March 29th, 2022 
*
The Ford-Firestone Case
Quality Assurance, Management, and Control
*
Case Overview
- Introduces quality management
- Set in Fall 2000
- Backdrop of confusing information, as claims against Ford and Firestone mount following a TV segment
we need to assess whether the problem is critical enough and warrants a recall
- If there is, indeed, a problem, how can we use data to zoom in on the source of the problem.
- Recent happenings in the auto industry: recalls, new regulations …
*
Very old business connections
*
Questions for
Quality Analysis and Improvement
How serious is the situation? (recognize)
At what point in time?
How did it happen? (identify)
Fishbone diagram
What is a defect? (define)
Tire design, vehicle design, in process, in use?
Who/what is responsible? (evaluate)
Analysis: Pareto chart
Other tires also fail?
Could it have been prevented? (organize)
Cost of quality
Steps in future to prevent? (control)
Organizing, systems, ..
*
This are the typical steps towards quality analysis and improvement.
*
1999
- Ford: Profitable, Most Admired Corp.
Explorer fatalities 1 in 100 mil. miles
Other SUVs 1.3/100 mil. miles fatalities
Explorer: fewer rollovers relative to vehicle population
2 % of the market, 0.6% of rollovers in that market
- Firestone: Global Market Leader
“NHTSA received 46 complaints about Firestone tires. Over the same period, the agency logged 970 complaints about Goodyear tires and 725 gripes about Michelins.”
(“Firestone had a smaller market share but the complaints are still out of proportion.”)
Based on overall performance, the recall was unexpected.
*
The message is that the recall was unexpected based on overall performance
*
What are the risks?
Fatality rates
Air (1970) .3 / 100 million miles
Air (1997) .06 / 100 million miles
Passenger cars 1.6 / 100 mil. mi.
All compact SUV’s 1.3 / 100 mil mi
Ford Explorer 1.0 / 100 mil mi
100 deaths for 3.5 million Explorers on the road from 1990 to 2000
Roll-over fatalities (1995-1998)
Explorer 26 / 1 million vehicles
Jimmy 34 / “
Chevy 45 / “
Jeep 23 / “
This doesn’t look so bad. So what are the problems?
*
These are the original brochures for the AT tire in 1996 as given to the tire dealers. Notice the designed values at lower left corner
Firestone: Models Recalled
*
These were all recalled – went as OE on these trucks
*
Magnitude of Crisis 2000
- RECALL COST
6.5 million tires at $50 ($325 mil) + legal +liability +lost profits
= $800 million to $1 billion
Later data shows that these numbers were exceeded
- Brand Image
- Most profitable segment for Ford
- 40% of sales for Bridgestone
- …..
- Was this a real crisis? Was a recall necessary? Why? The rollover rate and fatality rates on Explorers were lower than the competitors.
*
Was it a crisis? Was it over reaction? You will get many answers. The question really is when did it become a crisis? When should the managers of either company have reacted?
*
What would you have done?
- 1989-1990 (design stage), new tire (1996)
- 1997: Claims (?), Venezuela
- 1998: 21 cases, Saudi Arabia
- 1999: Ford memo to Firestone
- 2000 February: TV Show
- 2000 May: 90 complaints + 4 deaths
NHTSA opens formal investigation
- August 9, 2000: issued recall notice for 6.5 million tires (went up to 13 million out of 47 million made in the 1990s)
- 2000 September: 2200 complaints + 103 deaths + 400 injuries
- Perspective: 3.6 million Explorers
(0.00286% deaths, 0.06% complaints = 2.86/100,000, 61/100,000)
When should they have reacted?: Ethics, Economics
Tread separation in Firestone tires on Explorers
Assignable cause
*
This time line is shown and students are asked when would they have reacted and why? One type of answer is when is the liability greater than the cost. The other type of answer is when there is a problem. These two viewpoints need to be sharpened as the case discussion proceeds. But before proceeding ask students HOW to detect such a problem? Will final inspection do? What should be the sample size? They can use a p*(1-p)/n type of estimate for variance and see how big n has to be.
