Research analysis

A 58-page review of wireless energy transfer for space introduces an R0 to R7 reference-plane framework so that beaming efficiencies reported at different electrical boundaries stop being compared as if they were the same quantity. Its closure audit of six flight and ground experiments finds that no demonstration, from Goldstone 1975 to MAPLE and Tianzhou-7, has validated a complete source-to-useful-load chain, and that the segments everyone skipped are exactly where space power engineers live.

An Oxford and Toronto team simulated how a swarm of 4 m, 20 kg wind-propelled tumbleweed rovers disperses inside a Martian global dust storm. The swarm holds together as a radio mesh over a full sol with a per-rover range of just 1.6 to 4.1 km, which converts almost the entire surface mobility budget into communications, sensing, and storage instead of drive power.

Guzman, Babiceanu and Pekoz built a Boeing 787-inspired 400 Hz power system model, generated 21 classes of faults and disturbances, and trained a Compact ResNet that classifies them at 96.94% with only 175,685 parameters. Deployed with 8-bit quantization on a Xilinx ZCU102 FPGA it holds 95.87% accuracy at 6.90 ms mean latency, but every waveform it was tested on came from the same simulator that trained it.

LAAS-CNRS researchers built a rectenna that either harvests incoming RF power or modulates its reflection in two orthogonal polarizations, controlled by two GPIO lines. Bench tests show 25 to 29% RF-to-DC conversion at minus 5 dBm and a 16-byte AES-128 payload backscattered over a 2-meter link, a meter-scale proof of concept whose polarization trick has a real space-power parallel.

By 2003 a NASA-led, multi-institution program had moved laser-photovoltaic wireless power transmission from the bench to initial open-air tests at the AMOS facility on Maui. The record documents what was tested, but it publishes no range, power, or efficiency figure, and that absence is as instructive as the milestone.

A NASA white paper from the 2023 Moon to Mars Architecture Concept Review sizes human Mars surface power from about 10 kW at the low end to megawatt class for ISRU-intensive architectures. Its harder message is operational: the power plant must deploy itself, years ahead of crew, with no possibility of repair.

NASA Langley's OPAL concept study beams laser power from lunar orbit to surface assets through the night, and its central finding is about the beam, not the laser: focusability is what makes the system safe and light. Every efficiency figure in the chain is an assumption, and the study says so.

Todd Tofil's ASCEND 2022 presentation is the document where NASA's Fission Surface Power requirements first appear as a fixed table. The core numbers survived into 2024, but the schedule did not.

Flores et al. demonstrate in simulation that a rendezvous with Halley's comet in 2061 is feasible with a 36 mN Hall thruster drawing 640 W, powered by radioisotope generators. The power ceiling, not the rocket equation, is what forces two gravity assists and a 24-year flight.

NASA Glenn's 2022 hibernation study found ordinary 18650 lithium-ion cells survive cryogenic lunar nights in vacuum, after one-atmosphere tests killed two of five through a packaging failure mode. The harder problem is the dawn: a small set of cryo-capable power electronics must detect, preheat, and pre-charge a frozen battery before the rest of the spacecraft can wake.

NASA Langley's 2023 technical memorandum examines a Nuclear Thermionic Avalanche Cell (NTAC) that would heat atmospheric CO2 ingested through a nose port to 2800 K for Mars entry retropropulsion, then separate after landing to serve as a multi-megawatt surface power plant. Every number in it is analysis, not hardware, and the document says so.

NASA Glenn's 2024 IAC paper lays out the growth of lunar surface power from self-sufficient elements to a utility grid, and commits the Universal Modular Interface Converter to a 3000 VAC, 3-phase, 1000 Hz backbone. The driver is not fashion: space-rated power switches cap realistic high-voltage DC at about 1.5 kV, and AC wins on mass and extensibility.

