Workshop on Artificial Intelligence
in Natural Hazards Engineering

February 18-19, 2020

Texas Advanced Computing Center
Advanced Computing Building
Austin, TX

The application window has closed for in-person attendance, but registration is now open for attendance via webinar.

The NHERI DesignSafe team in collaboration with the NHERI SimCenter will host a 1.5 day workshop on February 18-19 to bring together expertise in Natural Hazards Engineering and Artificial Intelligence/Machine Learning/Deep Learning (collectively known as AI/ML/DL).

The purpose of this workshop is to build a research agenda determining where AI/ML/DL can and is being productively employed in natural hazards engineering research, what techniques hold key promise, how these techniques may be adapted to the hazards engineering domain moving forward. The workshop will be held at the Texas Advanced Computing Center (TACC), of the University of Texas at Austin.

The workshop will be an interactive event, including anchor talks by researchers across the hazards engineering and AI/ML/DL domains as well as interactive breakout, reporting and discussion sessions. Interested participants should apply by submitting a 2 page NSF formatted Biosketch and a Statement of Interest of no more than 250 words. The Statement of Interest should succinctly describe your research area and interests with respect to the workshop theme. Former experience in applying AI/ML/DL to hazards engineering problems is not required.

Participant support is available for a select number of applicants; please indicate if you are requesting support when applying. Applications are welcomed from all areas of AI/ML/DL, disciplines of natural hazards engineering, and career stages, including faculty, post-doctoral researchers, and late stage PhD students.


Lodging

The workshop will be held in the ACB building on the J.J. Pickle Research Campus (PRC) of The University of Texas at Austin at 10100 Burnet Road, Austin, TX, 78758. A hotel room block is available at The Hyatt House Austin Arboretum, you can book your room using the information below.

Website: https://austinarboretum.house.hyatt.com/en/hotel/home.html?corp_id=G-TAC4
Phone: 1.866.974.9288
Reference Group Code: G-TAC4

AI Bootcamp

There will be an additional AI Bootcamp (including hands-on use of TACC resources) on Monday, February 17th from 1pm to 5pm for those who signed up during the application process.

The topics that will be covered in the training are:

  1. AI/ML/DL and HPC at TACC
  2. Introduction to ML/DL
  3. Introduction to Keras and TensorFlow
  4. Hands-on with ML/DL Examples in Natural Hazards

Agenda

DAY ONE — February 18, 2020

8:00am

Light Breakfast

8:30am

Opening Remarks

9:00am - 9:45am

AI/ML Overview Talks

Machine Learning for Prediction Tasks: Applications in the Natural Hazards Domain
Devika Subramanian, Rice U

Machine Learning for Inferring Scientific Models: Hope or Hype?
Omar Ghattas, UT

9:45am - 10:30am

Inventory Characterization Talks

Machine Learning for Building and Structural Information Modeling
Stella Yu, UC Berkeley

Aerial imagery analysis with minimal labeling
Nebojsa Jojic, Microsoft

10:30am - 10:45am

Break

10:45am - 11:45am

Breakout 1
  – Breakout Slides (Powerpoint)
  – Breakout Slides (PDF)

11:45am - 12:30pm

Reporting 1

12:30pm - 1:30pm

Lunch

1:30pm - 2:15pm

Surrogate Modeling for Hazard/Response/Risk Assessment Talks

Surrogate Modeling Applications for Storm-surge Predictions
Alex Taflanidis, Notre Dame

Big data and machine learning for improved modeling in Natural Hazard Engineering
Youssef Hashash, UIUC

2:15pm - 3:15pm

Breakout 2

3:15pm - 4:00pm

Reporting 2

4:00pm - 5:00pm

General Discussion + "Pass the Cards" Activity

5:00pm - 7:00pm

Reception at TACC

DAY TWO — February 19, 2020

8:00am

Light Breakfast

8:30am - 8:45am

Recap

8:45am - 9:30am

Damage Assessment and Classification Talks

Machine Learning to Enable Visual Data Analytics
Shirley Dyke, Purdue

Application of Machine Learning to Post-Earthquake Damage Assessment of Building Structures
Henry Burton, UCLA

9:30am - 9:45am

Break

9:45am - 10:45am

Breakout 3

10:45am - 11:30am

Reporting 3

11:30am - 12:30pm

Final "Activity" + Wrap Up Discussion

12:30pm

Lunch