Michael’s focus is on scientific computing and HPC modeling for:
- endpoint computing and edge computing: reduce data at the sensor
- scalable OS-agnostic HPC algorithms
- precision radar and passive indoor radar / optical remote sensing
- broadband wireless systems including: broadcast, public safety, two-way voice/data radio, UAS, satellite
Michael is a core developer for the Meson build system. The Meson build system design philosophy centers around an object-oriented non-Turing complete DSL. Meson is a modern, fast, complete alternative to CMake and other build systems. Meson can be used to bootstrap embedded systems and seamlessly build large software stacks for HPC where only a compiler is available.
Michael code contributions to widely used projects include:
- Python type hinting subsystem
- Meson: core developer focusing on HPC and Fortran applications
- GNU Octave
Michael has published over 110 geoscience and aerospace programs, including 28 packages listed at http://heliopython.org/. Michael’s work blends data science with full-stack hardware development, using data science with radar, GNSS, optical and other remote sensing techniques.
- Research and development in applied scientific computing, particularly for aerospace and geospace remote sensing
- geoscientist with expertise in RF systems
- expertise in remote sensing (multiple patents) using computer vision and machine learning
- solved the century-old problem of auroral discrimination based on physical driving mechanism via machine learning and computer vision techniques
- technology rights and geoscience outreach
- award-winning STEM educator
Growing engineering endeavors
Professor Hirsch uses comprehensive model-based engineering frameworks to turn an idea into a product. Broad engineering expertise combined with years of experience bring quick resolve to vexing system issues for geospace/geoscience research as well as corporate and government entities.
- comprehensive practical and theoretical design experience with RF systems from antenna to CPU and detailed propagation modeling
- remote sensing: radar & radio science–from microwatt to megawatt with EMC mitigation
- remote sensing: optical–terabytes/day reduced 1000x by machine learning/machine vision algorithms using edge computing
- model-based design reduces iterations and speeds path from concept → bench prototype → fielded system
- serves as outside expert for design review, expert witness or pre-money planning and prototyping (whether research grants or startups)
Michael works with public safety and transportation entities on their wireless infrastructure, whether in modeling upgrades, solving coverage problems or helping write RFPs and reviewing responses. As an independent voice of experience, Michael helps police, fire, EMS, transit and other public agencies as they maintain and transition from legacy analog radio and data systems to broadband digital networks, including FirstNet. Michael works with agricultural operations from farmers to companies working with and for agricultural advancement as humankind struggles to feed more with less.
Michael teaches hundreds of Electrical and Computer Engineering undergraduates and graduate students each year. His courses cover Senior Capstone as well as Software Engineering for data pipelines in concurrent cloud and endpoint applications.
Michael’s Mathematics Genealogy Project entry
PhD in Electrical Engineering
Michael developed novel ionospheric remote sensing system techniques via optical tomography, radar, and radio instruments. The kickoff to Michael’s doctoral research was an AFOSR-funded deployment to Greenland on multiscale (temporal and spatial) features in the aurora. The lessons learned from this system built confidence in the field deployment of a multi-site auroral tomography system to the Poker Flat Research Range near Fairbanks, Alaska.
Michael earned the Master of Science in Electrical Engineering degree from Boston University in the Space Physics: Remote Sensing & Instrumentation track.
BSEE (High Honors)
My undergrad ECE accomplishments include: