Research analysis

An NSF-funded project at Michigan State University, led by Principal Investigator Navid Yazdi, investigates multi-layer optical interconnects and thermal-management technologies for high-density chiplet-based 3D heterogeneous integration. The central objective is to develop high-density vertical optical waveguides and optical I/O for high-speed communication between stacked layers. The second objective is a scalable thermal-management architecture using modular thermal-interface chiplets that provide low thermal resistance while mechanically decoupling chip stacks from shared cold plates. The project will measure loss, crosstalk, data rate, bit-error rate, alignment tolerance, thermal resistance, heat-removal capability, and pressure drop, and will integrate both technologies in a module-scale 3D test vehicle.

In a 1984 conference paper at NASA Lewis Research Center, F. J. Redd and E. V. Fornoles of the Air Force Space Technology Center argued that growing interest in new classes of military and civil space systems demanding substantial power was driving renewed interest in space-qualified nuclear power. They identified power as a limiting technology for future space goals and noted that development speed was constrained by the absence of clear system requirements. The needs they described for robust surveillance, survivable communications with anti-jam, electric propulsion, and weapons applications remain relevant today.

An NSF EPSCoR Research Fellowship led by Efren Lopez Morales at New Mexico State University, collaborating with UC Santa Cruz, develops a formal framework for modeling how AI systems transform inputs into spacecraft commands, characterizing AI-enabled command manipulation attacks, and building runtime validation mechanisms that check AI-generated commands against mission and safety constraints before execution. The project uses an AI-enabled satellite command prototype with realistic spacecraft command and telemetry interfaces. It is funded research at the framework development stage with no deployed results as of the award date.

An NSF-funded project at Penn State Behrend combines a curated photovoltaic device database, physics-based drift-diffusion modeling, nanoscale electrical characterization, and interpretable machine learning to discover design rules for stable mixed-cation perovskite solar cells. The framework links material composition, fabrication processes, and device architecture to efficiency and stability, but no device results have been demonstrated yet as of the award date.

NSF Award 2610941, led by PI Ifana Mahbub at the University of Texas at Dallas, is a three-year, $506,406 program that develops the Digital-Twin-Accelerated Inverse Metasurface (DT-AIM) framework for designing large reconfigurable metasurfaces. The framework integrates electromagnetic simulation, reduced-order modeling, physics-informed machine learning, and hardware-in-the-loop validation. Goals include a tenfold reduction in design time and a 90 percent decrease in the need for full-wave electromagnetic simulations. The project will fabricate and test real metasurface tiles. Applications span antennas, photonics, acoustics, and wireless power transfer including power-beaming systems. No deployed system is claimed.

An NSF-funded project at Virginia Tech, led by Principal Investigator Mantu K. Hudait, investigates high-mobility germanium-on-aluminum-arsenide nanosheet field-effect transistors and complementary FET (CFET) architectures for sub-nanometer technology nodes. The project uses ultra-high vacuum molecular beam epitaxy to develop stacked Ge-on-AlAs nanosheet structures, targets 0.5V operation with minimal power consumption, and employs Synopsys TCAD and Verilog-HDL predictive modeling with thermal effects. Circuit-level demonstrations of inverters, ring oscillators, and 6T SRAM cells are planned. No devices have been fabricated yet as of the award date.

NSF Award 2421927 (EAGER: CHIPS in Space) was a $300,000 project led by PI David A. Staack at Texas A&M that explored the scientific, technical, and economic feasibility of semiconductor manufacturing in space. The project reviewed more than 180 prior microgravity experiments, found substantial evidence that reduced gravity improves crystal quality and material performance, produced one of the first techno-economic assessments of space-based semiconductor manufacturing, organized two national workshops, and established the In-Space Semiconductor Manufacturing Alliance. The findings are assessment-level, not flight demonstrations.

An NSF-funded project at Arizona State University develops a hierarchical device-to-circuit co-optimization framework for gallium nitride high-electron-mobility transistors operating from room temperature down to cryogenic conditions. The framework connects full-band Monte Carlo transport physics to compact circuit models, targeting applications in quantum computing, space missions, and superconducting equipment where electronics must operate reliably at extremely low temperatures.

