How to Create a Global Reliability Program

Like many organizations, pharmaceutical company Eli Lilly has been in pursuit of reliability for many years. Pinpointing when the journey started is difficult, since manufacturing is always asking for increased performance or throughput. However, usually the improvements are achieved through brute force: longer hours, faster response, stocking many spare parts, etc.

The desire to improve manufacturing output created the need to look for ideas to help reduce the risk of interruptions. Since the 1990s, several programs have been implemented at various manufacturing sites with variable success. A few examples of the programs include:

• Vibration routes and analysis to determine bearing health and to plan corrective actions before failure.
• Infrared routes and analysis to determine electrical distribution equipment health and corrective actions before failure.
• A lubrication program for correct lubricant handling and equipment lubrication.
• Failure mode and effects analysis (FMEA)/reliability- centered maintenance (RCM) on critical systems to determine appropriate maintenance strategies for equipment.
• Root cause failure analysis (RCFA) to determine how and why failure occurred along with actions implemented to prevent recurrence.
• Precision maintenance training to improve craft understanding and practices for the assembly and installation of equipment.
• Computerized maintenance management systems (CMMS) to capture equipment information and manage maintenance activity.
• Planning and scheduling to improve the efficiency of maintenance activities.

These programs have impacted Eli Lilly’s manufacturing business in various degrees. In some areas, the value generated by the program has been very high. In other areas, less value was created. Two critical attributes of successful implementations were determined. The first was the implementer of the program. These individuals had to be passionate in their belief that the program truly added value and that not implementing it would be a mistake. The second was if the receiving organization recognized the benefit of the program and required the benefits from the program. Unsuccessful implementations were missing one or both of these attributes.

As a corporation, overall reliability was not improving at the rate required for the changing business climate. Senior leadership recognized that a few manufacturing areas were delivering consistent reliability improvements while other areas were not. They started believing in the importance of improving reliability. Eventually, the manufacturing units that did not have reliability wanted to get it, and the areas that had some success wanted more. This was a significant change. The organization began to ask for a reliability program. Typically, reliability professionals were “pushing” programs into the organization. Now, manufacturing had created the desire for improved reliability by establishing a reliability philosophy and implementing the appropriate programs.

In 2011, a small reliability steering team was commissioned to develop the next-generation reliability process for Eli Lilly. The team was composed of successful reliability engineers and managers. The team’s goal was to define and document reliability as well as the governing principles and tools, prove concepts were valid through demonstration projects, and form a recipe for implementation.

Early in 2013, a “reliability book” was written, training was established, three demonstration projects were providing significant results, and a flexible rollout plan was developed.

Training and the Reliability Book

The purpose of the reliability book was to present a consistent methodology to improve equipment reliability in manufacturing. This document was written for personnel involved in the design, operation and maintenance of equipment used in manufacturing, both direct production and support equipment. This included all the interactions of people, processes, raw materials, spare parts and utilities associated with the manufacturing equipment. Although this focus was on manufacturing equipment, the principles described generally apply to all equipment.

Asking the Right Questions

Management behavior is critical for the success of a reliability-based culture. Merely asking certain questions during an unexpected equipment failure could send unintended messages to the organization. The following excerpt is based on Winston Ledet’s article, “How Does Plant Management – and Possibly Corporate Management – Enable Unreliability.” Given an equipment failure, these are the types of questions management has historically asked, along with their unspoken message that created the culture:

Questions more appropriate for a failure in a reliability-based culture would be:

The reliability book consists of eight chapters and several appendices in 93 pages. It describes Lilly’s reliability vision, important reliability principles and models. Three training classes accompany the book: Reliability Overview, Reliability Leadership and Reliability Book Understanding. The overview class is intended for all employees involved in the manufacturing of products. The class provides general training on the content of the reliability book. The leadership class offers training for all employees in positional leadership roles. The class is an open discussion with leadership to discuss behaviors and methods to advocate and support the content of the reliability book. The final class is simply a test to verify reading and understanding of the reliability book.

