I am a software engineer working as the IT manager for the Institute for Astronomy at University of Hawaii, Manoa. I manage a team of 6 IT specialists who support faculty, grad students and staff at 3 island campuses. I actively contribute my software engineering expertise in advanced computing technologies to research projects that apply AI and machine learning to astronomy, astrophysics and information security. I collaborate with researchers across multiple academic institutions and industry on using high performance computing technology for astronomy and astrophysics. I have over 30 years of extensive software engineering and over 20 years of IT management experience.
AI and Machine Learning
Solar Stokes Inversion
In 2019 I used Hinode SOP SP data to develop a neural network for fast Stokes inversion of solar spectropolarimetry which was accepted for publication as a poster at the 33rd Annual Conference on Neural Information Processing Systems in 2019. I continue to work on improving neural network Stokes inversion using Hinode and eventually SDO and DKIST DL-NIRSP data.
Network Intrusion Detection
I am working on applications of AI and machine learning to the problem of network intrusion detection using auto encoders for anomaly detection.
High Performance Computing (HPC)
I lead the “AstroFlows” group within the UH IRNC PIREN project. We work on performance tuning of data transfers of big data and real time data in astronomy over long distance, high latency network paths, e.g. Hawaii to Europe, Asia and the Middle East.
Working with the UH Applied Research Laboratory, I developed a method of efficient and complete information security assessment using the Risk Management Framework and the NIST 800-53 standard.
Semiconductor Test and Evaluation
While at Sof-Tek Integrators, Inc. I helped develop 3 patents related to parametric testing of semiconductor light emitting diodes and lasers.
- Dodds, S., Cunnyngham, I., et al Inverting Solar Spectropolarimetric Observations with Deep Learning, 33rd Annual Conference on Neural Information Processing Systems (2019)
- U.S. Patent 9,551,669 Morrow D., J. Dummer, and S. Dodds. Method and system for characterizing light emitting devices. Jan. 2017
- U.S. Patent 9,128,144 Morrow D., et al. System and method of quantifying color and intensity of light sources. Sep. 2015
- U.S. Patent 8,593,148 Morrow D., J. Dummer. System and method of testing high brightness LED (HBLED). Nov. 2013
- AI/ML neural network applications to astronomy and astrophysics
- AI/ML neural network applications to network and information security
- FISMA compliance, RMF, NIST 800-53/171
- High performance, velocity, volume and latency data transfer
- Network and database administration for Pan-STARRS Published Science Product PB scale distributed database
- Association for Computing Machinery
- Astronomical Society of the Pacific
- Co-I - NSF 2008344 Critical Early DKIST Science: Spectropolarimetric Inversion in Four Dimensions with Deep Learning, 07/01/2020-06/30/2023, $668,938 (PI Dr. Xudong Sun)
- Co-I - NASA 80NSSC19K0413 Cometary Observation Metadata Archive (COMA), an Interactive Science Portal for Primitive Solar System Objects, 07/01/19-6/30/22, $1,069,228 (PI Dr. Karen Meech)
- Collaborator - NSF OAC 2029312 IRNC Core Improvement: SC-Transport: Pacfic Islands Research and Education Network, 08/15/2020 - 07/31/2025 ,$4,494,140 (PI Dr. David Lassner)
- PI - USNO Task 19F8810 Information Management and Security Upgrades for United Kingdom Infrared Telescope (UKIRT), 05/13/2019-05/12/2020, $777,710
- PI - NSF OAC 1541471 CC*DNI INSTRUMENT: High Performance Reliable Network Access to MLO, 09/01/15-08/31/18, $398,207
M.S. 2010, California State University, Fullerton, Software Enginering
- SPIN - Convolutional Neural Network for Real Time Stokes Inversion
- AstroFlows - High Throughput Long Distance Data Transfers
- Pan-STARRS Published Science Products Database
- RMF/NIST 800-53/171 Compliance Assessments (UKIRT, IRTF, ATLAS, Pan-STARRS)
- MKOCN/MLONET/HONET - Remote Site High Performance Networking