Covestro – Meng Yuan – 2025

Summary

Yuan Meng achieved energy consumption attribution and energy-saving optimization through data-driven modeling at Covestro.

Goals

Covestro, as one of the world’s leading manufacturers of high-performance materials, is also a pioneer in low-carbon green manufacturing. To identify potential energy-saving opportunities and reduce energy consumption, Covestro engaged Yuan Meng to conduct energy consumption modeling and production optimization using data-driven approaches.

Solutions

Yuan Meng conducted production energy consumption attribution analysis and optimization through data analysis:
1. Established a standardized analysis process based on a Ridge Regression model. This quantified the contribution of different factors to energy consumption fluctuations and was applied across three factories, demonstrating its potential for repeatable deployment.
2. Developed a steam production prediction model utilizing AI algorithms (such as Random Forest and Neural Networks). Building upon this predictive model, mathematical optimization of load scheduling was performed to maximize steam output. This approach was implemented using one factory as a case study.

Potential Impact

Through Yuan Meng’s work, Covestro achieved multi-factor real-time visualization of energy consumption fluctuations across three factories, identifying an annual energy-saving potential of $190,000 USD. Furthermore, the load optimization implemented at one factory demonstrated an additional annual energy-saving potential of $112,000 USD, while ensuring stable output of the primary products. These contributions have helped Covestro further enhance its green and low-carbon credentials.