T+0 · Stage 1 raise
A laser rail to the stars, built one satellite at a time.
Fuel-less sail probes, pushed outward by a chain of solar-powered laser stations instead of one giant laser. Stations go up a few per launch and are dropped off outward from the Sun. Each rung produces measured data — one in the lab, three in orbit.
Stage 1$25M4–5 years3 launches
01 / The problem
Sixty years of proposals. The same wall every time.
Laser sails have been proposed since the 1960s. The physics has been accepted since Forward's 1984 paper. Nobody disputes the equations.
Every design since has asked for one enormous laser or lens — from Forward's 1,000 km lens to Starshot's kilometre-scale array — and gigawatts on day one.
Every attempt died on the same two things: scale and the cheque that had to be written before any data existed.
02 / The concept
The rail.
Instead of one enormous beamer, a chain of stations passes the probe along. Each relay holds focus over only 0.1 AU, so the track can be thirty times longer than a single beamer's — and a longer track is what lets the same power push a kilogram instead of a gram. Power stations sit near Mercury's orbit, at roughly 0.3 AU, where sunlight is about 11 times stronger than at Earth. Relay stations along the outward track keep the beam focused. A probe sails inward on sunlight, whips around the Sun, and is pushed out from behind. The same sail does the braking and the flying: no propellant.
The probe sails inward on sunlight, the same sail, no propellant. Leaving Earth it carries Earth's 30 km/s orbital speed and must shed about 9.5 km/s to drop its perihelion to 0.3 AU. A 1 kg probe with a 22 m sail has a lightness number that depends on the sail's measured solar reflectance and transmission — a Rung 0 measurement — and brakes against its orbit on sunlight alone. Projected: 2–3 months of sailing, then 96 days of fall. The dive is not for speed — 67 km/s is 0.2% of the target. It is how a probe from Earth reaches the head of the rail without propellant.
It rounds the Sun at about 0.3 AU, just inside the power stations' orbit near Mercury. Projected: about 67 km/s at perihelion. At 0.3 AU the sail runs at roughly 350–600 K (projection); silicon nitride is stable well past 1,500 K.
The three power stations behind it fire; the relays along the outbound leg refocus the beam in turn. Each relay is kilometre-scale, but holds focus over only 0.1 AU. Projected: about 9 hours on the rail to 10% of light speed.
Projection On the order of months from launch to the start of the 43-year cruise, pending trajectory optimisation.
03 / Why now
Three things changed.
A Delft team hit a 2 mm tethered silicon-nitride sail with a 230 W laser at 0.3 GW/m² for 4 seconds, in air, and measured about 100 g of acceleration. The material survives the intensity. Vacuum, duration and free-flight are the open questions — and they are Rung 0 and Rung 2. Delft also found the membrane buckles under ~20 °C of uneven heating, and expects vacuum to be worse — the reason Rung 0 measures shape, not just survival.
Sun-synchronous orbit at $350k for up to 50 kg, and $7k per kilogram above that. Laser-communication pointing hardware is commodity. A Stage 1 rung is a smallsat, not a program.
The one funded laser-sail program has been paused since 2024; most of its pledged $100M was never delivered. No one is flying this. Scientific American, September 2025.
Never demonstrated. No published experiment has pushed a free-flying sail with a laser in space, or handed a sail from one station to the next. The rail depends on both. Both are cheap.
04 / Physics in one line
Thrust on a reflective sail is 2P/c.
It scales linearly. The physics scales. What doesn't — pointing, handoff, and how a real sail behaves under a real beam — is exactly what Stage 1 measures.
05 / Stage 1
The ladder.
Four rungs, a bench relay breadboard and a pre-flight gate, each with a numeric pass/fail metric, a cost band, and a timeline. Hardware and launch by category, they sum to $17–28M; $22–36M with a 30% reserve on those bands; program-level staff, operations and integration are not yet estimated. $25M is the planning figure and the first tranche, released rung by rung. Four to five years with Rungs 1 and 2 overlapping, five to seven if strictly sequential. First measured data in the lab for under $1.5M; first flight data at about $2.5–5.5M, about two years in.
A sail under a beam, in vacuum.
1 kW fibre laser, 1070 nm, ~2–3 mm spot on a cm-scale silicon-nitride sail.
- Cost
- $0.5–1.5M
- Timeline
- 6–12 months
Pointing at range, no sail.
Holds a laser on a beacon-equipped passive target at 5–10 km to ≤10 µrad rms per axis. 100 km is a stretch objective nothing depends on.
- Cost
- $2–4M
- Timeline
- 18–24 months
First laser push of a free-flying sail in space.
An ESPA-class smallsat fires a 1 kW fibre laser through a 20 cm aperture at a spin-stabilised 10 g sailcraft 5–10 km away, in 9-minute battery bursts in Earth's shadow.
- Cost
- $8–12M
- Timeline
- 2–3 years
Handoff.
The second station about 5 km along-track, so the first station stays inside its tested 5–10 km range through the handoff. It acquires the sail while the first is still firing, then the first cuts.
- Cost
- $6–9M
- Timeline
- +12–18 months
Stage 1 is three launches: one rideshare slot each for Rungs 1, 2 and 3.
06 / The deliverable
What $25M buys.
Bench, bench relay, pointing, push, and handoff. Each with a numeric metric that was set before the money was spent.
