Text from the interactive explanation
Picogram-Scale Interstellar Probes via Bioinspired Engineering
George Church
Original paperDrag from a gram to a picogram and watch the mission change: one chip that flies past, or a swarm that can stay.
Each glint is a probe riding a light sail. White sparks are fatal dust hits: on average two per probe over the whole trip, so about 13% get through. Blue flashes on the planet are a light signal programmed in DNA. Launch energy at a fixed speed scales with mass. Impact energy uses E = mc²(γ − 1) at 0.2c; sail area scales with mass from the paper’s 10−8 m² per picogram.
Today’s boldest plan would fling a gram-sized chip past Alpha Centauri. This paper proposes shrinking the probe to the size of a living cell, launching it by the billion, braking it with starlight, and letting the survivors land, multiply and grow a light that signals home. Follow one swarm from launch to reply.
Twenty-one powers of ten
People have imagined probes that copy themselves since the mathematician John von Neumann described a self-reproducing machine in 1948. In 1978 the physicist Paul Davies even suggested sending “biological machine seeds” at near-light speed. But every engineered design for a self-copying probe came out enormous. Breakthrough Starshot went the other way, tiny and fast, but its chip can’t stop, let alone build anything. Tap the stops on the ladder of probe masses.
Predict For the cost of launching one gram-scale probe, about how many picogram-scale probes could you launch?
Small also means gentle
Starshot’s chip is meant to fly past at 20% of light speed. If something that fast hit a planet, its kinetic energy would be released at once. The paper compares a gram and a nanogram arriving at 0.2c. The bars use a logarithmic scale: each step is ten times more energy.
Gram and nanogram values are from the paper. *The picogram value is computed with the same formula, and the apple comparison is ours. Scale: 1 J to 1013 J.
But even a probe light enough to land softly arrives at thousands of kilometres per second. With no fuel on board, how does it stop?
Braking with starlight
Light carries momentum. When it bounces off a mirror it gives a tiny push. That’s how a laser on Earth can launch a light sail, and it’s also how the destination star can slow one down: as the sail falls toward the star, the starlight pushes back against it. The lighter the sail for its size, the harder that push works.
The paper, citing calculations by Heller & Hippke (2017), puts numbers on it. A sail of graphene , a sheet of carbon one atom thick weighing 0.763 milligrams per square metre, can be captured at Alpha Centauri only if it arrives no faster than 4.6% of light speed . Punching holes in the graphene halves its mass and raises the limit by √2, to about 6.5%. Sirius A, a much brighter star, could stop a sail arriving at 14.9% .
Your mission: get a probe captured
Trip time is distance ÷ arrival speed, ignoring the short launch. The leftover speed of a probe that is too fast uses a simple energy budget (vout² = v² − vlimit²) and is our illustration, not a figure from the paper.
A century in transit, through space that isn’t empty. What happens to a cell-sized probe along the way?
The dust gauntlet
Between the stars drift grains of dust, typically smaller than half a micrometre. At a few percent of light speed, hitting even one is likely fatal for a cell-sized probe. The paper estimates that a probe 8 micrometres wide, crossing the 4 × 1016 metres to Alpha Centauri, would meet about two grains on average .
Predict If the average probe meets two grains and a single hit is fatal, what fraction of the swarm survives?
Fly a swarm of 300 through the dust
About one in seven arrive and are steered onto a planet with liquid water. Now: what can a single cell do there?
One cell, one planet
This is the step no machine can do: a living cell copies itself from simple chemicals around it. Current microbes need carbon, hydrogen, oxygen, nitrogen, phosphorus, sulfur, magnesium, potassium, sodium and calcium, plus traces of a few metals. The paper asks how fast one lander could spread.
Predict One cell lands. It divides in two every 1.4 hours. How long until its descendants could coat the whole planet in a single layer?
Watch a colony take a world
The living toolkit
To build a transmitter, cells must make hard, shiny, precise materials. They already do: organisms grow minerals that human industry makes only at high temperature or pressure, and they pump chosen elements to extreme concentrations. Selected entries adapted from the paper’s table of biominerals (reformatted).
A planet coated with engineered cells. How would anyone on Earth find out they made it?
A lighthouse programmed in DNA
To “phone home”, the colony must make light that outshines its surroundings, at some wavelength, at least for a moment. Fireflies, glowing fungi and ocean plankton already make light with an enzyme called luciferase. The paper’s idea is to make that light spell out a message written in DNA, using the chemistry behind an early DNA-reading method called pyrosequencing .
The message is DNA. A strand called the template stores it in four letters: A, C, G, T.
Offer one letter at a time. Building blocks are supplied in a fixed order, A, C, G, T, over and over. A copying enzyme (polymerase) adds one only if it pairs with the next template letter: A pairs with T, C with G.
Every addition makes light. It releases pyrophosphate (PPi), which two enzymes turn into ATP and then a flash. No match, no flash. A clean-up enzyme (apyrase) clears leftovers before the next offer.
