group
What campus are you from?
Daytona Beach
Authors' Class Standing
Kylie Nager, Senior Akash Joseph, Graduate Student James Kirk, Senior Payton Miller, Sophomore
Lead Presenter's Name
Kylie Nager
Faculty Mentor Name
Dr. Cagri Kilic
Abstract
Satellite constellations in near-Earth orbit are increasingly vulnerable to solar flares, coronal mass ejections, and high-speed solar wind streams that disrupt communications, power systems, and orbital stability. Current mission architectures rely on delayed ground intervention and do not leverage predictive solar data for autonomous response. The HELIO project addresses this gap by developing an AI-driven resilience layer that converts real-time heliophysics forecasts into coordinated operational decisions for CubeSat swarms. HELIO integrates pretrained solar forecasting models such as NASA and IBM’s SURYA foundation model to predict active-region evolution, flare probability, and solar-wind velocity. A lightweight decision-intelligence framework interprets these forecasts and maps them to disturbance levels (Normal, Moderate, Severe), which in turn trigger corresponding spacecraft operating modes (Nominal, Harden, Safe). This framework accounts for onboard constraints such as battery margins, communication windows, and network redundancy to autonomously reconfigure CubeSat operations before solar-weather impacts occur. The project will validate this forecast-to-decision architecture through two complementary paths: (1) back testing HELIO forecasts against Solar Dynamics Observatory (SDO–AIA/HMI) data to verify physical consistency, and (2) conducting simulation-based swarm trials to quantify improvements in network availability, power stability, and response latency. By linking foundation-model forecasts with spacecraft decision logic, HELIO introduces a forecast-informed autonomy paradigm for space systems. The outcome will be a validated prototype, open-source dataset, and performance benchmarks that demonstrate measurable resilience gains under solar-weather stress. These results will directly support proposals to NASA ROSES Space Weather R2O2R and NSF Cyber-Physical Systems programs, advancing the frontier of AI-enabled operational resilience in CubeSat constellations and laying the groundwork for higher-TRL flight demonstrations.
Did this research project receive funding support from the Office of Undergraduate Research.
No
Heliophysics Enhanced Learning for Intelligent Orbits (HELIO)
Satellite constellations in near-Earth orbit are increasingly vulnerable to solar flares, coronal mass ejections, and high-speed solar wind streams that disrupt communications, power systems, and orbital stability. Current mission architectures rely on delayed ground intervention and do not leverage predictive solar data for autonomous response. The HELIO project addresses this gap by developing an AI-driven resilience layer that converts real-time heliophysics forecasts into coordinated operational decisions for CubeSat swarms. HELIO integrates pretrained solar forecasting models such as NASA and IBM’s SURYA foundation model to predict active-region evolution, flare probability, and solar-wind velocity. A lightweight decision-intelligence framework interprets these forecasts and maps them to disturbance levels (Normal, Moderate, Severe), which in turn trigger corresponding spacecraft operating modes (Nominal, Harden, Safe). This framework accounts for onboard constraints such as battery margins, communication windows, and network redundancy to autonomously reconfigure CubeSat operations before solar-weather impacts occur. The project will validate this forecast-to-decision architecture through two complementary paths: (1) back testing HELIO forecasts against Solar Dynamics Observatory (SDO–AIA/HMI) data to verify physical consistency, and (2) conducting simulation-based swarm trials to quantify improvements in network availability, power stability, and response latency. By linking foundation-model forecasts with spacecraft decision logic, HELIO introduces a forecast-informed autonomy paradigm for space systems. The outcome will be a validated prototype, open-source dataset, and performance benchmarks that demonstrate measurable resilience gains under solar-weather stress. These results will directly support proposals to NASA ROSES Space Weather R2O2R and NSF Cyber-Physical Systems programs, advancing the frontier of AI-enabled operational resilience in CubeSat constellations and laying the groundwork for higher-TRL flight demonstrations.