Steven Myint

Robotics Software Engineer, NASA Jet Propulsion Laboratory

Perseverance and Ingenuity in Jezero Crater, a few meters apart, with wheel tracks between them.
Perseverance and Ingenuity, sol 46 (April 6, 2021). Taken by the WATSON camera on the end of the rover's arm. Credit: NASA/JPL-Caltech/MSSS.

About

I'm a robotics software engineer at JPL and group lead of Robotic Operations and Tools (358T). I've been at the lab since 2007, when I started as a student intern.

On Mars 2020 I led development of SSim (Surface Simulation), which runs the rover's flight software against every command sequence before we uplink it to Perseverance, and I used it to fuzz the flight software. I also wrote parts of the rover's flight software, including the module that commands the Ingenuity helicopter. In operations I've been a Helicopter Integration Engineer and, since 2022, a rover driver.

Since 2023 I've worked on the rover flight software and simulation for Perseverance's onboard global localization. It runs on a commercial Snapdragon processor in the rover's helicopter base station, so we also had to protect it from radiation in software. I'm now PI of a JPL task to make surface operations cheaper: SpaceWasm, portable global localization, learned navigation, and automated downlink and sequencing tools.

Before that I led robotic simulation for InSight, and later was simulation lead for Mars Sample Return and ground software lead for CADRE. Outside JPL I used to develop the Python tools pyflakes and autoflake.

contact [at] stevenmyint [dot] com · GitHub · JPL Robotics profile

Steven Myint under the F-1 engines of a Saturn V first stage.
Under the F-1 engines of a Saturn V.

Work

2026–present

PI, Low-Cost High-Performance Surface Operations for Mars, Moon, and Beyond

A JPL strategic R&TD task to make surface operations cheaper: more autonomy on the vehicle, more automation on the ground, for any mission and not just Mars 2020. Each piece is tested against Mars 2020 data on flight or flight-like hardware.

  • SpaceWasm: a flight-compliant WebAssembly interpreter, so one autonomy implementation can run sandboxed on different vehicles and flight software. It's integrated into F Prime. Firefly is integrating it, and Moonfall and Skyfall are evaluating it. GitHub featured it in Octoverse 2026.
  • Global localization without GPS, extended beyond Perseverance to the Moon and to helicopters. All of the lunar terrain vehicle vendors have said they intend to use it.
  • A traversability model pretrained on about 10,000 Perseverance image/waypoint pairs. Its maps feed the ENav planner, and it runs on the Scarecrow test rover.
  • DownlinkDigest, which writes telemetry summaries from downlink. It could save about two hours per planning cycle.
  • SequenceAgent, which generates, simulates, and validates command sequences. We demonstrated it on Perseverance abrasion planning.
Scarecrow test rover on a rocky slope in the JPL Mars Yard. Inset: NavCam image, frozen DINOv2 encoder, transformer, color-coded traversability map.
The traversability model on Scarecrow in the JPL Mars Yard: NavCam image in, traversability map out. From the task summary. NASA/JPL-Caltech.

2020–present

Group Lead, Robotic Operations and Tools (358T)

Robotics section, JPL. The group writes the flight and ground software used to operate robots: flight-software-in-the-loop simulation, sequence validation, 3D visualization, and the RSVP tool suite. Our open-source work includes Hermes, Honeycomb, and SpaceWasm.

2023–present

Perseverance Onboard Global Localization

Perseverance's position uncertainty grows as it drives. When it gets too large the rover stops and waits for the ground to relocalize it. Onboard global localization matches NavCam panoramas against a HiRISE orbital map on board instead. Over 264 panoramas the error averaged 0.36 m. It runs on the Snapdragon 801 in the Helicopter Base Station, the commercial processor that relayed Ingenuity's data, as a co-processor to the rover's RAD750.

