Robust Experimental Strategies for Improving Upstream Productivity

Identify the best experimentation methods for the data you need.
Jul 01, 2010
Volume 23, Issue 7


Experience using Quality by Design in upstream processes has identified several things that can improve the application of the method. When the experimental environment is diagnosed and strategies are developed to match the environment, experimentation moves more quickly and the critical process variables are identified with higher probability. This article presents a process for developing experimental strategies. The approach focuses on enhancing process understanding by developing the right data, at the right time, and in the right amount to maximize the time-value of the data collected. The development and operation of robust measurement methods that produce high quality data and streamlining of experimentation work processes also is discussed.

With the leadership of the FDA, there has been considerable focus on the use of Quality by Design (QbD) to accomplish the goal of speeding up development and reducing costs in research and development and manufacturing; producing better quality products and more effective, efficient, and robust manufacturing processes. Much has been learned about the use of QbD since it was first proposed by the FDA.1–4 QbD also can be used to enhance the performance of existing products and processes.

There are many variables involved in improving an upstream process. Hulbert, et al. have shown an effective way to prioritize the collection of variables.5 Now, we need a strategy to deal with the variables identified by the prioritization approach proposed by Hubert, et al.

Hubert, et al. point out that there are a lot of data involved in improving upstream productivity. The data can't be collected all at once. The strategy selected must enable us to collect the right data, at the right time, and in the right amount, as discussed by Lonardo, Snee, and Qi.6

Figure 1
Even when good QbD approaches are used, experimentation often is slowed down by poor processes. Personnel, materials, testing, and equipment often are not available when needed, resulting in wasted time and effort. Additional time and effort are wasted when analytical methods are of poor quality or the procedures in the analytical laboratory are inefficient.

Two methods can be used to increase the speed of upstream development, which in turn speeds up the development of process understanding:

  • using Design of Experiments (DOE)-based strategies to design, analyze, and interpret experiments, resulting in getting better information in a timely fashion
  • using Lean principles to streamline the availability of information, materials, equipment, measurements, and personnel for the experimental process, thereby accelerating the flow of the experimental process.

The overall strategy therefore is to speed up the experimental process by adopting strategies that collect the right data when needed and to improve the work processes used for experimentation. The result is that the experimentation is speeded up, in turn speeding up the development of process understanding, which is fundamental to improving upstream productivity (Figure 1). Some critical principles in conducting upstream experimentation are summarized in the sidebar and discussed in the following paragraphs.

lorem ipsum