Stellar Engines — Moving Stars

A civilization that can move its star does not need an ark ship. It takes the planets, comets, metals, habitats, and biosphere with it. Stellar engines are hypothetical megastructures that use a star’s own radiation or mass as propulsion, turning a solar system into the spacecraft.

The numbers are ugly in a useful way. A Sun-mass object is about 2 × 10³⁰ kg; even tiny accelerations need absurd power, but stars already produce it. The Sun emits about 3.8 × 10²⁶ watts every second. The trick is not making energy. The trick is pushing in one direction for a million to a billion years without wrecking the system you are trying to save.

How it works

The simplest design is the Shkadov thruster, proposed by Leonid Shkadov in 1987. Put a huge mirror on one side of a star. Radiation pressure pushes the mirror outward; gravity pulls it inward. At the right distance the mirror becomes a statite, hovering without orbiting.

Because half the star’s light is reflected back while the other half escapes, the star-mirror system emits more momentum one way than the other. The star recoils. For a Sun-like star, the acceleration is around 10⁻¹³ m/s². That is pathetic for a rocket and meaningful over geology: after 1 billion years, the velocity could reach hundreds of km/s.

Matthew Caplan’s 2019 design is less patient. It uses focused stellar energy to lift material from the star, separates hydrogen and helium, fuses part of the stream, and ejects plasma as a directed jet. The claimed acceleration is around 10⁻⁹ m/s², about 10,000 times a Shkadov-style push. Caplan estimates a Sun-like star could be moved roughly 10 parsecs in about 1 million years while consuming about 10¹² kg/s of stellar material.

A 2024 proposal by Clément Vidal pushes the idea into binary systems. “Spider” pulsars are real: millisecond pulsars that irradiate and strip material from a low-mass companion. Vidal asks whether such systems could be steered, using ablated companion material as propellant. The speculative SETI version is the “stellivore” hypothesis: some odd accreting binaries might not be natural feeding systems, but engineered ones.

What would give it away

Stellar engines are attractive SETI targets because they do not require anyone to say hello. They distort stars.

A Shkadov mirror could create asymmetric transit light curves if seen edge-on. It should also produce directional infrared excess, like a partial concept dyson sphere with a preferred axis. A Caplan-style thruster would be messier: anomalous stellar wind, strange spectral lines, or mass loss that does not fit the star’s age and type.

The best current map for this kind of search is Gaia. Data Release 3, published in 2022, contains positions, parallaxes, and proper motions for more than 1.8 billion sources. If a star is being pushed hard enough, its motion might look slightly wrong against galactic dynamics. The hard part is separating engineering from bad data, binary companions, past encounters, and ordinary stellar weirdness.

Why move a star at all

The obvious reason is evacuation. A nearby supernova within tens of light-years could damage atmospheres and biospheres. A close stellar encounter can disturb Oort-cloud comets. A Sun-like star has about 5 billion years before red-giant expansion makes Earth-like orbits unlivable.

But the cleaner reason is logistics. Interstellar ships move payloads. Stellar engines move supply chains: the star, planets, asteroids, volatiles, habitats, and power source. A star system becomes a mobile city-state measured in astronomical units.

That makes stellar engines a sharper concept fermi paradox question than radio beacons. If advanced civilizations exist and prefer survival over messaging, they may leave motion, heat, and missing-light signatures rather than speeches.

What's contested

No one has observed a stellar engine. The physics does not violate conservation laws, but the engineering assumptions are doing most of the work: mirror stability, material strength, stellar mass handling, beam control, and survival of planetary orbits under long acceleration.

The SETI claim is also fragile. Stars already misbehave. Binaries transfer mass, pulsars ablate companions, dust creates infrared excess, and catalog errors can fake anomalous motion. A credible detection would need multiple signatures pointing the same way: motion, spectra, heat, geometry, and no ordinary astrophysical explanation.

Why this has to do with other realms

A stellar engine is propulsion, but it is also political philosophy at astronomical scale. A civilization that can move its home star has chosen continuity over expansion by colony ships. That choice rhymes with concept longtermism: spend impossible capital now to keep options alive for descendants no one will meet.

It also touches biology. Earth life evolved under one star, one magnetosphere, one comet history, one day-night rhythm. Moving the Sun would not just move rocks; it would move an evolutionary laboratory. The page next to this is not only tech fusion drive, but concept evolutionary bottleneck.

An open question

If Gaia can see 1.8 billion sources, what would count as the first serious candidate: a star moving wrong, a star shining wrong, or a binary system eating itself too neatly?

Key Sources

Leonid Shkadov, “Possibility of controlling solar system motion in the galaxy” (1987) — original Shkadov thruster proposal.
Matthew E. Caplan, “Stellar engines: Design considerations for maximizing acceleration” (Acta Astronautica, 2019) — active stellar-engine design using stellar mass as propellant.
Clément Vidal, arXiv:2411.05038 (2024) — spider stellar engine and stellivore technosignature proposal.
Gaia Collaboration, Data Release 3 (2022) — astrometric catalog for large-scale proper-motion anomaly searches.
to verify: arXiv:1306.1672 — cited in prior draft for transit signatures of Shkadov-like artificial structures.

Abhishek