Skip to main content

DuPont

Submitter

Bernardo Tiburcio, Global Digital Innovation Leader

Company

DuPont

Industry

Specialty Chemicals


Business Challenge

DuPont has a focus on driving competitive advantage and eliminating complexities across supply chain, manufacturing and operations functions. Our focus is on predictability, reliability and agility across our global supply chains. As part of this challenge, we were tasked with leveraging our data in production processes to ensure consistently high-quality end products that differentiate us from competition. Specific focuses have been on improving customer demand forecasting, improving yield and releasing manufacturing capacity at sites for sold out products.

Analytics Solution

To address this and other challenges, we built our “Spark Digital” Center of Excellence to deliver advanced analytics and other emerging technology solutions to DuPont businesses and functions. The team has created eight digital products to support our business, many including advanced analytics and machine learning. Two of Spark’s digital products – Spark Smart Demand Forecasting and Spark Intelligent Process Optimization – target supply chain and operations effectiveness. Our Spark Digital Smart Demand Forecasting (SDF) solution improves networking capital and streamlines supply chains with more accurate demand predictions through machine learning models. SDF analyzes and finds correlations in data such as orders on-hand, historical sales, economic and market indicators — then creates much more accurate predictions of customer demand. Spark Intelligent Process Optimization focuses on increasing yield and improving quality with AI-based machine learning models that transform high volumes of data into predictive or prescriptive recommendations.

Impact

SDF benefits include higher revenue through a more accurate understanding of the market and increased production accuracy, reducing non-productive inventory generation and lowering the carrying cost of inventory. Our current models have increased forecast accuracy by about 14 points (~30% improvement in accuracy), resulting in multi-million dollars in benefits. Intelligent Process Optimization benefits include increased throughput, reduced raw material costs and reduced waste handling, shipping and disposal costs. Initial implementations have delivered double-digit millions of dollars in capacity release.