  • Worked on the rover flight software: commands to send arbitrarily large files to the base station, wait on it, and bring results back without making a data product, and to apply the localization result to the rover's pose.
  • Extended SSim with part of the base station's F Prime flight software and a simulated serial link to the rover. It ran the whole rover–base station exchange end to end in under a minute, and was the main testbed for the flight software updates.
  • The Snapdragon isn't radiation hardened. On Mars, weak DRAM bits made redundant runs disagree. We fixed it in software: a memory test, a program that keeps bad pages out of use, CPU throttling, and a rover flight software hot patch that accepts two runs that agree within 10 cm.
  • With Columbia University, worked on software radiation protection for commodity processors in space (HotNets '23, ASPLOS '26). Radshield's version of the localization ran in 26% of the time of the serial, radiation-hardened approach.
  • First closed-loop use was on sol 1762 (February 2, 2026), after a 159 m autonomous drive. Position uncertainty went from 8.4 m to 1.5 m.
Perseverance on Mars and a lab photo of the open rover body. Blue circle: Helicopter Base Station.
The Helicopter Base Station (circled). Its Snapdragon 801 talks to the rover's RAD750 over a serial link capped at 10 KB/s. From Verma et al., IEEE Aerospace 2024. NASA/JPL-Caltech/MSSS.

2015–present

Mars 2020 (Perseverance)

Simulation Lead (2015–) · Flight Software Developer (2016–) · Helicopter Integration Engineer · Rover Planner (2022–) · Mobility Downlink Engineer

  • Led a team of ten building SSim. SSim takes the command sequences, the onboard state from the latest downlink, terrain maps, and onboard files, and runs the rover's flight software on them. It outputs the event records and the rover's kinematic state history. Every plan is simulated in SSim before uplink. Robotic operations runs about 60 simulations per planning day, about 20 seconds each.
  • Led development of RSVP for Mars 2020. MobSketch shows SSim results next to the planned drive, so rover planners can change drives until uplink lockdown instead of freezing them early as on MER and MSL.
  • Fuzzed the flight software through SSim with AFL and with a Python fuzzer against SSim's web service. Both generate command sequences from the flight command dictionary. They found tens of fatal flight software bugs, which were fixed before we hit them on Mars. Static analysis with commercial, open-source, and in-house tools found hundreds more before integrated testing.
  • Wrote the rover flight software module that commands Ingenuity. In operations I was a Helicopter Integration Engineer, the rover-side role that plans each sol's helicopter commanding and the rover's relay of its data.
  • Wrote a flight software module for rover navigation, and led two other developers writing the PIXL instrument manager.
  • Worked around a hardware camera problem in software. Without the workaround the rover assembly schedule would have slipped about a month.
  • Set up continuous integration for flight software in place of the weekly build. Wrote a build system that built the flight software in minutes instead of an hour.
A directed drive I planned through sand ripples on sol 1849 (May 3, 2026). My own 3D reconstruction, from the Navcam stereo pairs taken along the drive, using only public data. Navcam images NASA/JPL-Caltech.
MobSketch, drive over orbital imagery of ripples. Planned path (magenta) and SSim path (blue) nearly overlap between waypoints. Green keep-in zone, red keep-out zone.
The sol 555 drive in MobSketch. Planned path in magenta, SSim's simulated path in blue, keep-in zones green, keep-out zones red. From Verma et al., IEEE Aerospace 2023. NASA/JPL-Caltech.

2023–2025

Mars Sample Return: Simulation Lead and Flight Software Developer

SSim was the main simulation tool for Mars Sample Return operations.

2024–2025

CADRE: Ground Software Lead

Ground software for Cooperative Autonomous Distributed Robotic Exploration, a team of small lunar rovers.

2014–2019

InSight: Robotic Simulation Lead

  • Led development of RSVP for InSight's arm, used to plan and check the arm motions that placed the seismometer, heat probe, and wind and thermal shield. Connected InSight flight software to RSVP.
  • Wrote algorithms that scored candidate instrument placement sites by terrain roughness and arm workspace.
RSVP HyperDrive, top-down: InSight lander and solar arrays, arm over a terrain mesh, seismometer on the ground with its tether to the deck.
Sequencing the seismometer placement in RSVP HyperDrive. NASA/JPL-Caltech.

2007–2014

DARTS Lab and Mars Technology

Student intern in 2007, full time from 2009.