Before concluding that space-based solar power loses on cost, NASA's Office of Technology, Policy, and Strategy had to invent a way to cost it. The 2022 methodology paper shows an Aerospace Corporation bottoms-up estimate with learning curves, a first-of-a-kind to nth-of-a-kind progression, and a lifecycle carbon accounting that includes launch and disposal.

Under the Mars Transportation Assessment Study, NASA, DOE, and ORNL traded reactors, shielding, conversion, and radiators for a 1.9 MWe crewed Mars nuclear electric propulsion power system. The reference concept lands near 25,000 kg at about 13 kg/kWe, and the study's own parametrics show the design is capped not by reactor physics but by stowing 2500 m2 of radiator in an 8.4 m fairing.

NASA Glenn's Tom Kerslake presented a full performance model for Mars surface solar arrays at the FISO working group in June 2017: dust costs 0.14% of power per sol, reddened light costs triple-junction cells about 10%, and a 10 kW module pairing IMM arrays with a regenerative fuel cell weighs in near 1.5 tonnes. It is a modeling study, not flight hardware, and it argues simple power estimates can miss by more than 5X.

Giudici's 1987 assessment of Mars power systems, N87-17795, derived the scaling law that still governs space power beaming: laser transmission closes at hundreds of kW if 6.3 percent end-to-end efficiency is achievable, microwave links need GW scale, and a Phobos station at 14 km orbit changes the answer. Every figure is an analysis assumption, not a measurement.

NASA/TM-20220006678 trades three power architectures for a 21-day crewed Mars rover excursion: a 3.4 kWe dynamic radioisotope system with a 52.7 kWh battery, regenerative fuel cells, and a 73 square meter solar wing. The RPS baseline is 2.5 times lighter than a 21-day fuel-cell system, but the study is explicit that the winning system is TRL 3 to 4.

A NASA Johnson Space Center study team sized surface power for a conceptual 500-day crewed Mars mission and found the Mars Ascent Vehicle's oxygen production, not the crew, sets the power budget. Trading one 40 kWe fission reactor for a farm of small Kilopower units saved up to eight metric tons on paper, at the price of connectors, enriched uranium, and field wiring.

Before the ISS existed, NASA Lewis engineer Fred Teren drew the Space Station power management and distribution system down to the component level: rotary power transfer through the solar-array joints, remotely switched buses, and a planned 40 kW, 20 kHz AC ground demonstration. The paper is a schematic record, but its trade tree still frames every PMAD decision made since.

NASA Glenn engineers trading AC against DC distribution for a 100 kWe nuclear-electric deep-space spacecraft found the DC bus 218 kg against 257 kg for AC, then recommended the heavier system. The reasons are a masterclass in what mass-optimal power design misses.

NASA's Fission Surface Power project procured a lunar reactor with three minimal requirements, three competing $5M industry studies, and a government reference design built in parallel as the evaluation yardstick. The 2022 plan targeted a flight unit by end of 2028; by late 2024 the Phase 2 solicitation had still not been released.

A NASA webinar from August 2025 lays out the draft instrumentation and controls development path for the 40 kW lunar fission surface power system. The hard part is not the reactor physics; it is fielding sensors that survive in-core neutron fluence and gamma doses orders of magnitude beyond terrestrial nuclear practice, with no maintenance for a decade.

NASA's Office of Technology, Policy, and Strategy evaluated two representative space-based solar power designs and found baseline lifecycle electricity costs far above terrestrial alternatives. A specific, jointly necessary bundle of launch, manufacturing, lifetime, and efficiency assumptions is the only modeled path to competitiveness.

The NASA Glenn team that proposed lunar-night hibernation in 2022 reported its first measurements in July 2023: two named commercial 18650 cells survived freeze-thaw under vacuum at about 50 K with 14-day dwells, and a cryo-operable shunt regulator started cold at 50 K after a 16-hour soak. The evidence base is two cycles on a small cell set, with control electronics and cycle life still open questions.