In 1981, while Lockheed's Skunk Works was deep into the classified Have Blue program that would become the F-117 Nighthawk, a small team at Messerschmitt-Bölkow-Blohm (MBB) in Ottobrunn, Bavaria, was independently solving the exact same problem — and getting there by reading the same Soviet paper Lockheed had used.

The team was led by Dr. Gerhard Löbert, an aerodynamicist who had read Pyotr Ufimtsev's 1962 paper "Method of Edge Waves in the Physical Theory of Diffraction." Ufimtsev had published the equations that let you predict the radar cross-section of an object built from flat triangular facets. Lockheed's Denys Overholser found it in 1974. Löbert found it in 1975. Neither knew the other was reading it.

By 1981 MBB had launched Lampyridae ("firefly" in Latin) as a classified West German Defence Ministry study. The aircraft was a deeply faceted, diamond-planform stealth interceptor — not a strike bomber like the F-117, but a slow-flying point-defence fighter intended to ambush Warsaw Pact ground-attack aircraft from above. Three subscale demonstrators were built. Static radar cross-section testing in MBB's anechoic chamber from 1983–1985 confirmed signature reductions that German engineers later described as "two to three orders of magnitude" below conventional fighters. Free-flying 1/4-scale models were tested at the Manching flight test centre between 1985 and 1987.

The aerodynamics worked. The stealth worked. Then, in late 1987, the program was abruptly shelved.

The official German position has always been vague — "cost," "shifting priorities," "end of Cold War." The version told by Löbert himself, and corroborated by multiple MBB engineers in interviews after declassification in the late 1990s, is sharper: the United States learned of the program, realised Germany had independently developed faceted stealth, and made it diplomatically clear that NATO did not need a second stealth aircraft program. The Pentagon had spent billions to keep faceted-stealth math a state secret. A West German firefly buzzing around Bavaria with the same RCS as a B-2 demonstrator was an intelligence headache nobody wanted. Funding evaporated within months of the conversation.

Why it matters in 2026. Faceted stealth was abandoned almost everywhere by 1990 because curved-surface stealth (B-2, F-22, F-35) gave better signature reduction across more bands once compute power could solve Maxwell's equations on curves. But faceted designs have one decisive advantage: they are cheap to manufacture and cheap to model. Every facet is a flat plate. Every edge is a straight line. There are no compound-curvature composite layups, no thousands of hand-fitted radar-absorbent tiles, no $44,000-per-hour maintenance bills like the B-2.

This is exactly what the emerging class of attritable stealth drones needs. Programs like the USAF's Collaborative Combat Aircraft, Anduril's Fury, and General Atomics' Gambit are racing toward $20 million stealth UCAVs designed to be lost. Faceted geometry — Löbert's geometry — is suddenly the right answer again, because additive manufacturing and CNC sheet-folding can produce a faceted airframe in days. Modern radar-absorbent paints handle the residual edge diffraction Ufimtsev couldn't kill in 1962. The Lampyridae was a slow, subsonic, point-defence aircraft built to be cheap and stealthy. That is the 2026 attritable-drone spec sheet, written in 1981.

The hardware sits in storage at the Deutsches Museum's annex in Oberschleißheim. Nobody is asking to look at it.