Laser-photovoltaic power beaming reached open-air tests in 2003, and published no numbers

What the work claims

Fikes, Henley, Mankins, Howell, Fork, Cole and Skinner report the state of laser-photovoltaic wireless power transmission (WPT) as of 2003: that it can be accomplished over long distances with laser sources and photovoltaic receivers, and that a multi-institution effort had matured the technology through systematic component testing to the point of initial open-air tests at the AMOS facility, with safe operating procedures established¹. The tested matrix is specific: gaseous argon and krypton lasers, solid-state diode lasers and fiber-optic sources, spanning wavelengths from the visible to the near infrared, each fired at silicon and gallium arsenide photovoltaic receivers¹. The institutions named are NASA (Marshall Space Flight Center and Headquarters), Boeing, the University of Alabama in Huntsville, the University of Colorado, Harvey Mudd College and the Naval Postgraduate School¹. The record states the work is progressing toward longer-distance ground demonstrations and practical near-term space demonstrations; it contains no flight result¹.

The record also documents the program's institutional frame: Fikes was the contracting officer's technical representative for the Space Solar Power Technology Demonstration for Lunar Polar Applications, a contracted effort with Boeing North America to mature space solar power technology applicable to a lunar polar mission through ground demonstrations¹. This is an extended conference abstract, not a full paper; one page carries all of its technical content. Everything below weights it accordingly.

How it works

Laser-photovoltaic WPT is a five-link chain. DC electricity drives a laser (wall-plug efficiency); the beam is formed, pointed and propagated to a receiver (pointing and divergence losses); the receiver collects the light and converts it to DC electricity (receiver efficiency). The engineering heart of the concept is the match between laser wavelength and receiver bandgap: a photovoltaic cell converts photons at or above its bandgap energy, so the laser's wavelength must sit on the cell's useful response curve, which is why the program tested sources from the visible to the near infrared against both silicon and gallium arsenide rather than committing to one pairing¹.

The companion program paper supplies the measured component figures the 2003 record omits. Using an 830 nm laser wavelength, selected for high atmospheric transmissivity, with gallium arsenide cells, the University of Colorado at Boulder measured close to 70 percent efficiency in converting laser light into electricity (Figure 12); Japanese researchers measured 34 percent at 805.8 nm using silicon cells (Figure 12, citing Takeda and Kawashima, 1999)². Harvey Mudd College built a specialized laser beam expander for precise pointing and focusing, and initial tests held near-constant beam dimensions over a one kilometer range (Figures 9 and 10)². The measurement boundary of the 70 percent figure matters: it covers the receiver link only, laser light in to electricity out at the cell. It excludes laser wall-plug efficiency, beam propagation, pointing and collection. Neither record publishes an assembled end-to-end chain efficiency, so no such number can be quoted honestly from this program.

The strongest case for it

The test matrix is exactly the right work for a maturing technology. Four laser families against two photovoltaic materials across the visible to near infrared is a systematic mapping of the wavelength-bandgap trade space, the kind of unglamorous component engineering that has to exist before any architecture claim deserves trust. The safety work is equally significant: the record says safe operating procedures were established and only then were initial open-air tests conducted, which is the correct order and the order regulatory reality demands for anything that will eventually fire energy beams across spaceports, ranges or the lunar surface¹. And the receiver choice is quietly brilliant economics: silicon and gallium arsenide were, in 2003, the standard materials of the space solar array industry, so a laser power receiver rides the existing space photovoltaic manufacturing, qualification and radiation-hardness base rather than requiring a bespoke receiver industry of its own¹. A program with a named government technical representative, a named prime contractor and a named application (lunar polar power) is a program past the slide-study stage.

Where a skeptic should push

The load-bearing problem is that the milestone record contains no numbers at all. We read the complete 3-page NTRS record: there is no range, no aperture, no beam power, no link efficiency, no beam quality figure anywhere in it. "Recent accomplishments" announces a phase of work, not a result. The phrase "progressing toward longer distance ground demonstrations" concedes that even the ground link had not yet been demonstrated at range when the record was written, and "practical near-term space demonstrations" is 2003 intent; the record itself documents no flight hardware and no space demonstration, so nothing here should be read as flown or flight-qualified capability. On the standard maturity scale, tested commercial components and initial open-air ground tests amount to breadboard validation in a relevant ground environment: not thermal-vacuum tested, not flight-qualified, not flown.

The most load-bearing assumption behind the program's optimism is that receiver conversion measured at benign bench conditions scales to kilometer-class ranges through a real atmosphere on Earth and, later, to thermal-vacuum and radiation conditions in space. The records are also silent on laser wall-plug efficiency, historically the weakest link in any laser power chain; a 70 percent receiver behind a weak laser is a weak system, and neither record lets a reader assemble that product. Treat every efficiency associated with this program as receiver-boundary or component-boundary only, because that is the only boundary the sources actually measured.

What the AMOS tests mean for power in space

The non-obvious implication is that optical beaming's receiver risk was never the receiver. Because the 2003 test matrix paired lasers with exactly the silicon and gallium arsenide cells the industry already flew on sunlight¹, a laser receiver inherits the entire space photovoltaic base: qualification heritage, radiation data, manufacturing capacity, and the option of an aperture that accepts both sun and laser. Beaming programs compete on lasers, pointing and safety cases, but the receiving end is a commodity solar array, which means receiver cost and reliability improve on the space solar array curve whether or not beaming itself succeeds.

The threat is the demonstration-without-numbers pattern this record exemplifies. Two decades later, NASA's OPAL lunar orbital beaming concept still has to assume its client receiver optical-to-DC efficiency at 40 percent in its key assumptions slide, because published end-to-end laser-PV chain measurements remain scarce enough that a 2026 architecture has nothing better to anchor on³. Programs that demonstrate in public without publishing per-link numbers force every successor study to guess, and the guesses drift optimistic. The calibration discipline this site applies to beaming chains, measured links versus assumed links, exists precisely because records like this one leave the gap unfilled.

The opportunity is geometry plus precedent. The companion program paper's lunar polar analysis shows that a beam from a 1 kilometer peak reaches roughly 50 km and from a 6 km peak roughly 120 km before grazing the horizon, and that relay mirrors extend light into permanently shadowed craters (Figure 5)². Pair that geometry with an established open-air safety-procedure baseline¹ and the path opens to powering small science assets inside shadowed polar craters from sunlit peaks, an application that needs beamed power at modest scale rather than a landed reactor. The concept is a direct ancestor of the same team's lunar polar demonstration study, covered in our earlier analysis of the companion paper.

The bottom line

High confidence: by 2003 the program had tested four families of commercial lasers against silicon and gallium arsenide receivers across the visible to near infrared, written safe operating procedures, and conducted initial open-air tests at the AMOS facility; the record states each of these plainly¹. Moderate confidence: receiver-side conversion in the 70 percent class (gallium arsenide at 830 nm, measured by the University of Colorado at Boulder under the same program) was real bench hardware performance, bounded to the receiver link only². Unestablished: any end-to-end chain efficiency, any space demonstration, any flight capability; the record claims progress and intent, not results. What would confirm the claim: a published end-to-end measurement with distance, apertures, delivered power and per-link efficiencies. What would break it: a full-chain efficiency that collapses once laser wall-plug and pointing losses are included, which is exactly the measurement this program's public records never made.