  • 3D visualization for the DARTS simulation tools (Dspace, DspaceOgre), including GPU terrain rendering of planet-scale datasets.
  • Real-time 3D visualization of Curiosity during entry, descent, and landing on landing night.
  • Interfaced MSL flight software with FPGA stereo vision for Mars Rover Fast Traverse and tested it on the Vehicle System Test Bed.
  • Simulation for lunar surface operations, ground vehicles, and the Asteroid Redirect Mission capture testbed.
Simulated rover leaving wheel tracks on rendered Martian terrain.
Wheel tracks drawn on clipmapped terrain as a simulated rover drives. From Myint et al., SMC-IT 2011. NASA/JPL-Caltech.
Lunar Electric Rover, a pressurized cabin on wheels, on cratered lunar terrain.
The Lunar Electric Rover, which I modeled and rendered, in the Lunar Surface Operations Simulator. From Howard et al., IEEE Aerospace 2011. NASA/JPL-Caltech.

2008

B.S. Computer Science, minor in Mathematics

California State Polytechnic University, Pomona.

Tools

Languages and Platforms

C++, Python, C, Bash. VxWorks on RAD750, RTEMS on LEON, Linux, macOS. AFL, libFuzzer, AddressSanitizer, UBSan, Coverity, Valgrind. OpenGL, GLSL, OGRE, Coin3D.

Publications

Submitted

  1. Learning to Drive on Mars: Visual Multimodal Traversability Estimation for Off-World Navigation. D. Chiu, C. Wilson, A. Tumbar, G. S. Sukhatme, S. Myint. Submitted to ICRA 2027.

2026

  1. Radshield: Software Radiation Protection for Commodity Hardware in Space. H. Wang, S. Myint, V. Verma, Y. Winetraub, J. Yang, A. Cidon. ASPLOS '26, Pittsburgh, PA.
  2. A Software-Based Approach to Radiation Mitigation for Planetary Missions. J. Nash, V. Verma, E. Graser, H. Wang, G. Kubiak, M. Maimone, S. Myint, J. Biesiadecki, P. Romano. IEEE Aerospace Conference, Big Sky, MT.

2024

  1. Censible: A Robust and Practical Global Localization Framework for Planetary Surface Missions. J. Nash, Q. Dwight, L. Saldyt, H. Wang, S. Myint, A. Ansar, V. Verma. ICRA, Yokohama, Japan.
  2. Enabling Long & Precise Drives for the Perseverance Mars Rover via Onboard Global Localization. V. Verma, J. Nash, L. Saldyt, Q. Dwight, H. Wang, S. Myint, J. Biesiadecki, M. Maimone, A. Tumbar, A. Ansar, G. Kubiak, R. Hogg. IEEE Aerospace Conference, Big Sky, MT.

2023

  1. Mars Attacks!: Software Protection Against Space Radiation. H. Wang, S. Myint, V. Verma, Y. Winetraub, J. Yang, A. Cidon. HotNets '23.
  2. Robotic Deployment and Installation of Payloads on Planetary Surface. A. Trebi-Ollennu, K. Ali, C. Sorice, W. Kim, S. Myint, O. Khan, P. Bailey, H. Abarca, R. G. Deen, J. Yen, J. N. Maki, G. Lim, N. Udomkesmalee, J. Umland. Handbook of Space Resources, Springer.
  3. Results from the First Year and a Half of Mars 2020 Robotic Operations. V. Verma, M. Maimone, E. Graser, A. Rankin, K. Kaplan, S. Myint, J. Huang, A. Chung, K. Davis, A. Tumbar, I. Tirona, M. Lashore. IEEE Aerospace Conference, Big Sky, MT.

2022

  1. First 210 Solar Days of Mars Perseverance Robotic Operations: Mobility, Robotic Arm, Sampling, and Helicopter. V. Verma, F. Hartman, A. Rankin, M. Maimone, T. Del Sesto, O. Toupet, E. Graser, S. Myint, K. Davis, D. Klein, J. Koch, S. Brooks, P. Bailey, H. Justice, M. Dolci, H. Ono. IEEE Aerospace Conference, Big Sky, MT.
  2. Testing Mars 2020 Flight Software and Hardware in the Surface System Development Environment. S. Brooks, T. Litwin, J. Biesiadecki, N. Abcouwer, T. Del Sesto, M. McHenry, S. Myint, P. Twu, D. Wai. IEEE Aerospace Conference, Big Sky, MT.

