LDRD seminar: Nov. 27
Three Argonne researchers will discuss their Laboratory-Directed Research and Development (LDRD) sponsored work at the LDRD Seminar Series presentation Tuesday, Nov. 27, 2018, at 12:30 p.m. in Building 212, Room A157. All are welcome to attend.
Visit the LDRD website to view upcoming seminars.
“Using Adaptive Optics to Improve Ultracold Atom Transport,” by Assistant Physicist Michael Bishof (PHY)
Laser-trapped, ultracold atoms are used for a myriad of projects spanning basic and applied scientific fields. In the Physics division, we use them to look for physics beyond the standard model, test the fundamental symmetries of nature and understand the properties of unusual nuclei. Often these sensitive experiments require us to transport these atoms over meter-scale distances to isolate them from environmental perturbations. Current techniques to transport atoms over such large distances are cumbersome and inefficient. We have developed a laser-based atom trap using adaptive optics that not only allows us to transport atoms, but also allows us to manipulate the trap in ways that are impossible with traditional optics. Our apparatus will soon be tested using ultracold radium atoms and we expect significant improvements to how efficiently we can transport atoms, resulting in more precise scientific results.
Michael Bishof is an assistant physicist in the Medium Energy Physics group of the Physics division. He works on improving measurements of the 225Ra electric dipole moment and on building a new atom trap trace analysis (ATTA) instrument for Argonne’s TRACER Center. Previously, he was a Director’s Postdoctoral Fellow in the Physics division. Michael earned his Ph.D. from the University of Colorado at Boulder in 2014 under the supervision of Prof. Jun Ye. His work focused on using the precision of atomic clock systems to understand atomic interactions, study many body physics, and create quantum-limited sensors. He also contributed to the most accurate and precise clock ever made.
“Agent-based Modeling for Biological Design and Biosecurity Applications,” by Senior Systems Engineer and Argonne Distinguished Fellow Charles Macal (DIS)
This talk describes an agent-based model that connects simulated gene editing of individual organisms, using techniques such as CRISPR-Cas9 and gene drive, on future generations by modeling population dynamics. We have scoured the scientific literature and connected with leading researchers in the field for the most recent data, which is the basis for a finite state machine description of all possible known outcomes in the gene editing (drive) process. We use multiscale stochastic agent-based simulation to predict the gene editing outcomes at the micro-(individual) level to the macro-(population) level. Although the conceptual modeling framework is applicable to all organisms, this talk focuses on insect disease vector applications. The model could have application in predicting the generational impacts of field experiments in which genetically-modified organisms are released into the environment and mingle with wild populations.
During his more than 30 years at Argonne, Charles (Chick) Macal has conducted and directed multidisciplinary systems projects developing innovative computational simulation models in the areas of national security, energy and infrastructure, the environment, and healthcare, including infectious disease models. His work focuses on modeling large-scale social-technical systems, composed of people and their social environment and their interfaces to the technologies they use.
“Computer Vision and Machine Learning for Low-Cost Wide Area Drone Detection with Distributed Urban Sensor Networks,” by Principal Software Engineer Adam Szymanski (SSS)
This LDRD was focused on the implementation and demonstration of machine learning algorithms for drone detection using a low-cost sensor network based on the Waggle platform (an Argonne technology). Drone or unmanned aerial system (UAS) detection is a large area of concern for National Security organizations. While a number of systems exist for drone detection, their usage is limited by cost, detection area, sensor type and current regulations. This project attempts to find a solution to some of these limitations through the use of existing urban infrastructure and automated machine learning techniques that can be executed on embedded low-cost platforms. In this presentation I will discuss the machine learning and computer vision methods used as well as the challenges encountered during implementation on this platform.
Adam Szymanski is a computer scientist in the Strategic Security Sciences division at Argonne. He works on a variety of modeling and simulation (M&S) projects. He graduated from Carnegie Mellon University with a degree in computer science and robotics and is currently pursuing a masters in analytics at the University of Chicago with an emphasis on advanced computational models including computer vision and machine learning algorithms. Szymanski’s current work involves advanced logistics modeling for the U.S. Transportation Command through the Analysis of Mobility Platforms (AMP) project. He is also the lead principal investigator on a program for the Naval Research Laboratory that focuses on electronic warfare M&S. This project includes both EW system modeling as well as detailed Radio Frequency (RF) propagation modeling in complex environments. Before coming to Argonne, Szymanski spent 10 years at the Naval Research Laboratory working in the EW M&S Branch of the Tactical Electronic Warfare division.