*
What is a defect??
- There is a difference between a tire and an electronic component.
Tire Parameters:
Miles (types of miles)
Load
Temperature
Average speed at which car is driven
Tire pressure
- When is a tire failure the fault of the manufacturer and when is it the fault of the owner?
Which manufacturer?
Design Quality, Production Quality and Quality in Use
*
This is another major debate. When do you say a tire a defective? Main idea here is that there is design quality, quality in production, and quality in use. You could also briefly mention that Ford is responsible for the tire even though Firestone made it. This is the liability rule that was established in a famous case many years ago (80?) in a case when a tire burst and the vehicle hit a pedestrian. The judge ruled that that the vehicle maker and the tire maker were liable if the defect were found due to them and could have been prevented.
*
Data: Ford and Firestone
- The population of Ford Rangers is much smaller than Explorers.
- After accounting for the number of Explorers with Firestone tires and Goodyear tires, the ratios are clearly out of proportion.
*
The Firestone argument is weak because the relative population of rangers is much smaller. Moreover their use may be different. Ford’s argument controls for the vehicle as a cause of defect
*
Isolating Causes by Controlling for other variables.
*
Once again the lower right picture is key to showing that tires are probably a major cause. Lesson: control for other variables to isolate cause of defect
*
What contributes the most to it?
Analysis performed by Ford in early August 2000
Age is not controlled. (Not a valid Pareto Chart.)
*
Most of the problems are cause by a handful of causes. Not all these are valid Pareto charts (the last one is not because of age not being controlled). You may also refer to Fig 6 in case.
*
Tread Separation: Industry Wide Problem
What was the additional pressure on Firestone to recall?
Causes for defects: Random vs Assignable (Attributable)
| Manufacturer | Recalled Tires | Fatalities | Injuries | Complaints |
| Firestone | 6,500,000 | 150 | 500 | 3000 |
| Goodyear | No Recall | 15 | 125 | —– |
| Continental – General | 160,000 | 0 | 0 | —— |
Management is responsible
Reason to believe that there is a process problem
*
Thus, we ask okay there is a problem but all tire makers have problems. So why bother?
The main idea here is random versus assignable cause. Mother nature versus Ford or Firestone management attributable defect. Thus partly answers the question when the recall should be made – when there is a sufficiently strong reason to believe that there is a process problem. More on detection techniques later …
*
Customer Focus: What would customers say about the problem ? Is the product too risky?
- Expectation
- Benchmark:
UK Road (0.6/100 mil)
Air Traffic (0.29/100 mil in 1970’s now 0.06)
Severity + Frequency of Occurrence + Detection = Risk
*
Here we try to give a measure of risk that combines all the ideas so far. The importance of doing so in the world economy with enormous volumes and even ppm being insufficient to avoid injury and death should be made. Thus set the stage for later discussion of six sigma – it is not just the process but the customer needs and impact on customers – how to make the two meet? These are the two faces of quality – process and use.
Acceptable risk levels:
Mother NatureAssignable problem
Random-cause known
Customers willing to take risk-get caught
Variation inherent in processNow Mgrs. are responsible
They didn’t do anything when they found the problems!
*
Information on Goodyear
“Goodyear quietly holds 'silent recall’ ” : (July 01) http://www.eagletribune.com/news/stories/20001107/BU_001.htm
- “Goodyear … has been quietly offering free replacements for thousands of tires in a practice known as a "silent recall,"
- Goodyear acknowledged last month that tread separation involving its 16-inch Load Range E light-truck tires has been linked to 120 injuries and 15 deaths. Company officials have said a recall wasn't necessary because it found no defects in the tires.”
- “Tires Under Fire: Goodyear Tires Draw Lawsuits, But No Recall” (May 03)
http://abcnews.go.com/sections/GMA/US/GMA030519Tire_investigation_hunter.html
- “NHTSA … 87 reported crashes involving Goodyear Load Range "E" tires between 1991 and 2001. … low percentage of the millions sold, but … — 158 injuries, in all — and some were serious. Eighteen people died.”