Demonstration Projects

Three demonstration projects were performed in 2011-2012 to help validate the reliability tools and concepts that were included in the reliability book. The first project consisted of improving net output from a device assembly work center. In 2010, the demonstrated production rate of the manufacturing line had a standard rate of 77 units per minute. By the second quarter of 2012, the same manufacturing center had a demonstrated standard rate of 125 units per minute. This was a 60-percent improvement in output - a record output. The production output value to the business far exceeded any change in maintenance costs. Examples of projects that impacted reliability included a redesigned glue system, an improved changeover time from 4.3 hours to 3.5 hours and better service practices of the glue system and robot head.

The second project involved improving the output from a vial-filling work center. In 2011, the demonstrated production rate of the manufacturing line had a standard rate of 190 units per minute. By the second quarter of 2013, the same manufacturing center had a demonstrated standard rate of 230 units per minute. This represented a 21-percent improvement, which was a record output. Examples of projects impacting reliability included scale modifications, automated tank filling, improved changeover times, better setup instructions, enhanced machine knowledge and run rules, and better service practices.

The third project consisted of improving the mean time between failures (MTBF) of a rotary screw conveyor. This asset showed up on the site’s list of 20 worst MTBF assets. Upon investigation, it appeared the conveyor was tripping off-line because the downstream transfer line was plugging. Further research determined the reason for the line plugging was due to the automation sequencing of the various system assets. Once this sequencing was changed, the MTBF increased by a factor of four. The problem wasn’t with the screw conveyor; it was just the first visible sign of a downstream problem.

The Importance of Culture

Any change must take into account the effect on the culture. The prevailing culture may help or hinder the desired changes. An entire chapter of Eli Lilly’s reliability book was dedicated to the cultural challenges of moving a successful reactive manufacturing organization toward a proactive philosophy.

The nature of a reactive culture is described in both the reliability book and in the reliability training. The first diagram below shows a reactive culture in action. This culture accepts equipment failure as a normal event. The organizational belief is that equipment is going to fail. The goal is to fix the equipment as fast as possible to get production back into operation. The repairs generally do not address the root cause of the failure but have only a short-term focus on fixing what broke. Once the equipment is running, the mechanics, supervisors and engineers receive praise and recognition for the prompt fixing of the problem and then move onto the next problem. These actions translate into eventual raises and promotions. This behavior establishes reactive craftpersons, supervisors, engineers and management as role models for the manufacturing organization to follow, which reinforces the reactive culture.

In contrast to the reactive culture, the second diagram below represents a proactive culture. It starts with various reliability-based actions. These actions include many things such as equipment design, equipment setup/changeovers, predictive maintenance, operating techniques, equipment walk-throughs, world-class lubrication, etc. Unfortunately, these reliability actions, by their nature, are much less visible than a failure in the reactive culture. For example, a pump failure is much more visible than inspecting incoming lubricants for possible contamination.

The outcome of these reliability actions is fewer equipment failures. This will take time to be noticed, as there may be a significant time delay of up to several years. This delay, along with the lesser visibility of reliability-based actions, is one of the reasons why managing a reliability-based culture is much more difficult than managing a reactive culture.

Rewards and recognition should follow the improved equipment performance. However, due to the amount of time required before the results of equipment improvement may be seen, rewards and recognition may be better tied to performing reliability-based actions. For example, recognition might be given to a craftsperson for completing a good root cause exercise or proper recording of a precision alignment following an intervention. Another example would be to recognize the person who detects and reports an early defect before it becomes a significant failure or quality issue. As these people are identified and recognized, they become role models for others to follow. They start performing and valuing reliability-based actions. This in turn reinforces the culture around reliability principles and values.

For the proactive culture to survive over the long term, it takes significant management energy as well as a strong commitment from all levels of the organization.

Leadership

The role of site management is essential for the successful creation of a reliability-based culture. Merely having management support reliability is insufficient. The changing of a reactive culture will take considerable time, energy and focus. Site management must be passionately engaged in order to overcome the organizational inertia built around a reactive culture. In the past, companies that were successful in making the proactive transition were under intense economic pressure - they had to become reliable or go out of business. If management energy is removed, the culture will quickly return to being reactive.

There are many similarities between a proactive safety culture and a proactive reliability culture. In the past few decades, new ways of thinking about safety have emerged. The previous safety culture assumed that accidents “just happened” and that safety was the concern of the safety department. Now, many sites have adopted a culture that views accidents as preventable and an injury-free workplace as achievable. In order to reach the goal of an injury-free environment, safety had to become a major concern of everyone and not just the safety department. In a similar manner, this same type of cultural shift also applies to a reliability-based culture.

