By Niraj Goyal

Background
Customer Delight (synonymous with Quality, referred to in this paper as the big Q) has 4 components: product quality, on-time delivery, service and cost.

 This case study demonstrates 3 significant aspects of Change Efforts:
 

1                     Improving three components of Q - product quality, delivery and/or service - inevitably leads to improvement of the fourth - cost – often much more effectively than pure cost reduction initiatives. This principle is embodied in the COQ concept.

2                     The Lean-Six Sigma combine is much more powerful than either technique individually. Real life problems typically consist of problems in all three components of Q simultaneously. Six-Sigma attacks defectives frontally while Lean attacks the problems of cycle times, on time delivery and faster service responses.

3          Time is a very effective overarching metric for measuring and directing change in all the dimensions of Q. It helps answer the often asked question – “Where do I start? Which technique should be applied first?” It relentlessly forces the blurring of artificial boundaries into seamless supply chains.

This case was carried out in the high speed printing operation of a large diversified, multi-location technical services provider in India. The plant, engaged in outsourced printing work for a variety of clients, struggled to achieve the turnarounds demanded. The project was the pilot “demonstration” project for Lean-Six Sigma and achieving the mindset change of operating personnel was as important as the result’s achievement for the company to commit more resources to the Q initiative.

The Case:  The narrative follows the chronological sequence of the seven steps of problem solving.  

Step 1              1.1)       Selecting the Theme (CTQ) – Following a 2-day introductory “Quality Mindset Programme” for senior management a brainstorm produced a long list of pain areas for the organisation. These were separated into two categories:

-  “End result” problems faced by external customers

-  Problems that were internal and therefore the causes of customer problems rather than intrinsic problems in themselves.

A quick discussion led to the realization that end result problems were the “real” problems (CTQs). 

Using the weighted average table followed by a quick discussion established the CTQ attribute -  “Consistency in Quality”.

1.2)       Selecting the project    - The management team then selected a location, a customer (Bank A), and the team for a pilot project. The job involved the printing, packing and despatching monthly bank statements for Bank A’s clients. The work was therefore sporadic (once per month) with unpredictable volumes.

 A “Quality Mindset Programme” for the team was preceded the project’s commencement. 

Step 2  Defining the Problem: a group brainstorm regarding problems in customer line A was prioritised. “Delayed turnaround” emerged as the critical customer problem. The problem was defined numerically as 

Problem = Customer Desire – Current Status 

Considerable debate was required for all to agree the start and end points of turnaround as “Arrival time of data on site” and “End of packing of all statements for despatch” 

Customer desire was ascertained from the Service Level Agreement (SLA) to be specified at two stages of the process:

Printing turnaround        8133 images per hour

Job turnaround              8133 images per hour

TAKT time                    0.43 seconds                                        (3600/8133 seconds)

For determining the current state one run was followed through for careful recording of throughputs and completion.

The TAKT time obtained was 1.35 seconds for a run of 56000 statements.

Problem: reduce TAKT time from 1.35 seconds to <.43 seconds i.e. 67%.

Phase 1 Objective:  reduce TAKT time 50%

Step 3              Finding Root Causes:  Construction of the “Current State Value Stream Map” (A powerful Lean tool) was undertaken to unearth problem areas for root cause and countermeasure analysis. This was done in 3 stages

i)                    Outline process stages

ii)                   Agree “Best estimate” standard times/cycle times expected for each activity

iii)                  Conduct a time study to ascertain actual times.  

Essentially the process has three major “loops”:

i)                     Pre-processing: processing customer data, printing samples and getting customer approval. This takes a fixed time independent of the volume and is customer dependent.

ii)                   Printing and Slitting – high speed automatic machines 

iii)                  Stuffing – labour intensive manual process with several stages

Based upon data collected the Current State Value Stream Map shown in Appendix 1 was constructed. The Root Cause/Countermeasure analysis for each loop proceeded as follows:

Loop 1 – Pre-processing

Step 3 Find the root causes In run 1 Pre-processing took 390 minutes as shown below:

Step 4 Generate Countermeasure Ideas: With each activity timed, three successive brainstorms resulted in an idea to reduce this time from 390 minutes <30 minutes. Essentially it meant asking the customer to transfer sample data in “Flow” mode before the main body of data arrived to enable approval to be before the turnaround clock started.

