Inventors at Georgia Tech have developed an ultra-fast on-demand data-driven prediction of polymer properties. This technology implements machine learning algorithms and requires the polymers to be fingerprinted numerically so that the fingerprint components correlate to the property to be predicted. Using our capability, polymer properties may be predicted instantaneously prior to actual synthesis and testing. The prediction models may be used in an initial line of screening polymers, thus saving significant and unnecessary production and manufacturing cost and time. A prototype is currently available where a user may predict the property of a polymer- www.polymergenome.org
- First of its kind – no effective and efficient alternative approaches to predict polymer properties currently exist
- User friendly – implementable online and easily accessed by users
- Cost and time savings – prevents unnecessary manufacturing
- Screening polymers
Polymers are materials made of long chains of molecules, and have unique properties depending on the type of molecules being bonded. Polymers have a broad scope of applications, from water bottles to clothing. A need for companies manufacturing polymers is the ability to predict polymer properties; at this point in time, there is no effective solution to do so.