NSF Award 2629145 funds a four-year, $499,837 project led by PI Edward J. Oughton at George Mason University to build the first end-to-end framework that translates uncertainty in space weather forecasts into actionable risk information for satellite operations. The framework integrates machine learning and physics-based models to characterize geomagnetic storm occurrence, severity, and timing, propagate atmospheric drag effects through orbital dynamics, and produce probabilistic measures of conjunction risk, maneuver burden, orbital degradation, and service disruption. The project is at the framework-development stage and has not yet produced deployed results.

A 1989 NASA Lewis concept paper describes a microwave electrothermal thruster that heats propellant gas through plasma discharge using 915 MHz CW microwave power. The design targets 30 kW input power and over 95 percent energy absorption efficiency, but the paper describes apparatus being assembled, not measured test results.

An NSF SBIR Phase I award to Verde Technologies investigates perovskite solar cells fabricated directly on metal foil substrates instead of conventional glass or plastic, targeting a lightweight, flexible, and potentially recyclable solar technology. The key technical risk is that metal foil surfaces are rougher and harder to process, which can introduce defects and limit device stability, and the project aims to overcome this through interface engineering and transparent top contact design.

NASA Marshall Space Flight Center studied an operational space-based solar power station placed in lunar orbit to beam energy continuously to a lunar village, removing two of the strongest constraints on base location: the need for south pole sunlight and the need for a fission reactor. The concept is a system-level architecture study presented at the 2017 IAA Symposium on the Future of Space Exploration in Turin, Italy. It is not flight hardware or a funded development program.

A 1984 NASA conference paper by C. R. Baraona at NASA Headquarters outlined the strategies, reasoning, and planning guidelines for the US Space Station Program power system. The paper argues that power is a key driver of overall Space Station design, and conversely, that Space Station requirements drive power technology. It discusses technology options, mission analysis, weight, area, payload, and altitude requirements. The architecture it describes became the foundation for the International Space Station power system, the largest space power system ever flown.

NSF Award 2616563, led by PI Besma Smida at the University of Illinois, is a four-year, $950,000 theoretical program that challenges the conventional split between power harvesting and information decoding in SWIPT systems. The project introduces unified receivers that use the same low-power analog hardware to rectify energy and demodulate data from a composite RF signal, without local oscillators. Three goals span analytical modeling, memoryless low-rate bounds, and memory-aware high-rate bounds. No hardware demonstration is claimed.

An NSF-funded project at Texas Tech University, collaborating with Los Alamos National Laboratory, develops a physics-based understanding of how radiation alters charge transport in AlGaN/GaN HEMTs used for RF power amplifiers in satellite communications and deep-space missions. The project challenges the gamma=2 invariant assumed in compact models, showing physics-extracted gamma values that rise from approximately 4.5 at 4K to 6.5 at 550K. The goal is an open-source Radiation-Aware ASM-HEMT Library that predicts post-irradiation RF performance without empirical adjustment.

An NSF-funded project at the University of Washington develops a unified modeling framework for understanding how internal power converters, control loops, and energy storage in data centers transform computing activity into the effective load and impedance seen by the electric grid. The framework aims to enable scalable stability analysis and control design for power grids serving dynamic AI and computing workloads, and the modeling approach is transferable to in-space microgrids.

A government team from NASA Glenn, Los Alamos, and Idaho National Laboratory developed an independent 40 kWe lunar fission surface power concept, trading Stirling against Brayton conversion, sodium-potassium against water cooling, and shielding geometry against launch mass. The result is a reference architecture for architecture studies, not a flight-ready design.

An NSF-funded project at Penn State proposes using electron wind force, the momentum transfer from conducting electrons to defect structures, to recover gallium nitride power electronics degraded by radiation exposure in extreme environments. The method is non-thermal, electrically programmable, and aims to extend device life without turning systems off or applying heat.

In 1978, NASA Space Power Systems Branch chief J. P. Mullin presented a program overview at Space Congress in Cocoa Beach arguing that total energy demanded by future space missions would exceed past needs by orders of magnitude. Electric propulsion and high power sensors would require systems lighter by factors of ten or more than anything that had flown. The paper describes NASA's research program across energy conversion, storage, management, space environment interactions, and advanced concepts. It is a program overview, not hardware. The striking finding is that the specific power, cost per watt, and gap between needed and demonstrated systems that Mullin identified in 1978 are the same concerns driving space power R&D today.