At a Glance

  • Texas Spacecraft Laboratory collaboration with NASA Johnson Space Center (JSC) on the Seeker-1 CubeSat mission.

  • Goal of Seeker-1 is to demonstrate capabilities relevant to on-orbit, free-flying inspection of crewed space vehicles to improve safety and identify anomalies early.

  • TSL contracted by JSC to develop a novel visual navigation system for Seeker-1 that provides azimuth and elevation of a target spacecraft using only a monocular camera.

  • Shared first author on a paper covering this reseach published at the 2020 AIAA SciTech Forum and Exposition in January 2020. Read it here.

  • Personal Contributions:

    • Designed, implemented, and tested all flight software.
    • Integrated machine learning and computer vision algorithms on flight hardware.
    • Developed a suite of logging and data visualization scripts that characterize algorithm performance onboard flight hardware.
    • Created technical documents detailing setup of mission hardware/testing procedures to simplify integration for flight.

  • Project Timeline:

    • Seeker Vision delivered to NASA JSC in May 2018 and selected for integration among competing solutions.
    • Seeker-1 CubeSat launched on Cygnus NG-11 on April 17, 2019.
    • Seeker-1 primary mission was conducted on September 16, 2019.

System In Action

alt text Seeker Vision In Action, Ground Testing alt text Seeker-1 Deploy, Sep 16, 2019 alt text Seeker Vision In Action, Sep 16, 2019

Details

The Texas Spacecraft Laboratory (TSL) is a student-driven research group dedicated to designing and building small satellites, securing launches into space, and operating them once in orbit. I joined the lab as a sophomore undergrad and was privileged to see a project through from start to final delivery as a team lead during my sophomore year. I worked as the Flight Software and Hardware lead on the Seeker Vision project. Seeker Vision was a collaboration between NASA Johnson Space Center and UT Austin as part of the Seeker 1 CubeSat mission. The mission serves as a first technical demonstration of small, free flying, autonomous spacecraft which will one day be used to inspect stations and spacecraft for damage during flight. UT Austin was contracted to build a full vision system for the CubeSat, including flight software, novel vision algorithms, and robust error handling. The system was to use only a camera and microcomputing board - access to other sensors onboard the satellite was restricted. Using Tensorflow, our team built a convolutional neural network to detect the target spacecraft, a Northrop Grumman Cygnus resupply ship (pictured above), within the camera’s frame of view. Once the target was identified, OpenCV was used to determine relative azimuth and elevation of the target spacecraft. This information was then passed to the main flight computer and used as part of the Guidance, Naviation, and Control for the CubeSat. As the Flight Software and Hardware lead, I was present at all three mission reviews and presented slides about the state of the various areas of the project I was working on. My responsibilities were threefold:

  • Design, implement, and test flight software to handle boot sequence, facilitate vision algorithms, and communicate with main flight computer
  • Develop a suite of logging and data visualization scripts to characterize algorithm performance onboard flight hardware
  • Create technical documents detailing setup of mission hardware/testing procedures to simplify integration for flight

My team and I successfully built flight software using two-tier process monitoring to prevent crashes on orbit. The entire vision system was installable with a single terminal command and ran automatically on boot if a camera was connected. After rigorous testing, the system was delivered to JSC in May of 2018 with technical documents which detailed its use and integration.

The delivered vision system was selected over competing soltutions for integration on the Seeker 1 CubeSat mission that launched in April 2019. Its ability to successfully handle worst case scenarios in testing was cited as a primary reason for its selection.

The Seeker-1 60 minute demonstration mission took place in September of 2019 and results are still being analyzed.