2018

  1. InSight Mars Lander Robotics Instrument Deployment System. A. Trebi-Ollennu, W. Kim, K. Ali, et al. (incl. S. Myint). Space Science Reviews.
2015 and earlier

2015

  1. Testbed for Studying the Capture of a Small, Free-Flying Asteroid in Space. B. Wilcox, M. San Martin, L. Giersch, S. Howe, H. F. Grip, R. Nayeri, T. Litwin, J. Carlson, M. Shekels, A. Jain, C. Lim, S. Myint, J. Dunkle, A. Sirota, C. Fuller. AIAA SPACE 2015.

2014

  1. Vehicle Dynamics Models for Onboard Motion Planning. R. Mukherjee, T. Howard, S. Myint, J. Chang, J. Craft. ASME IDETC, DETC2014-34831.
  2. M3TK: A Robot Mobility and Manipulation Modeling Toolkit. R. Mukherjee, S. Myint, J. Chang, I. Kim, J. Craft, M. Pomerantz, J. Kim, L. Peterson. ASME IDETC, DETC2014-34832.
  3. Modeling and Simulation of Asteroid Retrieval Using a Flexible Capture Mechanism. H. F. Grip, M. Ono, J. Balaram, J. Cameron, A. Jain, C. Kuo, S. Myint, M. Quadrelli. IEEE Aerospace Conference, Big Sky, MT.

2013

  1. Real-Time and High-Fidelity Simulation Environment for Autonomous Ground Vehicle Dynamics. J. Cameron, S. Myint, C. Kuo, A. Jain, H. F. Grip, P. Jayakumar, J. Overholt. NDIA Ground Vehicle Systems Engineering and Technology Symposium, Troy, MI.

2012

  1. DspaceOgreTerrain 3D Terrain Visualization Tool. S. Myint, A. Jain, M. Pomerantz. NASA Tech Briefs.
  2. End-to-End Dexterous Manipulation with Deliberate Interactive Estimation. N. Hudson, T. Howard, J. Ma, A. Jain, M. Bajracharya, S. Myint, C. Kuo, L. Matthies, P. Backes, P. Hebert, T. Fuchs, J. Burdick. ICRA.

2011

  1. DspaceOgre 3D Graphics Visualization Tool. A. Jain, S. Myint, M. Pomerantz. NASA Tech Briefs.
  2. Large Terrain Modeling and Visualization for Planets. S. Myint, A. Jain, J. Cameron, C. Lim. IEEE SMC-IT, San Francisco, CA.
  3. A KML Interface for Dynamic Simulation of Robotic Planetary Exploration. T. Howard, J. Cameron, S. Myint, H. Nayar, A. Jain. IEEE Aerospace Conference, Big Sky, MT.

2010

  1. Surface Operations Analysis for Lunar Missions. H. Nayar, J. Balaram, J. Cameron, M. DiCicco, T. Howard, A. Jain, Y. Kuwata, C. Lim, R. Mukherjee, S. Myint, A. Malkovic, M. Pomerantz, S. Wall. AIAA SPACE 2010.

2009

  1. Recent Developments on a Simulator for Lunar Surface Operations. H. Nayar, A. Jain, J. Balaram, J. Cameron, C. Lim, R. Mukherjee, M. Pomerantz, L. Reder, S. Myint, N. Serrano, S. Wall. AIAA SPACE 2009.
  2. Dspace: Real-Time 3D Visualization System for Spacecraft Dynamics Simulation. M. Pomerantz, A. Jain, S. Myint. IEEE SMC-IT.

Open source

SpaceWasm and Honeycomb are JPL projects. GitHub featured SpaceWasm in Octoverse 2026. I created autoflake and later transferred it to PyCQA to maintain; it still gets about 5 million downloads a month from PyPI. I also used to develop pyflakes.

