Main Sequence — Early Access launch guide: first station, research and the 60-a-minute problem

Main Sequence is a first-person space factory builder: you fly a mining ship, walk around inside it, mine asteroids, and then turn the ore into an industrial base by bolting station modules together and running conveyor belts through the middle of them. It launched into Early Access on 12 November 2026 after three years of solo development, a public playtest and a demo that pulled 200,000 wishlists.

This page is a progression and systems guide, not a played-through walkthrough. Everything numeric here comes from the Steam store page, the developer's own announcements, and a fan database extracted from the demo build 26097974 (2 November 2026) — the last build published before launch. The last section names exactly what is verified, what is inferred and what is still missing.

The short version

The genre blend matters for expectations: it is not a ship-combat game that happens to have factories, and not a Factorio-in-space that happens to have a cockpit. The combat layer is thin and the factory layer is the game — but the factory is gated behind ship design, because your station is built from the same modular blocks as your ship.

What the loop actually is

Six systems, in the order they gate each other:

1. Mining. Two routes: your ship's Mining Laser plus tractor beam, or drone bays on a station. Station mining is range-limited, so station placement is a mining decision.
2. Refining. Ore becomes plates on your ship at the Workstation (hand-fed, cannot be automated), and on the station at a Refinery once you have one.
3. Station construction. Modules snap together inside a station the way ship modules snap together inside a hull. The Command Module is what lets you edit the station; the Shipyard Core is what lets you edit a ship.
4. Automation. Machines, conveyor belts, splitters and containers on a factory floor.
5. Research. Manufactured items are fed into Science Labs and converted into research points, which unlock recipes, machines and modules. This is the whole progression spine.
6. Defence. Alien relic forces attack what you build. Weapons are modules, so every gun is ship or station mass you had to design in.

The first hour, in order

The opening is a tutorial chain with explicit objectives, and the very first one is a placement instruction worth reading twice: build your starter station within 1 km of a large titanium asteroid. Mining drone bays reach 1,000 m, so a station parked at 1,100 m from the asteroid is a station that collects nothing.

1. Scan the belt before you commit. Scanning tells you which asteroids carry what. You need titanium first — everything in the opening build costs titanium plates indirectly, through Hull Plate and Lightweight Frames.
2. Mine with the laser, pull with the tractor beam. Secondary fire on the tractor beam pulls a mined chunk into your inventory. The Mining Laser is 125 m range, 10 damage per second against rock, 3 MW of ship power.
3. Refine on the ship first. The Workstation module gives you a refinery and an assembler in one block (10.5 MW, 3 Hull Plate). It cannot be automated — that is the point. It is how the first Hull Plate and the first Lightweight Frames exist at all.
4. Anchor the station. Command Module: 5 Hull Plate + 10 Lightweight Frames. It is 30 hull, 10 mass, and it is the block that opens the station designer.
5. Add a Dock (5 Hull Plate) so ships can connect and, once configured, unload automatically. Ships need a Docking Collar (3 Hull Plate) on their side of the connection.
6. Build the Reactor before the machines. 5 Lightweight Frames + 10 Magnetic Coils, and it produces 48 MW from water ice or hydrogen fuel. Ice is on the belt; a Refinery turns 2 Water Ice into 1 Hydrogen Fuel every 6 s, which is 10 per minute.
7. Workstation, then Refinery, then Assembler. Refinery: 2 Lightweight Frames + 5 Magnetic Coils, 4.5 MW. Assembler: 3 Lightweight Frames + 2 Capacitor + 2 Basic Motor, 6 MW.
8. Material Science Lab (2 Lightweight Frames + 2 Capacitor, 12 MW) and feed it Hull Plate — one Hull Plate is one Material Science point. That single lab is the entire tier 1 and tier 2 tech tree.
9. Container (2 Lightweight Frames, 16 slots) as the buffer between the dock and the machines.
10. Then the Shipyard (15 Material Science) so you can edit the ship you are flying instead of building a second one from scratch.