- “the company said that …, tread separation was caused in every case by either overloading of the vehicle, under-inflation of the tire, or other misuse of the vehicle.”
*
Thus, we ask okay there is a problem but all tire makers have problems. So why bother?
The main idea here is random versus assignable cause. Mother nature versus Ford or Firestone management attributable defect. Thus partly answers the question when the recall should be made – when there is a sufficiently strong reason to believe that there is a process problem. More on detection techniques later …
*
Another Way to Identify Causes Using Data: Fishbone or Cause-Effect Diagrams
- Example: Late flight departures
Using data, find the frequency of occurrence of each of these causes.
Delayed Flight
Equipment
Other
Material
Procedure
Manpower
Mechanical Failure
Late Push-back tug
Weather
Traffic
Late Food service
Late Baggage
Late Fuel
Poor Announcements
Delayed Check-in
Confused Seat Selection
Oversize Baggage
Gate Agents Cannot Process Passengers Fast
Too few agents
Under-motivated
Why does tread separation happen?
Cause-Effect Diagram or Fishbone Diagram
*
Another useful technique is to do the cause-effect diagram. From the case this can be constructed. Then one may put numbers against each cause and isolate the most frequent cause. Lesson 2: Focus on causes. Use measurements to determine the important cause
316.bin
*
Quality Problems due to new combination and applications of existing products
- Industry was not doing quality control the way they do it for trucks (each and every tire is checked and x-rayed)
- Drivers are operating SUV’s as if they are cars (high speed, high load, may not check tire pressure regularly)
- Characteristics of SUV’s
Higher center point of gravity
Higher propensity to roll over
*
The main point is the last one. Despite every manufacturer having access to same technologies how does Dell manage to put out a superior product? Maybe leave it as a question?
*
Ford-Firestone Interaction
- Firestone recommended 30 psi tire inflation
- Ford recommended 26 psi.
- When two different groups or products are combined, there is a need for or a value to over-designing the individual products or components.
*
KHOU-TV
- Investigative reporting awards for the Ford-Firestone coverage
- Example:
- … KHOU was the first to connect the dots between individual auto accidents and expose the "big picture." They found 30 deaths in Texas alone … began looking at similar incidents in other states. All of the accidents were the result of a problem called "tread separation," and all of them involved Firestone tires on Ford Explorers.
- Center for Auto Safety president Clarence Ditlow said, "Up until then it was a story about a local accident, and no one ever put it all together until your station did. I don't know that the problem would have ever been fully exposed without your report."
- Even Jacques Nasser, CEO of Ford, praised the station's investigative work. "They deserve a medal, actually," he said. "[KHOU] started everyone to think maybe there was really something there."
*
Testimony of Bridgestone/Firestone Inc.
(: Sep 2000)
- A small percentage of recalled tires have experienced tread belt separations in a number of serious accidents.
- Tread belt separations are usually caused by tire damage, underinflation, or simply by using worn-out tires, but separations can also be caused by defects.
- … working very hard to determine the root cause or causes of the problem. The process is … difficult because … a very small percentage of a large population of tires … highly variable operating conditions, which also affect tire performance.
- Firestone vigilantly monitors data on the in-service performance of its tire lines. We do product testing; we study warranty adjustment data; and, where possible, we analyze failed tires returned from the field.
- … satisfactory performance on the part of these tires. … passed Firestone internal design and development testing and Federal Motor Vehicle Safety Standard testing, as well as Ford's development and test track requirements.
*
Testimony of Bridgestone/Firestone Inc.
(: Sep 2000)
- February 2000, … KHOU ran a report on tread belt separations of Firestone ATX and Wilderness tires and their involvement in Ford Explorer rollovers. Following …, Firestone received an increased number of claims and lawsuits, the most serious of which seemed to be occurring in the warmest climates in the United States.
- In May 2000, the NHTSA began a Preliminary Evaluation of certain tires including the radial ATX and Wilderness AT lines. Following commencement of that Preliminary Evaluation, Firestone received notice of an even larger number of claims and lawsuits involving tread belt separations on Firestone tires mounted on Ford Explorers.
