118: A Program That Predicts the Properties of New Polymers (ft. Dr. Rishi Gurnani)

Bakelite was discovered in 1907. Nylon was discovered in 1935, polyethylene in 1936, Kevlar in 1966. All of these discoveries were revolutionary and had years of work put into their discovery. Each major discovery is years apart. Maybe scientists took a trial-and-error approach, trying new ideas until they worked. Maybe some materials were discovered by accident. What if discovering new polymers was possible in just a few minutes? 

 

In today’s episode, we discuss AI-assisted polymer discovery with Dr. Rishi Gurnani, Technology Development Lead at Matmerize. Check out the links below to see how to use Matmerize’s AI tool, PolymRize. Specifically, we discuss: 

 

🔹 Approaches for how MSE’s can learn coding. 

🔹 Case studies of how AI can assist in polymer discovery. 

🔹 The difficulty of predicting the properties of polymers, given the interdependence of many material properties. 

🔹 Challenges in AI, like getting high-quality datasets and overcoming the AI learning curve. 

🔹 And much more! 

 

We hope you enjoy the episode! And as always, let us know what topics you’d like us to cover next! 

 

#PolymerDiscovery #ArtificialIntelligence #MaterialsScience

 

Learn more about our MSE Career Development Online Course, which includes more industry-specific information and advice.

https://www.mseacademy.com/courses/students-tier-2

 

Join our Discord community! You can meet other passionate materials scientists and engineers from around the world, discuss the latest breakthroughs in MSE, share materials-related memes, and get career advice from experts in the field. 

https://discord.gg/49dCV3W4eN

 

Check out our MSE-themed merchandise if you want to support us or simply show off your love of materials science!

https://www.itsamaterialworldpodcast.com/shop

 

For shorter segments and full video podcasts, subscribe to our channel on YouTube. For bloopers, audiograms, and interesting materials science articles, follow us on Instagram, LinkedIn, and Twitter. Feel free to message us on our social media platforms if you have any feedback or recommendations for future episodes, or email us directly at itsamaterialworldpodcast@gmail.com. Finally, reach out to David Yeh and Punith Upadhya on LinkedIn if you'd like to chat about the latest breakthroughs in MSE!

https://www.youtube.com/channel/UCpUXt6ui8iC5npSizlfIJ2A 

https://www.instagram.com/itsamaterialworldpodcast/ 

http://linkedin.com/company/it-s-a-material-world-podcast/ 

https://twitter.com/ItsAMatWorldPod 

 

Disclaimer: Any opinions expressed by either guests or hosts in this show are their own, and do not represent the opinions of the companies or organizations for which they are affiliated.

Watch the full episode here:
https://youtu.be/srLJlK67tXc

#

 

 ~~~~~~~~~~~~~~

Timestamps:

-       Teaser (0:00 - 0:47)

-       Intro, travel, and episode highlights (0:47 – 5:24)

-       Introducing Dr. Rishi Gurnani (5:24 – 6:11)

-       AI in polymer discovery (6:11 – 8:17)

-       Matmerize (8:17 – 10:22)

-       Functional use of AI product (10:22 – 12:52)

-       Work environments of industry vs. academia (12:52 – 15:07)

-       Rishi’s work in Matmerize (15:07 – 18:42)

-       Communication & complexity in AI (18:42 – 21:01)

-       Learning coding in MSE (21:01 – 23:58)

-       Key learnings (23:58 – 26:09)

-       Case studies (26:09 – 28:20)

-       AI with inter-dependent material properties (28:20 – 29:51)

-       Gauging confidence in AI models (29:51 – 32:23)

-       Intimidation and learning curve of AI (32:23 – 35:35)

-       Technical communication with clients (35:35 – 37:14)

-       Getting good-quality datasets for AI (37:14 – 39:04)

-       Challenges in AI (39:04 – 42:34)

-       Advice (42:34 – 44:12)

-       Conclusion (44:12 – 44:37)

-       MSE Academy (44:37 – 45:42)