A laser kept a model drone flying. The heat nearly killed the experiment.

A model drone in a lab in Tianjin flew on laser power, and the device that turned the beam into electricity hit 38.49% efficiency. Researchers at Civil Aviation University of China and Tsinghua University published the result on 29 July 2026 in the journal Matter & Light, framing it as the first peer-reviewed measurement of laser-to-electricity conversion on a wing-mounted receiver in flight.

The drone was a model aircraft, not an operational one. A green laser struck a perovskite laser cell-thermoelectric (PLC-TE) tandem device fixed to the underside of its wing. The tandem is the equivalent of a solar cell tuned to a laser wavelength: incoming photons drive a perovskite layer that produces current, and the heat the cell cannot convert is captured by a thermoelectric stage below it. In the test, 38.49% of the incident light reached the receiver as electricity. The team compared that figure to the U.S. Department of Energy's 34% peak benchmark for perovskite-silicon tandem cells, a related but different device class designed for sunlight rather than a single-wavelength laser.

The same number that makes the result real is the one that makes it bounded. Under the test laser the receiver's surface temperature climbed to 80–90°C (176–194°F). Above that range the perovskite layer's efficiency drops sharply, and the team identified the heat as the dominant constraint on the system. Han's framing is that the work shifts the bottleneck rather than closing it. "The materials are no longer the limiting factor," senior author Jianhua Han of Civil Aviation University of China said in the institutional release. "The challenge is now systems engineering: mounting, cooling, and making the receiver compatible with a flying aircraft."

That reframe matters because the wire copy of this result is going to treat 38.49% as the story. A reader who walks away with that single number has the wrong model of where the field is stuck. The materials problem, which is getting a photovoltaic material to convert enough of a laser's light into electricity, is the part the paper actually solved to its own benchmark. The systems problem, which is keeping that material within its operating temperature while the aircraft is moving through air and connecting it to the rest of the airframe, is the part the paper explicitly flags as open. Without a published answer to the cooling question, sustained mid-flight laser charging stays a bench result, not an operational capability.

Two other gaps separate the Tianjin test from an in-service system. The drone flew inside a controlled environment under a stationary laser, so the test does not measure what happens when the laser has to track a moving target or when the beam passes through haze, dust, or humidity. The 38.49% figure is a single-paper measurement, not a replicated number across laboratories or laser wavelengths. Independent reporting by The National confirms the headline claim and Han's framing; a Spanish-language La Razón pickup is a translation of the same release, not an additional primary source.

What the next paper has to show is on the systems side. A receiver that holds its efficiency at 80–90°C during flight would need active cooling, a redesigned perovskite cell with a higher thermal ceiling, or a thermoelectric stage that can dump waste heat faster than the cell generates it. None of those have been demonstrated in this paper, and Han's framing suggests they are what the team's next publication will attempt. A 29 July publication date also means the next benchmark is likely to come from the same group rather than from an independent lab in the near term.

The narrow conclusion the evidence supports: light-to-electricity conversion is no longer the gate, and a specific peer-reviewed measurement now says so. The wider claim that drones can recharge in flight is still gated on heat. The 38.49% figure is the start of a thermal-engineering problem, not its solution.