary Vehicle disabled 8
function Vehicle drivable, reduced 7
power
Impaired or lost secondary Loss of comfort 6
function Reduced comfort 5
Inconvenience Noticed by 75% of customers 4
Noticed by 50% of customers 3
Noticed by 25% of customers 2
No consequence No perceptible consequence 1
Table 2. FMEA--occurrence [10]
Consequence Occurrence in respect of the customer Score
Major 100000ppm 10
Medium 50000ppm 9
20000ppm 8
10000ppm 7
Medium 2000ppm 6
500ppm 5
100ppm 4
Low 10ppm 3
1ppm 2
Very low Eliminated by prevention 1
Table 3. FMEA--identification [10]
Chance of detection Probability of identification Score
of the cause
Detection impossible Cannot be detected 10
Near impossible Low probability of detection 9
Negligible E.g. visual inspection 8
Detection of the problem in the 7
Very low workplace
Low Measurement after the operation was 6
completed by the operator
Medium Detection by operator with the 5
aid of automated tools
Medium to high Inspection with automated tools 4
and separation
High Inspection with automated tools 3
and separation
Very high Automated tools for error detection 2
Fault prevention Error prevention 1
Fig. 9. FMEA analysis--manual operator [10]
Process
Step Function Requirements
number Potential Failure
Mode
Operation Item.
1 Components Wrong part
loading into orientation
the machine
Insuficient
part position
Insufficient part
position (
missing component
extra component
2 clamping clamping clamp quality
parts
3 part assy Part damage
unloading
from the
machine
4 packaging Box Wrong Container type
Wrong number of
parts in container
Missing label
part skipped operation
condition
Step Potential Effects(s) Severity Class Potential Cause(s)/
number of Failure Mechanism(s) of
Failure
1 Improper fit Operator mistake,
to vehicle 7 Standardize working
instruction not
followed
Tooling damage, Operator mistake,
time loss Standardize working
instruction not
8 followed
Tooling damage, Operator mistake,
time loss Standardize working
8 instruction not
followed
Improper fit to Operator mistake,
vehicle 7 Standardize working
instruction not
followed
Improper fit to component inserted
vehicle twice
7
components sticked
with oil
7
2 Part may not meet parts weren't
system design 7 clamped
requirements
parts weren't
7 clamped
parts weren't
7 clamped
Part may not meet unsifficient
system design 7 clamping
requirements
unsifficient
7 clamping
3 Improper fit Falling of the part
to vehicle
6
4 Time loss, wrong Material handler
number of parts 3 mistake
Inventory Operator mistake,
discrapency 3 packaging
instruction not
followed
Lost of Operator did not
tracebility 4 print label or
forget to place
printed label on box
Improper fit operator mistake
to vehicle 7
Occurence Current Process Current Process Controls
Step Controls Detection
number Prevention
Process steps sensors control correct
1 4 description in orientation
Standardize work
instruction
Process steps pin is controling correct
description in position (sensor on end
Standardize work stop)
4 instruction, pin is
controling correct
position (sensor on
end stop)
Process steps visual control, operator
description in has to ensure position
4 Standardize work with small equipment
instruction
Process steps sensor checking right
4 description in position
Standardize work
instruction
Work instruction install sensor to control
right material thickness
3 before clamping
Work instruction install sensor to control
right material thickness
3 before clamping
operator training sensor checking right
2 4 position of clamps
operator training sensors checking right
4 position of clamps
operator training sensors checking right
4 position of clamps
operator training sensor checking right
4 position of clamps
operator training sensors checking right
4 position of clamps
Design of the holders Operator training
3 allowing easy part
2 manipulation
no control Operator training,
4 4 packaging instruction
Operator training, Discrapancy for next
4 packaging instruction station
Operator training, Reading lable on next
4 automatic printing process step
after last piece
inspection
No control 100% Touchpoint
3 inspection on another
work station
Detection R.P.N
Step
number
1 2 56
3 96
5 160
5 140
3 63
3 63
2 4 112
4 112
4 112
4 112
4 112
3
8 96
4 8 96
8 96
4 64
7 147
Fig. 10. FMEA analysis--robot [10]
3.2. Automated process
Process
Step Function Potential Failure
number Requirements Mode
Operation Item.
1 Components components Wrong part
loading into (Stamping) orientation
the machine
Insuficient part
position
Insufficient part
position (
missing
component
extra component
2 clamping clamping parts clamping quality
3 part assy Part damage
unloading
from the
machine
4 packaging Box Wrong Container
type
Wrong number of
parts in container
Missing label
part condition skipped operation
Step Potential Effect(s) of Severity Class
number Failure
1 Improper fit to vehicle
7
Tooling damage, time
loss
8
Tooling damage, time
loss 8
Improper fit to vehicle
7
Improper fit to vehicle
7
7
2 Part may not meet
system design 7
requirements
7
7
Part may not meet
system design 7
requirements
7
3 Improper fit to vehicle
5
4 Time loss, wrong
number of parts 3
Inventory discrapency
3
Lost of tracebility
4
Improper fit to vehicle
7
Step Potential Cause(s)/ Current Process
number Mechanism(s) of Occurrence Controls
Failure Prevention
1 Operator mistake, Process steps
Standardize working 2 description in
instruction not Standardize work
fo1lowed instruction
Operator mistake, Process steps
Standardize working description in
instruction not Standardize work
followed 2 instruction, pin is
controling correct
position (sensor on
end stop)
Operator mistake, Process steps
Standardize working 2 description in
instruction not Standardize work
followed instruction
Operator mistake, Process steps
Standardize working description in
instruction not 3 Standardize work
followed instruction
component inserted Work instruction
twice
2
components sticked Work instruction
with oil
2
2 parts weren't operator training
clamped 2
parts weren't operator training
clamped 3
parts weren't operator training
clamped 3
unsifficient clamping operator training
3
unsifficient clamping operator training
3
3 Falling of the part Design of the holders
2 allowing easy part
manipulation
4 Material handler no control
mistake 4
Operator mistake, Operator training,
packaging instruction 4 packaging instruction
not followed
Operator did not print Operator training,
label or forget to 4 automatic printing
place printed label after last piece
on box inspection
operator mistake No control
3
Step Current Process Controls
number Detection Detection R.P.N
1 sensors control correct
orientation 1 14
pin is controling correct
position (sensor on end
stop)
1 16
visual control, operator
has to ensure position 2 32
with small equipment
sensor checking right
position
2 42
install sensor to control
right material thickness
before clamping 2 28
install sensor to control
right material thickness
before clamping 2 28
2 sensor checking right
position of clamps 2 28
sensors checking right
position of clamps 2 42
sensors checking right
position of clamps 2 42
sensor checking right
position of clamps 2 42
sensors checking right
position of clamps 2 42
3 Operator training
3 36
4 Operator training,
packaging instruction 8 96
Discrapancy for next
station 8 96
Reading lable on next
process step 4 64
100% Touchpoint
inspection on another 7 147
work station
COPYRIGHT 2018 DAAAM International Vienna
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.