Owens Corning Science and Technology Center Internship
I developed an unsupervised machine learning method to quantify shingle damage and guide impact-resistant roofing design.
Improving roofing impact resistance requires more useful damage measurements than a pass/fail result alone can provide.
What I did: During my Owens Corning internship, I led a roofing-system investigation and developed an unsupervised machine learning method to quantify shingle damage. I presented the findings to senior technical leaders and documented the work in a peer-reviewed internal technical memo. The method remains proprietary.
Tools and methods: Unsupervised machine learning, quantitative damage assessment, and technical reporting.
