Aim for Operational Excellence

But first understand what targeting that goal involves and what can go wrong.

By Bernard McGarvey, Eli Lilly and Company

3 of 3 1 | 2 | 3 > View on one page

Many years ago, we had a process that wasn't delivering the throughput required. This generated a lot of emphasis on the process. Many people were involved and were working really hard to improve the throughput — but with limited success. Then, a new manager for the line started and changed how the process team worked. This led to a significant improvement in the throughput.

He enhanced the level of OE in the process and the team in several ways:

• He changed the structure within the team so resources were more focused on certain tasks. The first part of the team concentrated on the day-to-day running of the process and dealing with minor issues that could be fixed quickly and easily. The second part zeroed in on solving the major special-cause variation problems while the third part tackled the major common-cause variation problems. The manager did his best to ensure that people kept focus on a problem until it had been resolved, not jumping from one issue to the next.
• He instituted a program to improve people's understanding of process variation so they were better able to differentiate between special- and common-cause variations. As part of this, he hired an external consultant to provide objective feedback on what they were doing and how they could do better.
• The process had a very good system for collecting data. However, the system's capabilities weren't being exploited. So, he spent time defining the ways in which people should look at the process data and how to present these data to the process team in ways that facilitated easy interpretation. This helped, for example, to better distinguish between special- and common-cause variations.
• He questioned the existing mental models about the impact of input variation on the variability of the process. For example, because there was a good set of work instructions, it had been assumed that people had minimal impact on variation. However, control charting the throughput data as a function of shifts showed that all the shifts were performing differently. This spurred a renewed focus on having all operators perform all tasks consistently — and had the added benefit of improving general communications between shifts, which is very important when it comes to reducing process variability. It also had been assumed that the variation of the input raw materials was insignificant. However, he was able to demonstrate this wasn't the case for one key input raw material; modifying the control strategy around this raw material led to improvement.

OE is an objective that can be defined in terms of expectations for high levels of process performance. However, it also must encompass the behaviors that support and sustain such enhanced performance over long periods of time, engraining them into the DNA of the organization. These behaviors only will occur if certain key elements (as outlined by the 7S model) are improved as part of the journey towards higher levels of OE.

BERNARD MCGARVEY is a senior engineering advisor for Eli Lilly and Company, Indianapolis, Ind. E-mail him at

1.    Deming, W. E., "Out of the Crisis," M.I.T. Press, Cambridge, Mass. (2000).
2.     McConnell, J. S., "Analysis and Control of Variation", 4th ed., American Overseas Book Co., Norwood, N.J. (1987)
3.     Wheeler, D. J. and Chambers, D. S., "Understanding Statistical Process Control," 2nd ed., SPC Press, Knoxville, Tenn. (1992).
4.     Waterman, R. H., Peters, T. J. and Philips, J. R., "Structure is not Organization," Business Horizons, Vol. 23, No. 3, pp. 14–26 (1980).
5.    Wheeler, D. J., "Advanced Topics in Statistical Process Control," SPC Press, Knoxville, Tenn. (1995).
6.    McConnell, J. S., Nunnally, B. and McGarvey, B., "Meeting Specifications Is Not Good Enough — The Taguchi Loss Function," J. Validation Technology, Spring 2011.
7.    Deming, W.E., "On Some Statistical Aids Toward Economic Production," Interfaces, Vol. 5, No. 4, pp. 1–15 (1975).
8.    Mitchell, T. M., "Machine Learning," McGraw-Hill, New York City (1997).
9.    Shmueli, G., Patel, N. R. and Bruce P. C., "Data Mining for Business Intelligence," 2nd ed., John Wiley & Sons, Hoboken, N.J. (2010).

3 of 3 1 | 2 | 3 > View on one page
Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.


No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments