No projects match the selected filters.
Advisor: Prof. Daniel Selva
Institution: Texas A&M University, Systems Engineering, Autonomy and Knowledge Laboratory (SEAK)
Funding: NASA Human Research Program (Grant #80NSSC19K0656) & NASA STTR Phase II (Award #80NSSC24CA028)
Advisor: Prof. Shinichi Nakasuka
Focus: Conceptual design of a cultural facility module for low Earth orbit space stations
Collaboration: JAXA
Advisors: Dr. Masato Sakurai & Prof. Shinichi Nakasuka
Focus: Equivalent System Mass (ESM) calculations and comparative analysis for lunar pressurized rover systems with regenerative life support
Advisors: Prof. Shinichi Nakasuka & Dr. Toshiki Tanaka
Funding: MEXT Space Exploration Fundamental Technology Advancement Program
Focus: Nano-satellite navigation and IoT communication services for lunar missions
Projects: SPHERE EYE-1 and ONGLAISAT geostationary satellite missions
Responsibilities: Structure development, CAD analysis, vibration testing, hardware assembly management
Project: EQUULEUS 6U CubeSat lunar mission (launched on NASA SLS, Nov 2022)
Responsibilities: Thermal operations from critical to nominal phases, heater control, attitude control
This research focuses on developing autonomous anomaly resolution systems for space habitats using physics-based diagnostic models and digital twin simulations. The work is part of a larger effort to support astronauts during long-duration space exploration missions where communication delays with mission control make real-time support impossible.
A web-based application with natural language interface connected to real-time telemetry from NASA's Habitat Support System (HSS)
Real-time telemetry data assimilation into physics-based models to create habitat digital twins
Integration with robotic agents for NASA's Gateway mission
Physics-based reasoning capabilities for proactive anomaly detection and diagnosis
solid = implemented · dashed = optional / not yet closed-loop
Telemetry stream · physical state
NASA Human Exploration Research Analog (HERA): 68 telemetry parameters with caution / warning limits and 31 anomaly scenarios across CO₂ removal, O₂ generation, fans / TCCS, fuel cells & power, water pressure & leaks, and trace contaminants.
Daphne-AT
When is “anomaly” declared? — limits vs learned binary score.
Daphne-AT
Shared catalog: detection and hypotheses stay fixed while the evaluation method is swapped.
Daphne-AT
Hosts several diagnostic methods side by side: knowledge-graph symptom matching, Bayesian network inference, time-series matching. Every evaluator outputs a probability distribution over hypotheses.
Daphne-AT
Optional / not yet closed-loop.
Daphne-AT
Run top hypothesis procedures in series. Verify on alarms; else eliminate & re-rank. Orderings: P/T · Highest P · Shortest T · Random.
PRIDE + CRAFTSMAN
Adjustably autonomous procedure execution, and semi-autonomous IVA and EVA robotic caretakers that perform inspection and repair. Robotic inspection evidence refines hypotheses.