Reliability Metrics

Three key global metrics for reliability were defined: downtime, mean time between failures (MTBF) and deviations.

Downtime

Downtime (along with its inverse, uptime) provides an indication of manufacturing performance. More important than the downtime/uptime value are the reasons for downtime. The manufacturing area must collect and analyze the causes of downtime. The organization then should focus resources to investigate and resolve the significant contributors to downtime.

Downtime measurement across the various manufacturing sites was inconsistent and underutilized. Some sites were measuring downtime but were using different parameters. One site had more than 50 different downtime reasons. Others were measuring downtime only during a production run.

To standardize this important metric, a list of 12 reasons for downtime was developed under two broad headings: unscheduled downtime and scheduled downtime. The metric is also based on a 24-hours-a-day, seven-days-a-week schedule or 8,760 hours per year. This way, all reporting times and percentages have the same common denominator. By using this metric, the chronic contributors to the loss of production can be determined, and the correct resources can be applied to the problem area.

MTBF

MTBF is used to identify assets that are experiencing high rates of failure. Assets are ranked from worst to best, and the worst 20 assets for each site are reported. Some of these assets have problems that are well-known, while others may be a surprise to the organization. The goal is to improve reliability by reducing their failure rate. Sometimes the downtime data and MTBF report overlap, while at other times the failures may have little impact on the overall manufacturing output. Both metrics should be reviewed to establish the best use of resources.

Deviations

Manufacturing upsets or failures are tracked as non-conformances. Each non-conformance must be investigated to determine the impact on the product and the corrective action required to reduce or eliminate recurrence. Tracking non-conformances helps identify the chronic equipment issues that are causing the upsets in the manufacturing processes and offers another way to focus on the high-priority equipment problems requiring further investigation and remediation. These upsets are usually the significant contributors to floor and factory loss.

Reliability Implementation Plan

Traditionally, reliability efforts have been implemented as programs. Various reliability concepts such as predictive maintenance, planning and scheduling, etc., have been applied as stand-alone programs or projects. All of the programs are elements needed to support the reliability model. Each will have an impact on the manufacturing output. However, universal organization support is difficult to achieve from the operational partners. They do not always see the immediate value, as manufacturing output may not significantly change.

The philosophy for Eli Lilly’s implementation plan was to have each site focus on improving the reliability of an asset or line that was impacting the site’s performance. By targeting chronic failures and the weaknesses limiting manufacturing output, significant improvements can be made. The organization is focused on the important and urgent issues. Employees have a higher sense of value by working on critical issues, and the plant’s net output increases. The reliability professional earns the credibility to continue the reliability journey.

Eli Lilly operates more than 20 manufacturing sites around the world. All of the sites are at different places in their reliability journey. Each site also has cultural or organizational differences. The implementation plan is not a one-size-fits-all approach. Instead, flexibility was built in the rollout plan to allow for sites to adjust the concepts to fit their business needs. The reliability book provided the framework for the sites to build their program.

For 2013, each manufacturing site was asked to perform the following:

• Name a reliability champion from the management ranks to be the site champion and point of contact
• Determine the site’s 2013 reliability plan, including a reliability- improvement project on a troublesome asset based on any combination of the three metrics discussed earlier, and a training plan involving the three reliability courses and a subset of the site personnel.
• Measure downtime on a 24-hours-a-day/seven-days-a-week basis and categorize into Pareto charts.
• Participate in global reliability forums.

The sites have responded well to this plan. Several have already completed their reliability-improvement project and are demonstrating improved performance of the asset or line. All sites are in various stages of training, and many people outside of engineering and maintenance are talking about reliability as a part of their normal job functions.

Keep in mind that reliability is an outcome. It is more than maintenance. It is the result of how equipment is selected, installed, operated, serviced and improved. The consequence of all this activity is the probability that the equipment or systems will perform their designed functions correctly when needed.

Changing the culture means eventually changing everyone. Most organizations have previously been successful with a reactive culture. The reactive culture is natural and normal to many people. However, the changing business environment means that your assets must produce more cost-effective, high-quality products. It will take significant leadership and management energy to change the culture from its formerly successful reactive culture to a proactive culture. Remember, the ability to improve long-term reliability business practices will be earned through sustainable manufacturing output improvements.