Applying Continuous Improvement (CI) Manufacturing

Traditionally, continuous improvement (CI) activities at Lonza were carried out and managed by staff who were also responsible for the delivery of product to internal and external customers. The CI projects usually competed for resources with product delivery. To address this issue, the potential benefit of having dedicated resources focused on improving working practices and procedures was often discussed. Separating direct accountability for product delivery from CI efforts alleviated this conflict.

In May 2003, Lonza restructured the manufacturing department and created two full-time, dedicated positions focused on CI activities in manufacturing. The new roles provided a resource for CI projects designed to achieve specific business goals. To highlight the importance of the projects, the CI leads reported directly to the head of operations.

The CI lead role was envisaged as a long-term secondment, allowing the skills and knowledge obtained from executing CI projects to ultimately be transferred back to the product delivery teams. This was done to ensure that CI did not become just the job of the CI leads, although the CI leads were responsible for driving and ensuring successful project execution. To allow the CI leads to be successful in their roles, training and coaching on CI techniques was provided by an in-house Six Sigma Master Black Belt. As well as executing projects, part of the CI lead's role was promoting the CI philosophy and encouraging the use of various CI techniques throughout the organization. By exposing as many staff as possible to the philosophy, Lonza hoped that CI would become part of the staffs' everyday work to ensure that the company continues to deliver value to its customers.

Lonza's CI philosophy is based on five main principles.

1. Understanding Process Variation

The first principle of the philosophy is understanding process variation and emphasizing on-target performance with minimal variation. Control charts are useful tools for monitoring process variation over time and determining if a process is in control (stable, predictable, and exhibiting only random, "common cause variation") or out of control (unstable, unpredictable, and showing "special cause variation").

¹

Traditionally, the number of deviations reported from one manufacturing batch was compared to the number reported for the previous batch. If the current batch generated more deviations than the previous batch, an investigation was carried out to try to understand why. Conversely, if the number of deviations generated was less than the previous batch, the teams were considered to have performed particularly well. In the past, there was no way to determine the typical level of variation in the process and, therefore, how many deviations could be expected. Control charting is now used to monitor the number of deviations generated by each batch. This has allowed Lonza to target investigations at batches that truly have special cause variation rather than wasting time investigating random and expected variation.

Figure 1. This control chart shows a shift in the process performance (a reduction in the mean number of deviations generated per batch) after implementing an on-plant review of documents. New control limits will remain temporary until more batches are completed using the new process.

Control charting also helps track and evaluate process performance over time. This is particularly useful after process improvements have been made to ensure the changes have a beneficial effect. For example, changes were recently made to the timing of batch record reviews. The review of manufacturing batch records now takes place in the plant, usually on the day they are completed. Before, the documents were reviewed in an office several days after they were completed. The improved process has reduced the time between document completion by manufacturing and the subsequent review by an independent lot review group and quality assurance (QA). One anticipated benefit of this change was reduction in the number of deviations generated per batch by having QA input available at the plant as batches are being processed. As control charts were already being used to monitor the process, the effects of this change in procedure were quickly evaluated. Changing the timing of the document reviews resulted in a reduction in the mean number of deviations generated per batch (Figure 1).

The successful use of this technique has led to the introduction of control charts to track the performance of other processes. For example, the number of process alarms generated during manufacturing is now monitored using this tool.

2. Error Prevention

The second principle of the CI philosophy is preventing errors — not just applying corrective actions once they have occurred. This has been achieved using a technique called failure mode effect analysis (FMEA). FMEA is a risk-based prioritization tool that helps to identify known or potential failure modes within a process or system.

²

FMEA has been successfully used for one of Lonza's fermentation harvest systems. The technique helped identify and prioritize areas requiring CI focus, engineering investment in the plant, additional operator training, and opportunities for standard operating procedure improvement. A secondary benefit of using FMEA is that the team (consisting of representatives from fermentation, engineering, and pilot and QA compliance) has enhanced its understanding of the harvest system. The team felt that the knowledge they gained would be useful to them in their normal roles.

Figure 2. Drill blanks have been mistake-proofed by attaching the correct size tubing for which the blanks are intended to be used. This helps to ensure that the correct size blank is used for the appropriate size tubing.

3. Structured Problem Solving

The third continuous improvement principle promoted at Lonza is using structured problem solving to help identify root causes of problems more rapidly. Cause and effect analysis has been used to structure problem solving by graphically displaying all possible causes of a problem. Brainstorming sessions are held with members of the relevant teams to generate the list of potential causes. It is essential to involve the people working on the system or process as they possess the best knowledge of likely causes of the problem. It is also useful to include an "outsider" at these sessions because of their ability to ask "obvious" questions that people with in-depth process knowledge frequently overlook. Once a cause-and-effect diagram has been generated, all the possible causes of a problem are investigated. Potential causes are only discounted if factual evidence can be found to back-up that decision. This helps remove subjectivity from investigations. Previously, individuals often presented their opinions as facts, leading to unsubstantiated conclusions being drawn about the actual root cause of a problem.

As well as a more rapid determination of root causes, this technique is also a useful method of documenting an investigation. This helps with the closure of associated manufacturing deviations as it provides clear evidence that an event has been fully investigated. The tool is now a standard problem solving and investigative method in manufacturing.

4. Eliminating Root Causes

Once root causes of problems have been identified, it is essential that actions are taken to ensure that they cannot recur. Root cause elimination is the fourth CI principle used at Lonza. Root cause elimination is achieved by using mistake-proofing techniques that have already proved successful in the automotive and aerospace industries.

²

Mistake-proofing allows systems and procedures to be designed or modified to help avoid the opportunity for error.

Figure 3. Instrument port blanking plugs used in fermenter vessels have been color coded to ensure that the correct plugs are used and that the plugs are correctly aligned.

For example, the root cause of a pump tubing split at Lonza was attributed to the incorrect setting of pump roller gaps. Gap measuring devices (blank drill bits) are now being used to ensure that pump rollers are set to the correct operating distance. These gap measuring devices were developed so that it is obvious which size tubing should be used (Figure 2). This has helped reduce opportunities for the wrong operating distance to be used when pumps are set up.

The mistake-proofing technique has also been successfully used to ensure that the instrument port blanking plugs used in fermenter vessels are correctly positioned and aligned prior to commencing a manufacturing batch (Figure 3). Now that the mistake-proofing system is in place, it is obvious to operators, at a glance, if a plug is correctly oriented.

5. Knowledge Transfer

The final principle in Lonza's CI philosophy is the transfer of key learning and knowledge to other systems, processes, and people. Transferring information can potentially prevent the same issue from occurring in other departments or facilities. This transfer of information helps reduce the resources that other departments and facilities spend trying to resolve the same or similar issues. This has been particularly beneficial where multiple facilities perform similar operations.

Due to the success of the manufacturing CI leads over the last 12 months, two new positions for engineering CI leads have been established at Lonza's UK operations. After a 10-month secondment, the two initial manufacturing CI leads moved to work on Six Sigma Black Belt projects and were recently replaced by two new manufacturing CI leads. Despite the relatively short time that the CI lead role has existed, the benefits of having dedicated resources are clear. Lonza is convinced that this investment in dedicated resources has produced and will continue to produce real returns in improved operational efficiency and effectiveness.

References

1.

Wise SA, Fair DC. Innovative control charting: practical SPC solutions for today's manufacturing environment. Milwaukee (WI): ASQ Quality Press; 1988.

2. London A. BioPharma Operations Excellence. BioPharm International 2003; 16:18-19.