All,

Argonne’s Physical Sciences and Engineering (PSE) directorate will host a “PSE AI in Science and Engineering” workshop Sept. 5-6, 2019, at the APS Conference Center. The workshop is open to all Argonne staff, postdocs and students.

REGISTER NOW! Attendance is limited to 85 people.

WORKSHOP DETAILS

During the workshop, we will assess the current status of AI methods and approaches in sciences and engineering. Breakout sessions will be used to develop recommendations for short-term goals and suggestions for longer-term strategic goals, focusing on:

1)     Areas in which Argonne can claim world leadership

2)     Identifying bottlenecks and required resources

3)     How to better leverage expertise and resources

4)     How to build or encourage synergy and collaborations

Further discussion will center on goals and strategies in the context of the DOE National Laboratories AI for Science Town Hall Meetings, one of which will be hosted by Argonne July 22-23, 2019.

BREAKOUT SESSIONS

Breakout questions are modeled after the DOE National Laboratories AI for Science Town Hall Meetings.

Application Breakout Questions  

•       In which “must-do” application areas is PSE behind and what is needed to catch up?

•       What are three to four significant opportunities in scientific domain areas in short-term (2-5 years) and longer term (5+ years) timeframes that combine Argonne’s unique expertise and capabilities and that address challenges articulated by the DOE?

•       For each opportunity identified: 

o   What new advances are needed in modeling and simulation, as well as, experimental capabilities to address this challenge? 

o   What are possible new roles for AI (e.g., machine learning, deep learning, statistical approaches, data analytics, automated control and other data-driven approaches) to address this challenge? What developments in AI are needed to realize these roles?

o   What are the opportunities for integrating modeling, simulation and experiments with AI to advance the scientific area? 

o   What are the major sources of data and what scales are required to construct data-driven models with sufficient validity/accuracy? What problems need to be solved to acquire and manage the needed data? 

Cross-Cutting Breakout Questions

•       What are the three to five open questions that need to be addressed in order to maximally contribute to AI impact in the science and engineering domains?

•       For each open question identified: 

o   To what extent is Argonne uniquely positioned to address this challenge?  

o   What capabilities are imagined in the near-term (2-5 years) and the long-term (5+ years) timeframes? 

o   What level of infrastructure and investment is needed to realize the impact? 

o   What is needed from the domain sciences to push this technical area forward?

For any questions, please contact Olle Heinonen (heinonen@anl.gov or 2-4877).