(59a) Rapid Development of New Product Formulations through the Efficient Utilization of Company Data

The continual development of new materials or products is the lifeblood of most chemical companies. Companies are constantly trying to: (i) develop products with novel properties for use in new applications, (ii) improve the quality and cost of existing products, (iii) reformulate products using new materials, and (iv) consolidate product formulations using fewer starting materials. However, their general success in these endeavors has been poor. In a recent Accenture benchmarking report (2015) entitled “Why Low-Risk Innovation Is Costly” (http://www.accenture.com/us-en/Pages/insight-low-risk-innovation-costly.aspx), executives indicated that indeed innovation is extremely important. 93% of executives surveyed said their companies depend on innovation for long-term success,but only 18% believed that their company derived a competitive advantage from its innovation strategy, and most of the executives were less satisfied with their company’s current innovation performance (2015) than they were in 2009(previous benchmarking report).

Product development at most big companies is characterized by:

• A lot of trial and error and intuition-based experimentation
• Some use of DOE’s – but usually only in a very local manner – no global overview
• No repository for the accumulated product development knowledge of the company
  • No easily accessible database of all past development data
  • No good database on raw material properties
  • No repository for past modeling efforts
  • No good way of sharing information among developers
    • Intuitive experience of individual researchers is often only way they carry knowledge forward.
    • Little of that information is effectively transferred to other developers.
  • No effective ways of rapidly finding formulations involving new raw materials
  • Processing conditions are often not included in development studies, but left to the scale-up group

There are basically 3 general degrees of freedom in developing, improving, or reformulating products: (i) the choice of raw materials or ingredients, (ii) the ratios in which to combine or react them (formulation or recipe) and (iii) the process conditions under which to manufacture them. These three degrees of freedom are highly interactive, making it important to consider them simultaneously.

In this presentation we discuss powerful multivariate latent variable data analysis methods for extracting information from diverse development databases (raw materials, processes, formulations and quality), for optimally augmenting the databases with multivariate DOE’s, and for using that information via optimization to rapidly formulate products that meet all the desired quality specifications at minimum cost.

A number of past development projects will be used to illustrate the approach. These include the development of novel functional polymers for use in medical devices and for the cores of the very popular Srixon golf balls (used by a large number of the touring pros), the development of novel high performance polymeric coatings, and the development of new food formulations.

Although this presentation will refer to advanced multivariate data analysis, design of experiments, and optimization, it will not be essential for the attendees to have a detailed knowledge of these in order to understand the essential ideas and the potential of the methods. The emphasis will be on the basic problem of rapid product development through the efficient quantitative utilization of company data, and on illustrating this approach with successful industrial examples.