Soft AE - Exploring Research and Educational Intersections of

SOFT MATERIALS, AUTONOMOUS
EXPERIMENTATION & SCIENCE POLICY

Soft AE Kickoff group photo by Lamont Abrams

Thursday, May 4, 2023
10:00 AM - 3:15 PM

Glandt Forum
Singh Center for Nanotechnology
3205 Walnut Street, Philadelphia PA

 

Opening Remarks [recording]

Beth A. Winkelstein, PhD
Deputy Provost, University of Pennsylvania

Chinedum Osuji, PhD
NRT Soft AE Director and PI

10:00 am

Opening Remarks
Beth A. Winkelstein, PhD
Deputy Provost, University of Pennsylvania

Chinedum Osuji, PhD
NRT Soft AE Director and PI

10:15 am

Keith A. Brown portrait

Unraveling Hierarchical Soft Matter using Self-Driving Labs [recording]
Keith Brown, PhD, Boston University

The development of advanced materials is slowed by the vast space of possible formulations and processing conditions that must be experimentally navigated. Our work focuses on the exploration of polymers using self-driving labs, or the combination of automation to perform experiments without human intervention and machine learning to select experiments to progress toward a user-defined goal. In this talk, we overview our progress in the areas of discovering architected materials for mechanical performance and identifying functional polymers using ultra-miniaturized formats.

11:00 am

Shorts: Talks from Trainees
Felipe Rodrigues Martins [recording]
Aria Zhang
[recording]
Kaitlin Wang
[recording]

11:12 am

Minki Lee portrait

Accelerating Electrospray Deposition Film Discovery through Bayesian Optimization [recording]
Minki Lee, PhD, University of Pennsylvania

Experiments are often tedious, time-consuming, and resource demanding. To address this problem, humans and robots have begun to collaborate to advance science and technology at a much faster pace than what humans can achieve today. This is done through the development and utilization of closed-loop, autonomous experimentation (AE) systems. AE systems comprise an iterative research loop that is carried out autonomously to achieve a research objective without human intervention. A key component of the AE process is the efficient interplay between decision-making and automated platforms in a closed-loop process. Bayesian optimization (BOA) is used as a decision-making method to direct the search of a global optimization problem by selecting the next possible point in the parameter space. The basic concepts of AE and its application in scientific areas will be covered.

11:30 am

Group Photo and Lunch

Due to unexpected circumstances Dr. Talati's talk will need to be rescheduled for later in 2023. Afternoon talks will start with Dr. Liu's 1:30 talk.

12:45 pm

Shuchi Talati portrait

Intersection of Technology and Policy for Addressing Societal Grand Challenges
Shuchi Talati, PhD, American University

The challenges we face today are deeply transdisciplinary. We must consider history, politics, policy, and science as we attempt to think about the best ways to approach societal decisions. This talk aims to look at the vital nexus of policy and technology to think about how to build careers in this space and what resulting work can look like.

1:30 pm

Andrea Liu portrait

Autonomous Experimentation for Better Autonomous Learning? [recording]
Andrea Liu, PhD, University of Pennsylvania

We have developed electrical networks of variable nonlinear resistors that can adjust themselves so that the network as a whole can learn to perform desired tasks that can be framed as delivering desired voltages at desired output nodes of the network in response to source voltages applied at input nodes, such as classification and regression. This raises the question: How does the ability to perform the task emerge as a collective phenomenon in this system as it learns? We have found that persistent homology provides a simple way of understanding the emergence of learning. The next question, which remains open, is whether we can harness persistent homology to obtain better learned solutions by the circuits, using autonomous experimentation.

2:15 pm

Heather Eckenrode-Stiffler portrait

Advancing Digitalization and Automation in the Coatings Industry [not recorded]
Heather Eckenrode-Stiffler, PhD, The Dow Chemical Company

Assessing coating performance is a very visual process. The Dow Coating Materials team continues to advance ways to analyze and predict film performance. Extensive work is being done to develop both digital and automated tools.

3:00 pm

Closing Remarks [recording]

Boon Thau Loo, PhD
Associate Dean, Graduate Programs, Penn Engineering

Beth S. Wenger, PhD
Associate Dean, Graduate Studies, Penn Arts & Sciences

Russell Composto, PhD
NRT Soft AE Associate Director and co-PI