2.7 Examples of DMAIC
2.7.1 Litigation Documents
Litigation usually creates a very large number of documents. These can be internal work
papers, consultants’ reports, affidavits, court filings, documents obtained via subpoena, and
papers from many other sources. In some cases, there can be hundreds of thousands of documents
and millions of pages. DMAIC was applied in the corporate legal department of
DuPont, led by DuPont lawyer, Julie Mazza, who spoke about the project at an American
Society for Quality meeting [Mazza (2000)]. The case is also discussed in Snee and Hoerl
(2005). The objective was to develop an efficient process to allow timely access to needed
documents with minimal errors. Document management is extremely important in litigation;
it also can be time-consuming and expensive. The process was usually manual, so it was subject
to human error, with lost or incorrect documents fairly common problems. In the specific
case presented by Mazza, there was an electronic data base that listed and classified all of the
documents, but the documents themselves were in hard copy form.
Define. The DuPont legal function and the specific legal team involved in this specific
litigation were the customers for this process. Rapid and error-free access to needed documents
was essential. For example, if a request for a document could not be answered in 30
days, the legal team would have to file a request for an extension with the court. Such extensions
add cost, time, and detract from the credibility of the legal team. A project team consisting
of process owners, legal subject-matter experts, clerks, an information systems specialist,
and Mazza (who was also a black belt in Dupont’s six-sigma program) was formed.
The team decided to focus on CTQs involving reduction of cycle time, reduction of errors,
elimination of non-value-added process activities, and reduction of costs. They began by
mapping the entire document-production process, including defining the steps performed by
DuPont legal, outside counsel, and the outside documents-management company. This process
map was instrumental in identifying non-value-added activities.
Measure. In the measure step, the team formally measured the degree to which the
CTQs were being met by reviewing data in the electronic data base; obtaining actual invoices;
reviewing copying and other labor charges, the costs of data entry, and the charges for shipping,
court fees for filing for extensions; and studying how frequently individual documents
in the data base were being handled. It was difficult to accurately measure the frequency of
handling. Many of the cost categories contained non-value-added costs because of errors,
such as having to copy a different document because the wrong document had been pulled
and copied. Another error was allowing a confidential document to be copied.
Analyze. The team worked with the data obtained during the measure step and the
knowledge of team members to identify many of the underlying causes and cost exposures.
A failure modes and effects analysis highlighted many of the most important issues that
needed to be addressed to improve the system. The team also interviewed many of the people
who worked in the process to better understand how they actually did the work and the
problems they encountered. This is often very important in nonmanufacturing and service
organizations because these types of operations can have a much greater human factor. Some
of the root causes of problems they uncovered were:
1. A high turnover rate for the contractor’s clerks.
2. Inadequate training.
3. Inattention to the job, causes by clerks feeling they had no ownership in the process.
4. The large volume of documents.
The team concluded that many of the problems in the system were the result of a manual document-
handling system.
Improve. To improve the process, the team proposed a digital scanning system for
the documents. This solution had been considered previously, but always had been discarded
because of cost. However, the team had done a very thorough job of identifying the real costs
of the manual system and the inability of a manual system to ever really improve the situation.
The better information produced during the measure and analyze steps allowed the team
to successfully propose a digital scanning system that the company accepted.
The team worked with DuPont’s information technology group to identify an appropriate
system, get the system in place, and to scan all of the documents. They remapped the new
process and, on the basis of a pilot study, estimated that the unit cost of processing a page of
a document would be reduced by about 50%, which would result in about $1.13 million in
savings. About 70% of the non-value-added activities in the process were eliminated. After
the new system was implemented, it was proposed for use in all of the DuPont legal functions;
the total savings were estimated at about $10 million.
Control. The control plan involved designing the new system to automatically track
and report the estimated costs per document. The system also tracked performance on other
critical CTQs and reported the information to users of the process. Invoices from contactors
also were forwarded to the process owners as a mechanism for monitoring ongoing costs.
Explanations about how the new system worked and necessary training were provided for all
those who used the system. Extremely successful, the new system provided significant cost
savings, improvement in cycle time, and reduction of many frequently occurring errors.
2.7.2 Improving On-Time Delivery
A key customer contacted a machine tool manufacturer about poor recent performance they
had experienced regarding on-time delivery of the product. On-time deliveries were at 85%,
instead of the desired target value of 100%, and the customer could choose to exercise a
penalty clause to reduce by up to 15% of the price of each tool, or about a $60,000 loss for
the manufacturer. The customer was also concerned about the manufacturer’s factory capacity
and its capability to meet their production schedule in the future. The customer represented
about $8 million of business volume for the immediate future—the manufacturer needed a
revised business process to resolve the problem or the customer might consider seeking a second
source supplier for the critical tool.
A team was formed to determine the root causes of the delivery problem and implement
a solution. One team member was a project engineer who was sent to a supplier factory, with
the purpose to work closely with the supplier, to examine all the processes used in manufacturing
of the tool, and to identify any gaps in the processes that affected delivery. Some of the
supplier’s processes might need improvement.
Define. The objective of the project was to achieve 100% on-time delivery. The customer
had a concern regarding on-time delivery capability, and a late deliveries penalty clause
could be applied to current and future shipments at a cost to the manufacturer. Late deliveries
also would jeopardize the customer’s production schedule, and without an improved process
to eliminate the on-time delivery issue, the customer might consider finding a second source
for the tool. The manufacturer could potentially lose as much as half of the business from the
customer, in addition to incurring the 15% penalty costs. The manufacturer also would experience
a delay in collecting the 80% equipment payment customarily made upon shipment.
The potential savings for meeting the on-time delivery requirement was $300,000 per
quarter. Maintaining a satisfied customer also was critical.
Measure. The contractual lead time for delivery of the tool was eight weeks. That is,
the tool must be ready for shipment eight weeks upon receipt of the purchase order. The CTQ
for this process was to meet the target contractual lead time. Figure 2.4 shows the process map
for the existing process, from purchase order receipt to shipment. The contractual lead time
could be met only when there was no excursion or variation in the process. Some historical
data on this process was available, and additional data was collected over approximately a
two-month period.
Analyze. Based on the data collected from the measure step, the team concluded that
problems areas came from:
1. Supplier quality issues: Parts failed prematurely. This caused delay in equipment final
testing due to trouble shooting or waiting for replacement parts.