ProjectWhat it doesLanguage
nasa/spacewasmFlight-compliant WebAssembly interpreter for running code on board spacecraftRust
nasa-jpl/honeycomb3D visualization framework for roboticsTypeScript
PyCQA/pyflakesChecks Python source files for errorsPython
PyCQA/autoflakeRemoves unused imports and unused variablesPython
perceptualdiffCompares images using a perceptual metricC++
syntastic-extrasExtra syntax checkers for Vim's SyntasticVim script

Awards

  • 2026 JPL Team Award: developing and launching JPL's Rover Operations Center, for surface operations and use of AI in operations
  • 2025 NASA Group Achievement Award: development, V&V, and delivery of onboard global localization to Perseverance
  • 2024 NASA Group Achievement Award: coordinated Ingenuity and Perseverance operations through low-power winter conditions
  • 2024 NASA Group Achievement Award: Mars 2020 mobility and robotic arm operations processes
  • 2024 NASA Group Achievement Award: Mars 2020 robotic operations processes, through commissioning and record-setting surface operations
  • 2024 NASA Group Achievement Award: commanding Perseverance's mobility and arm systems
  • 2024 NASA Group Achievement Award: R13 flight software for Curiosity's primary computer
  • 2024 JPL Team Award: Ingenuity helicopter campaign, 72 flights on Mars
  • 2024 JPL Voyager Award: improving SSim and its controls in RSVP, including faster execution and flight-like AutoNav timing
  • 2023 NASA Exceptional Public Achievement Medal, for leading development of SSim for validation and verification of all uplink sequences to Perseverance
  • 2023 JPL Explorer Award, for delivering SSim for operating Perseverance on Mars
  • 2023 NASA Group Achievement Award: NASA's first collaborative targeting, automated sequencing, and flight-software-in-the-loop simulation toolset for robotic operations
  • 2023 JPL Team Award: NPM-Direct, which analyzes flight software and generates an NPM-decoding map
  • 2022 JPL Team Award: 5 km of driving in one month during the Mars 2020 Rapid Traverse campaign
  • 2021 JPL Discovery Award: tactical support of SSim through SOX checkout and first-time activities
  • 2021 JPL Team Award: technical leadership in developing SSim
  • 2021 JPL Team Award: Mars 2020 RSVP and SSim for surface robotic operations
  • 2020 JPL Team Award: Mars 2020 cruise, entry, descent, and landing flight software
2010–2019
  • 2019 JPL Voyager Award: simulation and validation capabilities in SSim
  • 2019 NASA Group Achievement Award: InSight Ground Data System
  • 2019 NASA Group Achievement Award: InSight Instrument Deployment System
  • 2019 NASA Group Achievement Award: InSight Wind and Thermal Shield
  • 2019 JPL Team Award: command, simulation, and telemetry tools for the Mars 2020 V&V testbeds
  • 2018 JPL Discovery Award: integration, test, and operations readiness of the InSight IDS
  • 2017 NASA Group Achievement Award: Curiosity surface operations
  • 2016 JPL Team Award: visualization of robot arm target positions and orientations
  • 2016 JPL Discovery Award: leading the Mars 2020 SSim task
  • 2015 JPL Team Award: View Shed capability in RSVP
  • 2015 JPL Team Award: UGV night operations, sensor fusion, and coordinated tactical behaviors
  • 2013 NASA Group Achievement Award: landing of Curiosity at Gale Crater
  • 2013 NASA Group Achievement Award: Curiosity surface mission at Gale Crater
  • 2013 JPL Discovery Award: rendering of an orbiting aerosol cloud simulation for NIAC
  • 2013 JPL Team Award: real-time 3D visualization of MSL during entry, descent, and landing
  • 2011 JPL Team Award: Lunar Sample Return robotics demonstration
  • 2011 NASA Tech Brief Award: DspaceOgreTerrain
  • 2010 NASA Tech Brief Award: DspaceOgre
  • 2010 JPL Mariner Award: multi-gigapixel terrain modeling and visualization software