549362 Accelerating Product Innovation at Dow through Multivariate Modeling

Wednesday, April 3, 2019: 1:30 PM
Marlborough A (Hilton New Orleans Riverside)
Brandon Corbett1, Marlene Cardin1, Kristin Wallace1, Alix Schmidt2, Haseeb Moten3 and Rebecca Beeson3, (1)ProSensus, Burlington, ON, Canada, (2)Continuous Improvement Center of Excellence, The Dow Chemical Company, Midland, MI, (3)Dow Consumer Solutions, Dow Chemical Company, Midland, MI

Accelerating Product Innovation at Dow through Multivariate Modeling

In the era of Industry 4.0, manufacturers such as Dow are focused on gaining more value from their historical data than ever before. ProSensus has distinguished itself as a trusted, global leader in Big Data analytics by helping many Fortune 500 companies across numerous industries use their data to innovate for the future.

ProSensus has recently worked with Dow to accelerate innovation on one of their key product lines – silicone antifoams. For product development applications such as this, ProSensus combines powerful multi-block latent variable modeling with constrained optimization. This approach allows clients to simultaneously optimize the selection of raw materials, recipe formulations, manufacturing conditions and costs to reach targeted product performance properties while adhering to custom constraints. A model-based approach allows Dow to rapidly develop and scale up custom silicone antifoam formulations for the dynamic technical and regulatory needs of customers in the pulp and paper, food and beverage, wastewater treatment, metal working, and other industries.

This presentation will examine the approach taken (data assembly, multivariate modeling, model validation, and optimization), challenges encountered, and results obtained. Perspectives from both ProSensus and Dow will be included.


Extended Abstract: File Not Uploaded
See more of this Session: Big Data Analytics - Industry Perspective II
See more of this Group/Topical: Industry 4.0 Topical Conference