Towards an Extended Brain for Accelerating Research Discovery: A Proof-of-Concept for Next-Generation Adsorbent Discovery

Chengxiang Zhai (Computer Science)
Xiao Su (Chemical & Biomolecular Engineering)

Research Problem

The scientific knowledge in most engineering and science domains is complex and often buried and scattered in massive amounts of literature articles from multiple subject areas, creating significant challenges for researchers to get access to all the scattered relevant knowledge promptly and make effective use of the knowledge to reason about hypotheses for new discoveries. The recent rapid growth of scientific literature data exacerbates the situation. There is a critical need for the development of AI-driven approaches to assist experimental researchers in accelerating materials discovery to achieve adsorbents with higher separation factors which can lead to a lower chemical footprint, reduced energy consumption, and decarbonization of the processes in general.

Center on Accelerated Research Discovery (CARD) Vision

To accelerate research discovery in a general way, we envision the possibility of leveraging the recent progress of research in AI to develop a novel general intelligent discovery workbench system, Extended Brain (EB), to optimally assist human researchers in retrieving and digesting relevant knowledge, integrating all the relevant knowledge to a research problem, and exploring and reasoning about new hypotheses.

Larger Impact

By hosting EB as a general research-supporting infrastructure to serve many research units on our campus, CARD will enable all of them to increase research productivity and accelerate discovery. As EB is a virtual infrastructure on the cloud, once EB is tested internally at UIUC, it can be easily open to serve external researchers and generating broader impacts.