Amgen, a global biotechnology company that develops therapeutics through advances in cellular and molecular biology, recently began a new program to evaluate and improve its existing manufacturing infrastructure. To effectively manage a rapidly evolving pipeline of new products, Amgen recognized that developing new tools would not only help manage manufacturing capacity but would also support the continued success of its existing and new products.

The first phase of this program was a project at Amgen’s clinical fill and finish facility in Thousand Oaks, California. Amgen contracted H. B. Maynard and Company, Inc., an industrial engineering consulting firm, to facilitate the project. A cross-functional team of Amgen staff members and Maynard consultants was formed to combine traditional and modern industrial engineering techniques to improve labor utilization, improve workflow, and increase overall output capacity.

Background

Amgen’s success is built upon its ability to rapidly discover, develop, and test new therapeutics in a highly regulated industry. Prior to FDA approval and market introduction, all Amgen products run through a series of clinical trials. Building 20, located on campus in Thousand Oaks, California, formulates and packages lots of new products for use by doctors and researchers in clinical settings. Production consists of four major processes: Buffer Preparation, Formulation, Filtration, and Filling. Product enters the building as bulk protein and leaves the facility in vials or syringes. In addition to these value-added steps, numerous preparation and cleaning processes are required to comply with FDA regulations. As the number of clinical products increased, Building 20 needed to maintain its manufacturing flexibility and efficiency while increasing its throughput.

Objectives

The capacity evaluation and improvement project had the following objectives:

• Complete a thorough analysis of Building 20 operations
• Develop comprehensive engineered standards for each of the processes
• Identify the capacity constraints (production bottlenecks)
• Identify and document improvements to streamline the workflow
• Create a tool for modeling manufacturing output capacity

Approach

Standards Development

The foundation of the project was the development of engineered standards for all Building 20 production operations. This measurement was the basis for identifying the production constraints and evaluating workflow improvements.

Due to the long duration of some tasks (1+ hours) and infrequent occurrence of some operations (3 - 6 times per week), the team used a combination of work measurement techniques. MOST®, stopwatch time study, and work sampling were all used to develop standard task cycle times. This allowed the standards accuracy requirements to be effectively balanced with the constraints of the project.

Workflow Improvement

In conjunction with the standards development, the team assessed current production capacity and various support operations. As a result of this assessment, capacity enhancement and streamlining measures were identified for post-project consideration and implementation. Throughout the project, potential improvement measures were documented, reviewed, validated, and considered for implementation. The team members differentiated between value-added and non-value-added steps of each task-level event. This allowed the team members to pinpoint areas in which significant improvements could be made. One specific suggestion resulted in the combination of production tasks. This combination resulted in the elimination of non-value-added preparation.

The team also reviewed plant support functions including maintenance, documentation, scheduling, and material ordering. The primary discovery was the need to re-organize labor to maximize the return on individual efforts. Changes made to Building 20’s ordering/documentation procedures provide a significant example. Prior to the study, multiple staff members were assigned similar duties resulting in duplication of work and lost production hours. A recommendation was implemented to assign all material ordering and batch record documentation to a single staff member. The impact of this recommendation resulted in a 20% increase of operational staff availability for value-added production operations.

Bottleneck Reduction

The work measurement and operation analysis confirmed that Fill Operation was the limiting factor, and many of the improvement recommendations addressed product throughput in this area. One recommendation was to continue the usage of a fill machine previously scheduled for removal from service. Building 20 had purchased a new fill machine capable of running larger quantities of vials at a faster rate than the existing machine. After completing the analysis, the project team suggested keeping the old machine and alternating production crews between the two machines. This internalized setup, tear down, and sanitation processes to the production of the other line.

The application of quick changeover/setup reduction techniques proved to be another major improvement area. In the fill machine process, setups routinely took two operators several hours. These setups were not clearly choreographed. The method also required that numerous small parts be removed and replaced to complete the set-up, and all adjustments were fine-tuned by trial and error. In summary, the setup and operation of the machine relied too much on an operator’s skill and intervention. As recommended, implementing labor reduction measures would help mitigate this reliance.

As for support operations, the team also identified several redundant steps in the sanitation function. The elimination of many redundant steps provided a 22% increase in sanitation workforce efficiency.

Capacity Modeling

The objective of the capacity model was to give Building 20 a tool to define process capacity based on several key variables. The project team designed a spreadsheet application to set accurate standards for 3,000 potential process variations. Once calculated, the model uses standards to schedule processes determine capacity, and track performance.

Scheduled standards were subtracted from available production hours at the plant’s bottleneck operation, the Fill Process. To maximize the plant’s capacity, the model compared standards for unscheduled production runs against available blocks of time at the bottleneck. Management currently uses this information to minimize lost capacity and track its available hours for production activities.

The following screen capture shows the model’s input screen:

Conclusion

The model gave Amgen a tool to quickly set standards for new products, accurately forecast costs, optimize the use of facilities production space, and calculate plant capacity. Many of the workflow improvement suggestions yielded significant productivity gains.

Creative application of classic and modern industrial engineering techniques results in greater internal efficiencies. Despite the constraints facing industrial engineers in the highly regulated, pharmaceutical industry, Amgen and Building 20’s Capacity Evaluation and Improvement Initiative proved that significant improvements in pharmaceutical manufacturing are attainable. Amgen’s provision of a motivated team, empowered employees, and supportive management culminated in a greater commitment toward continuous improvement. Additionally, Building 20’s control and capacity gains have provided Amgen with continued motivation in rolling out similar projects in other production facilities.