Write a message and watch it flash
A run of identical template letters is filled in one cycle and gives a proportionally brighter flash, a standard feature of pyrosequencing (Ronaghi 2001, cited by the paper). Try AATTTCCCCGGG.
Be the receiver: decode a signal
A telescope on Earth records this pattern of flashes. Letters under the bars show what was offered in each cycle. What was the template?
Making a planet flash in step
One cell’s flash is hopelessly faint from four light-years away. A whole planet’s worth could count, but only if the flashes line up. Fireflies show both styles: in the Smoky Mountains, Photinus carolinus flash in rolling waves, while Southeast Asian Pteroptyx blink almost simultaneously. Marine plankton called dinoflagellates make 475 nm flashes lasting 80–500 ms.
Tune the synchrony
Tap the planet patch to start a trigger flash from that spot. In the paper’s design, a single spontaneous flash from one cell could trigger the rest. Built-in delays and feedback could push waves toward true synchrony.
Living lamp, or living laser?
Planet-scale bioluminescence. Conceptually far simpler and much closer to a demonstration, but more invasive, since it means coating a world. The paper’s summary lists its advantages as a distinct wavelength and “size: ×1014”; its challenges as synchrony, bandwidth and the km² extent it needs.
A biofabricated laser. Gather starlight with seashell-shaped “compound parabolic” collectors lined with biological mirrors such as nacre or beetle reflectors, and use it to drive a laser crystal like ytterbium-doped yttrium oxide. Organisms can gather both metals: Daphnia water fleas take up yttrium, and an E. coli protein binds ytterbium. A narrow laser line could stand out about 10,000× better, and frequency doubling another 10–100×. Advantages listed: small footprint, “focus: ×107”. Challenge: genetically programming shape and orientation to aim a beam across light-years. Some fungi already grow upward-facing cups.
Launch, brake, survive, grow, signal. So how does the whole proposal add up?
The scorecard
The paper closes with its own ledger of what each idea buys and what stands in the way. Tap a card for the detail you explored above.
Cards adapted from the paper’s summary table (reformatted, with explanations added).
What could be tried first
The author stresses that much work remains on the theory, design and testing, and that some of it can happen close to home: on Earth, where graphene sails have already been tested in microgravity, or within our own solar system, for example at Uranus or Neptune. There, deceleration, landing, replication and a bio-built communicator could each be tried.
Living probes, carefully
Sending life on purpose raises planetary-protection questions. The paper points to a 2018 National Academies report urging NASA to bring in a wider range of scientists, and calls for more study of biocontainment: engineered dependence on nonstandard amino acids, or circuits that count cell divisions and then stop.
Back to the question
What if we sent a billion seeds to the nearest star? Here is the swarm again, built from your choices.
—
Shrinking the probe a trillion-fold turns an interstellar spacecraft into something closer to a seed: cheap enough to send by the billion, gentle enough to land, and alive enough to build its own transmitter. This is a proposal, not a mission plan. Its contribution is the reframing, plus a concrete list of problems (sails, dust, braking, growth, synchrony, aiming) that can each be worked on here first.
Source & credits
The biomineral examples and the scorecard are adapted from the paper’s two tables (selected, reformatted and explained).
Numbers quoted from works cited by the paper: Heller & Hippke 2017 (braking limits), Parkin 2018 (Starshot power and cost), Lubin 2016 (Starshot), Gaudenzi et al. 2020 (graphene sail tests), Tabor & Loeb 2021 (detectability), Potvin-Trotter et al. 2016 (ring oscillators), Ronaghi 2001 (pyrosequencing), Ritchie et al. 2018 (photosynthesis on Proxima b), Freitas 1980, Borgue & Hein 2020, Davies 1978, von Neumann & Burks 1966.
All animations are original illustrations drawn with code and are not to scale. The dust and growth scenes compute their numbers from the paper’s figures. The braking trajectories, synchrony model and word-to-DNA code are illustrative. Values marked as computed or ours are not from the paper.
Topics in this iPaper
Engineering
let us consider the challenges of a picogram- to nanogram-scale probe to land, replicate, and produce a communications module based on biominerals at the destination
Abstract
Biology
The communications “device” could be constructed and aimed using engineered organisms
Section 5, Building Communication Devices
Physics
so the bulk of deceleration would rely on photons from stars
Section 4, Acceleration and Deceleration
interstellar travel
For exploring nearby stars, let us consider the challenges of a picogram- to nanogram-scale probe
Abstract
synthetic biology
Synthetic biology is exponentially improving in ability to programming photon outputs coupled to a variety of biosensor inputs
Section 5, Building Communication Devices
solar sail
One design is a highly reflective light sail, traveling a long straight line toward the gravitational well of a destination star
Abstract
bioluminescence
this might be engineered through bioluminescence
Section 5.1, Bioluminescence
astrobiology
Proxima b lies in a “habitable temperate zone” but receives only 3% of the photosynthetically active wavelengths of Earth
Section 8, Temperature and Energy Sources