This Six Sigma project used the DMAIC model to address the increasing quality costs of Reyco Granning’s highest grossing assembly line. Project took place from February through May 2017. Results after 5 months (June - October 2017) • Reduced defects by 47% • $44,000 annual scrap savings • 50 hours annual rework savings

1. REDUCING DEFECTS
AT IFS ASSEMBLY
CALEB HOUSE

2. COMPANY OVERVIEW
Reyco Granning LLC
• Located in Mt. Vernon, Missouri
• Employs 167 workers
• Revenues of over $37 million in
2016

3. COMPANY OVERVIEW
Reyco Granning manufactures
heavy duty suspensions for:
• Trucks
• Trailers
• Fire trucks
• Ambulances
• Buses
• Motorhomes

4. PROBLEM
Quality problems on the Independent Front
Suspension (IFS) assembly line:
Data from October 2016 - January 2017:
• 67 defects found on finished units.
• $22,000 in scrap and rework costs.
IFS main line accounts for roughly 40% of
the plant's business.
It is essential that IFS products remain
profitable for the entire company to
succeed. IFS Main Assembly Line

5. OBJECTIVES
• Determine current performance of
the IFS line.
• Identify the frequently occurring
defects.
• Determine the root causes of the
defects.
• Implement countermeasures to
improve First Pass Yield (FPY) and
Defects per Unit (DPU) by 50%.

6. THE SIX SIGMA TEAM
• Kenny Smith, IFS Team Lead
• Brandon Carsten, IFS Production Lead
• Rick Head, Quality Inspector
• Caleb House, Manufacturing
Engineering Intern
The team met every Friday for 11 weeks
from January – April 2017.

7. METHODOLOGY
The DMAIC Model for Six Sigma Methodology
was used for this project.
1. Define the Problem
2. Measure the Process Performance
3. Analyze the Data
4. Improve the Performance
5. Control the Improved Performance

8. DEFINE PHASE
Gained understanding of the project using scope affinity diagram as a team exercise.

9. DEFINE PHASE
Gained understanding of the product
requirements with a customer analysis.
• Created list of items that are critical to
quality (CTQs).
• Used existing control plans, failure
mode effect analysis (FMEA),
specifications on prints, and team’s
work experience to create CTQ list.
Any product that does not meet
customer requirements is a defect.

10. DEFINE PHASE
Gained understanding the IFS
assembly process with a high-level
flow chart.
• Listed the process steps.
• Listed the inputs and outputs.
• Listed CTQs by station.
• Used existing process flow diagrams
and six sigma team’s work
experience to create.

11. Pre-
Assembly
Station
Assembly
Station 1
Assembly
Station 2
Assembly
Station 3
EOL Quality
Inspection
Inputs:
WO Info
Cradle
ID Tag
Serial Tag
Relay Rod
Idler Arm
Pitman Arm
Gearbox
HCV
Inputs:
Shock mount brackets
Upper A-arms
Lower A-arms
Carrier
Inputs:
Airbags
Airbag Plates
Spindle
Kingpins
Kingpin Caps
Torque Plate
Steering Arm
Tie Rod
HC Connecter
Inputs:
Oil Seal
Bearings
Brakes
Brake Rotar
Brake Caliper
Hub Cap
Oil
Hydraulic Lines
Kit
Wheel Alignment
Output:
Completed Unit
CTQs:
PreA
PreB
PreC
PreD
PreE
PreF
PreG
PreH
PreI
PreJ
CTQs:
S1A
S1B
S1C
S1D
CTQs:
S2A
S2B
S2C
S2D
S2E
CTQs:
S3A
S3B
S3C
S3D
S3E
S3F
S3G
S3H
IFS Assembly High Level Flow Chart
S3I
S3J
S3K
S3L
S3M
S3N
S3O

12. MEASURE PHASE
Selected metrics for data collection.
• KPI 1: Improve the number of units to
pass First Pass Yield by 50%
• KPI 2: Reduce number of defects per
unit by 50%
Created data collection sheets.
• Six Sigma Unit Log to capture unit
information.
• Daily Issues List to capture individual
defects.

13. MEASURE PHASE
Created operational definitions. Created data collection plans.

14. BASELINE DATA
Current Performance of IFS assembly:
• Data collected for March, 2017
• 21 defect occurrences
• 92 units made
• 80 units without defects
Calculations:
• DPU:
Number of Defects Occurrences
Number of Units
=
21 defects
92 units
= 0.228 DPU or 228,261 PPM
• FPY:
Number of Units Pass with No Defects
Number of Units
=
80 units
92 units
= 86.96%
Sigma
Level
Defects Per Million Opportunities
2 308,537
3 66,807
4 6,210 (industry average)
5 233
6 3.4 (world class)
Note. Industry Standard PPM. Adapted from Six Sigma (p. 14) by M.
Harry and R. Schroeder, 2000. Crawfordsville, IN: Random House, Inc.

15. TARGET DATA
Used baseline data to determine target data.
• Objective is to reduce DPU by 50% and
improve FPY by 50%.
• Baseline DPU: 0.228
• Baseline FPY: 86.96%
Calculations:
Sigma
Level
Defects Per Million Opportunities
2 308,537
3 66,807
4 6,210 (industry average)
5 233
6 3.4 (world class)
• Target DPU: 0.228 DPU x 0.50 = 0.114 DPU or 114,130 PPM
• Target FPY: ((1 – 0.8696) x 0.50) + 0.87 = 93.48%
Note. Industry Standard PPM. Adapted from Six Sigma (p. 14) by M.
Harry and R. Schroeder, 2000. Crawfordsville, IN: Random House, Inc.

