Document Type
Paper
Publication Date
Spring 2025
Abstract
As small satellite missions grow in complexity, there is increasing demand for greater onboard computational performance. Traditional radiation-hardened processors, while reliable, are expensive, slow, and suffer from long procurement timelines. In contrast, commercial off-the-shelf processors like the Raspberry Pi offer high performance at low cost but are vulnerable to radiation-induced faults such as single-event upsets. SkyForge Core is an experimental computing system that addresses this challenge by integrating three Raspberry Pi Zero W units into a triple modular redundant architecture. The system executes identical tasks on each Pi, compares outputs using majority voting, and identifies discrepancies. If a failure occurs, the system continues operation while executing a recovery operation on the affected processor. This architecture will be demonstrated aboard a Near Space Launch smallsat in early 2026. SkyForge Core aims to validate the architecture, collect data on failure rates in low earth orbit, and characterize system performance in space.
Recommended Citation
Staritz, Peter; Schmitt, Harrison; McWatters, Thomas; Hoyt, Ethan; Ellman, Josh; Sikander, Sameer; Westrum, Nate; Clark, Jordan; Jung, Taebaeksanmaek; Williams, Micah; Myers, Lauren; Wright, Kelden; Homburg, Matthew; Roth, Alex; and Zhang, Jinran, "SkyForge Core: A Triple Modular Redundant Computing Architecture for Small Satellites" (2025). Physics & Engineering Student Papers. 1.
https://pillars.taylor.edu/physics-engineering-student-papers/1
Poster for SSC25-P1-76 Paper
Notes
Courses:
ENP 495, Engineering Senior Capstone III (Dr. Peter Staritz)
ENP 392, Junior Engineering Project (Dr. Peter Staritz)