Speakers BBD 2019

Ann Christine Catline, Purdue University

Ann Christine Catlin is a Senior Research Scientist in Research Computing at Purdue University. Catlin has collaborated with hundreds of faculty, physicians, pharmacists and other researchers in medical and scientific disciplines to build customized data and computing platforms for investigating research questions in civil engineering, electrical engineering, natural disasters, computational chemistry, ecotoxicology, pharmaceutical manufacturing, medical device informatics, adverse drug events, and cancer treatments. Her work has been funded by NSF, NIH, EPA, NIST, FDA, DoD, academic centers and national foundations, with more than 50 publications describing the design and implementation of her research platforms or the outcomes produced through shared use of these platforms. Catlin’s recent work is focused on the development of research platforms that support the full investigative lifecycle and are effective across science domains.
Talk:
Abstract:

Decades of government funding has targeted development of platforms that fully support team-shared research and are effective across science domains. Even so, most existing platforms offer support only for specific research activities, such as file sharing or HPC computing, and platforms that offer more complete support are heavily customized for specific scientific disciplines. As a result, most researchers continue to carry out their investigations in an ad hoc manner. This means their collected data, files, code, analyses and outcomes are fragmented, seriously complicating data preservation, sharing, interoperability, results traceability, publication, reuse and re-interpretation. In this presentation, we introduce the Digital Environment for Enabling Data-driven Science (DEEDS), a platform that provides comprehensive, end-to-end solutions for data, computing and scientific workflows. DEEDS is an extensible web-services platform that supports research projects throughout the investigation lifecycle, offering an interactive, user-friendly, team-shared dashboard where researchers can 1) organize research activities, 2) import, preserve, and manage data, 3) integrate and launch research computing tools, also supporting their execution on HPC resources, 4) connect data to tools, including capture of computational workflows for data provenance and reproducibility of results, 5) interactively explore, compute on, and visualize data and 6) apply access control as needed. The DEEDS platform was developed as a collaborative effort between computer scientists and domain scientists, including civil engineers, who jointly established requirements for functionality and usability. DEEDS datasets and operation will be demonstrated for two SMARTI research projects: Rural bridge health studies based on interactive data tables for collected NBI inspection data, trend analysis tools, Jupyter notebooks, and built-in R analytics and Bridge health monitoring and decision-making based on collected sensor data, machine learning algorithms for damage detection/classification, and outcomes visualization and analyses.

Sattar Dorafshan, Federal Highway Administration

Dr. Sattar Dorafshan is a postdoctoral research associate at Federal Highway Administration’s Turner-Fairbank Highway Research Center. He received his Ph.D. in structural engineering from Utah State University in 2015 and obtained his masters and undergrad degrees in civil engineering in Iran. He is an expert in the application of artificial intelligence for structural condition assessment. He also has expertise in Unmanned Aerial Systems (UAS) assisted inspections of infrastructure, accelerated bridge constructions, earthquake engineering, and numerical methods. At Turner-Fairbank highway research center, Dr. Dorafshan is developing the state of the art deep learning models for autonomous interpretation and fusion of nondestructive evaluation data.

Talk: Federal Highway Administration NDE Program Vision for 21st Century Bridge Data
Authors: Sattar Dorafshan, Ph.D., Postdoctoral Research Associate and Hoda Azari, Ph.D., NDE Program Manager.
Abstract:

The Federal Highway Administration (FHWA) Nondestructive Evaluation (NDE) Laboratory is a world-class facility to conduct state-of-the-art research, development, and implementation of NDE systems and technologies to improve the Nation’s highway infrastructure assets. Since its establishment in 1998, the NDE Laboratory has been maintained as an open resource for the FHWA, State departments of transportation (DOTs), industry, and academia. To address the needs of the 21st century, the NDE program has been moving towards automation to infrastructure construction, inspection, monitoring, and management. In this presentation, a brief overview of the FHWA NDE program is introduced. In addition, the past and present research investigations for using robotic data acquisition and autonomous data interpretation using artificial intelligence are presented. Finally, the existing and ongoing web-based platforms for data storing, sharing, analyzing, and training are discussed.