Experiment Design Strategies

In the dynamic world of chemical manufacturing, where competition is fierce, product innovation is in high demand, and research and development costs are sky-rocketing, a systematic approach to understanding and manipulating factors affecting a process has become crucial. This approach, known as Design of Experiments (DOE), is a method that determines the relationship between factors and the output of a process.

Julia O'Neill, a seasoned professional in this field, has spent over three decades bridging the gap between statistics and chemical engineering. Holding an MS in statistics from the University of Wisconsin-Madison and a BS in chemical engineering from the University of Maine, O'Neill has made significant strides in the application of DOE in the industry.

The traditional one-factor-at-a-time approach to experimentation in product development has been partly responsible for inefficient product and process development. However, DOE offers a solution by allowing for the manipulation of factors together, rather than one at a time.

One of the key sectors benefiting from DOE is drug development. In this field, dimension reduction, a statistical technique, can streamline the analysis process when dealing with a large set of measurements on a small number of subjects. When combined with DOE, this technique can lead to more efficient and effective drug development.

A real-life example illustrates the power of DOE. In a situation where unanticipated results in manufacturing threatened to disrupt the vaccine supply, a team was able to identify the cause and recommend changes, a process that once took several weeks and great cost, can now be solved by a single researcher in 30 minutes using JMP, a software tool that automates the computational work in DOE.

Deploying a methodical approach to experimentation throughout the development phase allows for designing quality into the process at the outset. However, appropriate training and know-how remain a challenge for the industry. Many chemists may not be equipped to work with statistics or DOE, and may not have been exposed to software like JMP.

The wide adoption of DOE in R&D by the chemical-based industries is necessary and strategic. Companies should invest in developing their DOE capability and comprehensive data collection schemes. Julia O'Neill, previously the director of engineering at Merck and Co., played a significant role in integrating continued process verification for all vaccines and biologics.

Modern approaches in DOE allow for designing experiments that fit various situations. DOE was first introduced by UK statistician Ronald Fisher in 1935 and has evolved since. The evolution of DOE has led to a series of design families adapted to specific situations, each offering a unique solution to the complexities of product development and manufacturing.

In conclusion, statistical analysis, particularly DOE, can be a vital tool for accelerating the discovery and creation of viable new products and engineering the processes through which they can be delivered at scale. As the chemical manufacturing industry continues to evolve, the strategic adoption of DOE will undoubtedly play a crucial role in ensuring efficiency, quality, and innovation.