Your first station, module by module

Station blocks are the shape of the base, and they are not cheap relative to the first ten minutes of production. Costs below are from the extracted demo data.

Two practical notes. Every factory floor module is a hull purchase first and a machine purchase second, so plan the floor before you buy the machines — a Refinery with nowhere to sit is 4.5 MW of nothing. And Station 2×2 C is locked in the demo data, so a twelve-module station in the demo is really a patchwork of 1×1 and 2×1 blocks.

Power: the MW budget

Power is distributed automatically across a station, and the failure mode is a warning that emergency power has reduced machine speed — not a blackout. That makes under-powering quiet and expensive: everything keeps running, slower, and your ore-to-plate ratios silently change.

The single most useful arithmetic on this page: one Refinery, one Assembler and one Material Science Lab is 22.5 MW, which one Reactor carries with room for a second assembler. Add an Enrichment Chamber (18 MW) and you are at 40.5 MW — still inside one Reactor, but with nothing left for a second lab. The moment you want Material and Computation science running together you are at 52.5 MW and need a second Reactor or four Solar Panels Mk. 1.

The ore ladder

Everything descends from six raw resources plus three late ones. The pattern is: ore → plates in a Refinery, plates → components in an Assembler, components → science in a Lab.

The enriched-ore line is the upgrade that changes the ratios. An Enrichment Chamber takes 3 ore → 2 enriched ore in 3 s (40/min) and a Refinery then turns 1 enriched ore → 3 plates in 6 s (30 plates/min). So enrichment is a 2× output multiplier on the same ore, at the cost of 18 MW and a second stage.

Titanium is the early bottleneck because it is three things at once: Hull Plate (2 Titanium Plates → 1 in 4 s), Lightweight Frames (1 Titanium Plates → 1 in 2 s), and Basic Motor (1 Titanium Plates → 2 in 2 s). A single Assembler running Hull Plate eats 30 titanium plates per minute — that is 1.5 Refineries on raw ore before you have built anything else.

The research economy: seven sciences

Research is not a currency you earn; it is a set of seven parallel streams, each fed by a different manufactured item into a different lab. This is the part that surprises people coming from Factorio: your science is your production line, and a lab with no input is a 12 MW paperweight.

Two things follow from that table. First, the later sciences are deliberately slow: Quantum gives 25 points per minute at best, Astronomic gives 6, and both sit behind long chains. Second, Relic Technology is a parallel economy — you do not mine it, you synthesise it: a Relic Synthesizer converts 1 Superalloy into 1 Relic Shard every 6 s, and the tier 3 Relic nodes cost 200–250 Relic Technology each.

Lab build costs, since they gate the mapping above:

Tiers 0 to 6: what the tree actually costs

The tree is 97 technologies across seven tiers (81 of them costed in the extracted data). Tier 0 is free or already unlocked; each tier adds another science stream to the bill.

Read that table as a warning about breadth. Tier 2 is where players tend to stall, because it is the first tier that wants two streams at once and the Computation Science Lab needs Basic Circuits — which needs Silicon Wafers and Magnetic Coils — which needs nickel and silica lines you probably have not built. Budget one full production line per science stream before you start spending points on it.

Technology can be queued in the Research Queue, so the correct play once a second lab exists is to keep both queues full rather than letting either lab idle waiting for you to decide.

The 60-a-minute problem

Conveyor Belt Mk. 1 moves 60 items per minute. Splitter Mk. 1 splits one 60/min lane into three and Belt Balancer Mk. 1 balances two lanes at 60/min. Underground Belt Mk. 1 is 60/min with an 8-block maximum run. Mk. 2 belts are 120/min.