A free-flying sail accelerated by a laser in orbit, measured by ranging and by sensors on the sail; the ranging architecture is chosen before any flight hardware is ordered.
The rail is three solar-powered laser stations feeding a chain of about twenty-seven passive relay optics. Stage 1 tests one emitter handing a sail to another; the relay optic — the operation the rail repeats — is tested on the bench in Stage 1 and in space in Stage 2.
Pointing on a passive target, a sail under a real beam, an emitter handoff, and one passive relay measured on the bench. Stage 2's own questions — a relay in space, beam-riding, a 100 kW-class solar station — are not settled by Stage 1; they are what Stage 1 data lets us propose to test.
What it doesn't.
$25M does not build the rail, and it does not send a probe toward Alpha Centauri. It ends with a sail pushed by one satellite and handed to a second in low Earth orbit — the first two-emitter handoff ever flown — plus the lab, pointing and thrust data. That is the evidence the next round needs.
$25M is the first tranche, released rung by rung; Stage 1 with a 30% reserve is $22–36M, before program-level staff, operations and integration.
$25M does not build the rail. It buys the first flight data a rail proposal needs.
07 / Beyond Stage 1
Where it leads.
Everything in this section is a projection. It is what the physics and the current cost of hardware allow, not a commitment.
Two stages sit between Stage 1 and the rail: a solar-powered 100 kW-class station, then a Mercury-orbit station and the sun-dive probe.
Coast time at 1% of light speed, from the Sun. Reaching the rail adds months.
Coast time at 1% of light speed, from the Sun. Reaching the rail adds months. New Horizons took 9½ years.
A 1 kg flyby of Proxima b returns images at tens to hundreds of km per pixel, a close-range spectrum of its atmosphere, a magnetometer reading, and the star's wind and dust environment. Fast flyby probes to the outer planets and the interstellar boundary are the near-term product.
Stages 2 through 4, the cost ladder, and what the probe returns →
08 / Risks we're not hiding
What could stop this.
Sail limits at gigawatt beams
Delft's result is at 0.3 GW/m², the intensity the rail needs on the sail, for 4 seconds in air on a 2 mm tethered sample. Metre-scale sails under gigawatt beams are a lab problem that is progressing, not a solved one. Rung 0 is where we find out on our own hardware.
Power-station cost at Stage 4
The financing cliff. Stage 1 costs tens of millions; the full rail costs far more. Stage 1 data exists to make the next cheque fundable on evidence, not on a slide.
Orbital shear
Stations at different solar distances orbit at different rates. The rail is not a straight line; it is a network that lines up in launch windows. That is a scheduling constraint we design around, not a physics objection.
Heat
150 GW on the sail means 197–259 GW emitted after relay losses, ~500–650 GW of electrical power and ~300–390 GW of waste heat. Fibre lasers run near room temperature, so the radiators are 385–939 km² of radiating surface at 300–350 K (half that in panel area if both faces radiate) — a lower bound this close to the Sun, and larger than the collectors. Not yet in the cost model.
Relay optics
Kilometre-scale membrane optics with per-relay losses that compound across ~27 stations. Stage 1 tests handoff between emitters; the passive relay is a Stage 2 problem. Across a 0.1 AU gap the light-time is 50 seconds, so the relay must predict where the sail will be; prediction, pointing, the sail's own stabilisation and passive beam-riding are four separate requirements, none yet numbered. Stage 2 tests the passive part at 1.4 ms of delay and injects the 50 s in software.
The probe itself
A 1 kg spacecraft that survives 43 years and returns data at 4.3 light-years has no flown analogue. The sail-as-antenna link is a concept, not a budget.
09 / Standing on prior work
The physics is forty years old. The build plan is not.
A solar laser near Mercury and a 1,000 km lens near Saturn. The physics has stood for forty years; the scale was judged absurd.
Stations in a line, and a four-phase incremental pathway to get there. NASA NIAC Phase I and a $500k Phase II (2013–15); a 450 g platform pushed on an air track in 2015. Funded to the bench, never to flight.
A proposed ground-to-orbit laser-sail demonstration ladder. Designed, never flown.
One 100 GW ground array. It funded lab work — Caltech's sail measurements acknowledge it (Source) — but most of the pledged $100M was never delivered and nothing flew.
A NASA-side beamed-energy demonstration study in the same period. Designed, never flown.
One modular phased array that grows by adding elements: "not all or nothing", starting with ground and CubeSat tests. The incremental-build idea in print before ours.
A CubeSat pushing a credit-card sail on a chipsat. Sought seed funding; never flew.
Solar-pumped stations firing in succession, 10 g probe to 7.7% of light speed.
What is different here: a few solar power stations feeding a chain of passive relays, rather than a laser at every station or one giant array; a first flight rung priced at a rideshare slot; and the sun-dive to the head of the rail. The incremental build is Lubin's, Bae's and the Caltech KISS study's idea; the chain of successive stations with handoff is van den Donker's. We are standing on them, not beside them.
10 / Team
Who is building it.
Mike Taylor
Mike Taylor is a repeat founder with several exits behind him. Heliorail is the first thing he's built that leaves the atmosphere.
11 / Stage 1 · $25M
Partner with us.
We are raising $25M for Stage 1 from investors and philanthropic funders who want to see the first laser push of a sail in space, and the first handoff, measured and published. First flight data at about $2.5–5.5M.