Step 5  Check effects:  In the third run 30 minutes was achieved. Time had helped create a seamless customer-vendor chain.

Step 6  Standardise the Process: this is in progess with the customer.

Printing and Slitting

 Step 3 Find the Root Causes: The problem solving went through the following cycle time (C/T) analysis using the classic Six Sigma cycle of improvement:

- Establishing ideal C/T achievable.

- Determine actual C/T through time study, and reason for all stoppages.

- Define the problem – (Ideal – actual) C/T

- Brainstorm root causes and countermeasures

- Implement countermeasures

- Check the results

- Repeat cycle

The pre project results are summarised in Table 1 below.

Table 1

Four cycles of Observation – data collection - root cause – countermeasure analysis were completed and implemented. Table 2 shows the analysis and Figure 1 the steady dramatic (50%) reduction of cycle time from .74 to .36 seconds.

Table 2

Figure 2 

The required C/T (.43) had been achieved from one machine. Besides 1 machine was doing the job of 2 machines.

Stuffing turnaround had five main components as shown below:

Pre-processing (1) and start of printing (2) had been reduced to almost zero minutes. 

Batch Processing between stages: stages of a process often leads to large turnaround times.

Pre project Slitting started 2 hours after printing due to batching in printing – printed material was passed onto slitting only on completion of “enough” material, as Slitting was much faster than Printing. Flow processing was introduced and it was agreed that half a batch (15 minutes) would be passed on to start slitting.

Likewise material from Slitting was passed on to Stuffing on completion of a half batch in 15 minutes as compared to 1.5 hours before the project.

With pre-processing reduced to zero this meant that stuffing would start 30 minutes after the arrival of data.

Check the result:  this result was achieved in successive runs as shown below in Table 3.

Table 3

Obtaining a stuffing C/T of .43 therefore became the central issue.

Stuffing was mapped through its six stages, which were run in batch mode. 2-3 large batches each of 1000 images envelopes were completed by manpower shifting between stages as the batch progressed. The C/T was .93 seconds and the productivity was 125 envelopes per man-hour against the desire of .43 seconds and 250 envelopes per man-hour.

Each operation was timed for its cycle time. A balanced line operating in Flow with batches of 50 was set up. The cycle times and the configuration of the flow line is shown in Table 4.

Table 4

The concepts were tested first on a limited one-hour trial and then a full scale run with dramatic improvements as shown in Table 5 below.

Table 5

Two teams of 9 operators obtained a C/T of .3 seconds with a productivity increase from 125 to 323 units per man-hour (i.e. 150%)

The problem had been resolved and the new process was represented in the Future State Value Stream Map shown in Appendix 2.

Summary of results:  the customer C/T .43 seconds was bettered to .36 seconds with a productivity of assets and manpower increasing by 100-150% depending upon the stages of the process as shown in Table 6.

Table 6

Conclusion: The judicious mixture of lean-six sigma techniques produced a much greater improvement than any one technique would have yielded. The Time metric clearly pointed the way to which approach’s tools were most appropriate at what stage of the project. Cost reduction though not targeted resulted automatically with the improvement of Q. Finally the company decided to continue the Q initiative.

Appendix 1 - Current State Value Stream map

Appendix 2 – Future State Map

About the Author: Niraj Goyal has 25 years of experience with multinational companies in various operating roles, among them operations director of Cadbury India Ltd., where he was among the leading implementers of the quality movement. He is the founder of Cynergy Creators Private Ltd. Mr. Goyal consults in India, the United States and Southeast Asia with manufacturing, IT, media and financial services industries. He specializes in training and facilitating the implementation of the techniques of Six Sigma-Lean and TQM. Several other real life case studies can be accessed on his website www.nirajgoyal,cjb.net
Mr. Goyal can be reached at nirajgoyal@vsnl.in.

Tell us what you think about this article. Send a note to the Editor.

www.manorhouseassociates.com