- Ford performed a statistical analysis using Firestone's data. The conclusion … was that the tires that eventually became the subject of the recall were overrepresented in the claim data. Tires manufactured in the Decatur plant were also overrepresented.
*
Testimony of Bridgestone/Firestone Inc.
(: Sep 2000)
- … number of serious accidents involving tread belt separations that surfaced after the onset of the NHTSA preliminary evaluation, and… Ford's analysis, … initiate the voluntary recall.
- Firestone initiated the … recall without identifying or pinpointing any particular cause or explanation for the tread separations and did so even though none of the yardsticks typically relied upon to measure tire performance indicated that the recalled tires were unsafe.
- …, the tire industry and the NHTSA have not historically used claims data in evaluating performance …, in hindsight, Firestone should have looked more carefully at claims data sooner than it did.
*
Testimony of Bridgestone/Firestone Inc.
(: Sep 2000)
- Firestone did not conduct any high speed testing … at 26 P.S.I. prior to the introduction of the product line in conjunction with the Explorer. It was Firestone’s understanding that Ford conducted testing of the high speed performance of these tires at 26 P.S.I .
*
Quality in Automobile Industry Today
- Wall Street Journal (March 4, 2004): “One in Twelve Cars Recalled Last Year”
“Car makers say …increase…doesn’t mean quality standards have fallen
proof of indutry’s greater commitment to safety
Advanced computer systems …spot problems sooner
76% of recalls in 2003 were voluntary; manufacturers initiated them without any involvement by NHTSA…
Internet has made it easier to report defects, and cars are more complex than they once were
Critics contend…that manufacturers are cutting corners to lower costs and speed development
*
Quality in Automobile Industry Today
- Wall Street Journal (March 4, 2004): “One in Twelve Cars Recalled Last Year”
… tougher safety laws in response to a rash of rollover deaths in Ford Explorers…with…Firestone tires.
… domestic auto makers …disproportionate share of the vehicles recalled.
Ford-Firestone …led to sweeping auto-safety law …criminal penalties of …15 years in jail …executives who knowingly hide information about product defects.
…new rules…car makers…submit…claims and warranty data
These rules were instrumental … Firestone … last week… voluntarily recall 300,000 tires…Ford Excursion…”
END OF FIRESTONE CASE DISCUSSION
*
How do we define Quality?
What does quality mean?
How do you define it?
In a restaurant? For an airline? At Donner?
Some defects are obvious, can be touched or seen
Others more difficult to define.
- TV shows: example, “Friends” series.
How do they measure quality?
- Quality: Exceeding customer expectations
*
Issues/Concepts
- Quality can make or break a company.
Profit = Revenue – cost of production – cost of quality
- So can an unethical decision.
How do we eliminate fear? Accept blame?
- For certain products, quality is
SUBJECTIVE, HARD TO MEASURE and DIFFICULT TO TEST FOR
- When the interaction of two (or more ) products is unknown,
then the quality of the two products should be OVER-DESIGNED for both
- What is a defect?
Product combination, design, process, testing, fitness for use
- How should data be stratified?
Pareto analysis.
- Assignable and random causes?
Issues/Concepts
*
- How to tailor products for specific use and prevent misuse?
driving under-inflated tires
- How should a firm educate its customers with regard to product use?
(Important service quality dimension. Fail-safing. What sort of failsafing?)
(Example: Water filter)
Product liability.
- What quality systems are required when there is joint responsibility for product design etc?
(The issue is feedback delay.)
Issues/Concepts
*
Quality- What’s next?
- How to know when your process is:
creating defects
creating too many defects
out of control
Just experiencing normal random variation
- How do we find the defects?
Inspecting
Calculating yields of the process
Determining where to inspect
Statistical Process Control
Acceptable risk levels:
Mother Nature
Assignable problem
Random
–
cause known
Customers willing to take risk
–
get caught
Variation inherent in process
Now Mgrs. are responsible
They didn’t do anything
when they found the problems!
Posted in 