16. ANALYZE PHASE
Newly Collected Data
• Collected for March 2017
• 4 defects types
• 21 defect occurrences
The team wanted to address more
than these 4 defects.
Note: Hub Cap Issue (BC) was a vendor
issue. Not due to assembly.

17. ANALYZE PHASE
Historical Data
• Collected May – September 2016
• 23 defects types
• 165 defect occurrences
Codes C and A were frequently
occurring in both newly collected data
and historical data.
There is no AV code in historical data
because it was not a collected code at
that time.

18. ANALYZE PHASE
Focused Defects Types
Defects were chosen to be analyzed based
on current and historical occurrences.
The 7 selected defects account for:
• 95% of defects found in March 2017
(20 of 21)
• 70% of defects from May – Sept. 2016
(115 of 165)

19. ANALYZE PHASE
Generating Potential Causes
• Each selected defect was
analyzed with a cause-and-
effect diagram for potential
causes.
• The 6 M’s were used to
trigger potential cause ideas.
• 44 potential causes were
generated for the 7 focused
defects.
(Code AJ)
Correctly
Installed
Fittings
All Lines &
Machine
Method
Environment
Man
Materials
Measurement
vendor
Defective line from
from vendor
Defective fitting
Forgot to tighten line fitting
Operator forgot a fitting
Wrong type of fitting
Fitting in wrong place
Cause-and-Effect Diagram for Code AJ

20. ANALYZE PHASE
Verifying Potential Causes
• Of the 44 potential causes, 19 were
chosen to be studied based on team
member preference.
• The 19 potential causes were tested
by further stratifying of statistical
data, additional data collection,
and/or six sigma team observation.
• 12 potential causes were found to be
actual root causes.

21. IMPROVE PHASE
X

22. CODE C: CHECK SHEET COMPLETE
Old Check Sheet New Check Sheet
• Worked with operators and QA to modify the assembly check sheet
• 37 improvements made. It took 4 revisions!

23. QA Assistant
notifies QA
inspector of
changes
QA inspector
informs and
trains IFS team
lead on changes
IFS team lead
informs and trains
IFS operators on
changes
IFS operators
follow new
changes on form
QA Assistant
updates IFS
check sheet
CODE C: CHECK SHEET COMPLETE
Check Sheet Change Communication Flow Chart
• Established process for updating check sheets.
• Previously, operators were not informed when changes were made.

24. CODE A: TORQUES RECORDED IN SYSTEM
Torque Rules Error
Operator scanned multiple unit barcodesData entry error
by engineer
Scanning Error

25. CODE AV: TORQUED BOLTS NOT MARKED

26. CODE AV: TORQUED BOLTS NOT MARKED
• Put check for marking
torqued bolts after
every station.
• Discussed with
operators about
importance of marking
the torqued bolts.
• Clarified procedure for
marking torques and
recorded in work
instructions.

27. CODE AJ: ALL FITTINGS INSTALLED CORRECT
• 7 different fittings used
for 5 different gear
boxes on 4 different
units.
• Some fittings look alike
but are different sizes.
• Defects occurred
because operators were
installing wrong fittings

28. CODE AJ: ALL FITTINGS INSTALLED CORRECT

29. PROJECT STATUS: BEFORE STRIKE
Targets Achieved!
• May 26 – last countermeasure was
put in place.
• August 14 – the union went on
strike and temps were hired.
• Improved FPY from 87% to 94%
(58% improvement).
• Improved PPM by from 228261 to
69124 (70% improvement).

30. PROJECT STATUS: INCLUDING DURING STRIKE
• Many quality issues during strike.
• August 28 – the strike ended and
union workers returned to work
with new temp workers.
*Data as of October 26, 2017
*
• Improved FPY from 87% to 89%
(15% improvement).
• Improved PPM by from 228261 to
120690 (47% improvement).
Targets not achieved, but quality was improved.

31. SAVINGS AFTER 5 MONTHS
• Rework Savings
• Saved over 50 hours of rework per year.

32. SAVINGS AFTER 5 MONTHS
• Projected Scrap Savings
• Improved scrap cost by 65%
• $44,000 annual savings

33. CONCLUSION
• Determined current performance of the
line to be 0.228 DPU and 87% FPY.
• Identified 7 defect types that account
for 95% of newly collected data and
70% of historical data defects.
• Identified 12 root causes.
• Implemented 10 countermeasure plans.
• Improved FPY by 15% in 5 months.
• Improved PPM by 47% in 5 months.
• Saved over 50 hours of rework per year.
• Saved $44,000 in scrap costs per year.

34. REFERENCES
Gijo, E. V., Antony, J., Kumar, M., McAdam, R., & Hernandez, J. (2014). An
application of six sigma methodology for improving the first pass yield of a
grinding process. Journal of Manufacturing Technology Management, 25(1),
125-135. Retrieved from
http://search.proquest.com/docview/1476442078?pq-origsite=summon
Harry, M., & Schroeder, R. (2000). Six Sigma: The Breakthrough
Mohan, R. R. (2012). Quality improvement through first pass yield using
Management Strategy Revolutionizing the World's Top Corporations (p. 14).
Crawfordsville, IN: Random House, Inc.
Mohan, R. R. (2012). Quality improvement through first pass yield using
statistical process control approach. Journal of Applied Sciences, 12(10),
985-991. Retrieved from
http://www.scialert.net/abstract/?doi=jas.2012.985.991