PREP: Advancing Research and Education in AI/ML for Science (AREAS)

University Of Illinois At Chicago, Chicago IL

Investigators

Abstract

Led by the University of Illinois Chicago (UIC), this project aims to Accelerate Research and Education in AI/ML for Science (AREAS) by recruiting, retaining, and seeding the careers of undergraduate physics majors. This project partners with the National Science Foundation funded Accelerated AI Algorithms for Data-Driven Discovery (A3D3) Institute. Students will work with A3D3 scientists/engineers on science driver projects associated with the A3D3 Institute. The project will select promising UIC physics majors who have completed the sophomore physics sequence to conduct research under the guidance of an A3D3 faculty supervisor and graduate student mentor. A UIC faculty mentor will complete the support team that will monitor student performance and progression towards degree. This project also funds curricular development of advanced undergraduate laboratory courses to incorporate FPGA technology, providing easy access to state-of-the-art infrastructure for data science to all UIC undergraduate students. A post-baccalaureate program will specifically support undergraduate students to work at an A3D3 university for one year following graduation, as a bridge to graduate school or a job in industry. The project will monitor the progress and track the success of students participating in this partnership, as well as enlist an external evaluator to measure the effectiveness of the program. The project is expected to increase graduation rates as well as improve performance in the core physics courses for physics majors at UIC. The program is also expected to increase the rate at which UIC physics majors are successfully admitted to graduate school, thereby bridging their pathway to a career in physics. In partnership with the National Science Foundation funded A3D3 Institute, this project brings AI/ML solutions to the enormous technical challenges that particle physics detectors will face at future high energy colliders. Disentangling the unprecedented numbers of particles expected in each collision requires detectors that have billions of readout channels and generate petabytes of data per second, which will be impossible to fully save for analysis. The project supports research that enables future scientific discovery via efficient data-reduction at the source, within strict bandwidth and latency constraints, and beyond. It is broadly categorized into three areas: tracking detectors for particle physics; level-1 triggering and data acquisition in particle physics; and event reconstruction and data analysis in particle physics. This research has the potential to enable the exploration of data that would normally not be saved for analysis and could lead to surprising future discoveries that transform our fundamental understanding of the universe. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.