IE 361 Lennox Project

IE 361 Lennox Project
Statistical Analysis of Run Test Data
By: Adli Shah Adnan, Josh Lamm, Peter Schulte, Yusuf Yigit
Project Contents and Outline
1. About Lennox
- Lennox International
- Branch : Lennox Marshalltown
- Client Information
2. Introduction to Run Test
- Run Test
- Category : Compressors and Fan Motors
3. Problem and Solution Methods
4. Solution Statement
5. Group Recommendations
About Lennox
Prologue : Lennox International Inc.
Lennox International was founded by Dave Lennox in 1895. Its operates through
three segments: Residential Heating & Cooling, Commercial Heating & Cooling and
Company name : Lennox International Incorporated
Headquarters :2140 Lake Park Boulevard, Richardson, Texas, United States
CEO : Todd M. Bluedorn
Products : HVAC (Heating, ventilation and air conditioning) equipment
Employees : ~ 9,700
NYSE : LII Stock Price : ~ $86.67
Prologue : Lennox Industries
Branch : Lennox Marshalltown
o Factory Location: 200 S 12th Marshalltown, IA 50158
o Plant Area: 996,000 sq.ft.
o Employees: 1,000 (approx.)
Products: Manufacturing residential heating and cooling
products: gas furnaces, split-system condensing units, splitsystem heat pumps
History: Original location of Lennox Furnace Company
Incorporated in 1904
Prologue : Client Information
Client Contact
Quality Engineer, Chuck Strobbe
o [email protected]
o 641-754-4095
Quality Engineer, Jason Kern
o [email protected]
o 641-754-4387
Introduction to Run Test
Run Test
Run Test is a station that is located at each manufacturing assembly line.
The station is to test the functionality and safety of their products
As part of Lennox’s quality goals, they measure the First Pass Yield (FPY)
of this process.
Our client indicates that there are errors in the test procedure.
o Two class error
 False Positive
 False Negative
o Tolerance are heavily biased against false positive
False Negatives are the target of this study.
Category : Compressor and Fan Motor
Fan Motors
Compressor and Fan motors are tested for Amp draw.
Amp draw are their way of testing motors for safety and quality assurance.
Each assembly number has different amp draws that will need to meet their
Both motors frequently fails by being outside of the parameters but will pass with
subsequent retest.
Problem and Solution Method
Identify The Problem
Test parameters (set by monitoring functioning properties with response history)
Parameters are not up-to-date for certain assembly models
New suppliers - different range of tolerance
Re-analysis of model parameters is necessary due to altering of
model parts that adjust their amp draw values
Auto generated data
(Fail - Red Column)
(Pass - Green Column)
Problem Statement & Analysis
Determining causes of error for First Pass
Yield (FPY)
This statistic is measured in parts per
million that were passed on the first test
Data Collection
Sample of the data
Amp measurements
are automated
Methods Outline
Pareto analysis (Top 80% of false negatives)
Creating histograms and control charts
Data Analysis (searching for patterns in the data)
Cp and Cpk analyses
Recalibration and new parameter calculations
Pareto Analysis
Pareto Analysis
Histograms and Control Charts
Control Charts & Data Analysis
Cp and Cpk Analyses
Cp and Cpk Example
Old parameter value
Old parameter value
Solution Statement
New Tolerance Calculations (Solution)
The new parameters through our process of
analysis should provide minimal retesting of
units that prove to be false-negatives
In this example, the parameters prove to be very
similar to what the model was previously set at.
Other models may show that parameter changes are
necessary based on past test data
Overall, this should provide fewer false-negatives
while still maintaining heavy bias against falsepositives
Sensitivity analysis of Cp and Cpk values can be
conducted to aid in the decision making of tolerance
setting for the managers and quality engineers
Group Recommendations
Due to this providing only a temporary solution we
New quality assurance program
More reliable testing equipment
Widen tolerance
Automate parameter adjusting
Organize worker instruction to remove over-retesting,
improper testing, and time waste
The new parameters have been provided so
that false-negatives may be minimized for
the time being and FPY may be higher
Considerations of a new-lasting quality
assurance process is recommended
Results will not be recognized during the
span of this course
Thank you for listening
For questions, feedback or comments
Please contact
Adli Adnan - [email protected]
Josh Lamm - [email protected]
Peter Schulte - [email protected]
Yusuf Yigit - [email protected]

similar documents