That number is lower than most of your machines want. Hull Plate needs 30 titanium plates/min in and puts 15 out; a mid-game Assembler on Superconductor Coil wants 15 Maglev Coupling, 15 Containment Coil and 60 Coolant per minute — three belts for one machine. Detector Array wants 20 Graphene and 20 Cobalt Plates per minute on top of 10 Basic Circuits. The Advanced Assembler's Advanced Circuits recipe is worse: 3 Detector Array, 3 Basic Circuits, 1 Superalloy and 6 Coolant every 12 s is 15 + 15 + 5 + 30 items per minute across four inputs.

Practical consequences:

• Feed each machine from its own container, not a shared trunk. A Container is 16 slots and a Large Container is 32 slots with +500% stack size; Smart Containers research adds per-item filters, which is the upgrade that turns a buffer into a traffic manager.
• Split the coolant line early. Coolant is one Enrichment Chamber recipe (2 Nitrogen Gas → 1 Coolant, 30/min) and it is an input to Superconductor Coil, Advanced Circuits and Missiles. It is the most oversubscribed item in the game's data.
• Count refineries per assembler. 20 plates/min per Refinery on raw ore vs 30 plates/min per Refinery on enriched ore: enrichment buys you 1.5 refineries' worth of output and costs one 18 MW machine. It is almost always the right call before you add a fourth refinery.
• Do not trust the repair-kit numbers as a ratio. The extracted data lists Electrical Repair Kit at 400 Magnetic Coils and 400 Capacitors consumed per minute, which no belt in the dataset can deliver. Either those recipes are hand-crafting-shaped, or the per-minute column is derived from a batch size the belts never see. Build one and time it before you design a line around it.

Ships: mass, thrusters and what the design layer rewards

Ships and stations are built from the same designer: the Command Module edits a station, the Shipyard Core edits a docked ship, and the main-menu designer is creative mode with saved blueprints. The two constraints are mass and power.

Thrusters are additive — "build more thrusters to increase the max speed of a ship" — so a heavy hauler is a mass-vs-MW problem, and a Small Reactor (24 MW, 5 Hull Plate + 20 Magnetic Coils) is the ship-side answer. The 0.1.80 balance pass trimmed exactly this: Ship Frame 3 → 2 t, Ship Corner Window 5 → 3 t, Small Habitat Window 5 → 3 t. Windows are cheap mass; frames are cheaper.

Two modules worth designing around early:

• Workstation (ship) — refinery and assembler in one block, 10.5 MW, cannot be automated. It is how you bootstrap away from the station.
• Warp Drive — 20 Hull Plate + 10 Containment Coil, one per ship, consumes Warp Fuel. It is not the same block as the Interstellar Warp Drive.

Defence: what actually threatens a station

Hostile relic forces are in the starting system from the beginning, and the developer's playtest notes are blunt that the Relic Destroyer sits at the edge of the solar system as a hard gate rather than a random encounter. Ship-side weapons from the extracted data:

Station defence is Plasma Turrets (built on stations, auto-targeting) plus armour blocks. The honest read on the numbers: a 10-point shield recharging at 0.5 per second is a soak, not a wall, so the defence that matters early is not being found — a station further from the relic patrol routes survives longer than one with a second turret.

The four progression gates

The developer's playtest briefing lays out the campaign shape explicitly, and it is four gates, not a level list:

1. Titanium and Material Science. Get a Reactor, machine floor and Material Science Lab running on Hull Plate. Everything up to tier 2 is paid for out of this one item.
2. Slipstream Drive (200 Material + 200 Computation, needs Shipyard and Enrichment Chamber). This is the moment the game opens the map: it unlocks the Warp Fuel recipe, and Warp Fuel is 2 Deuterium Fuel + 1 Containment Coil in an Enrichment Chamber (5/min). Deuterium comes from Hydrogen Fuel, Hydrogen from Water Ice. The chain is three machines deep, so build it before you research the drive, not after.
3. Cobalt, graphite and the Electromagnetic / Quantum tier. Other sectors of the solar system hold the cobalt and graphite you cannot get at home, plus abandoned stations to loot and a Trading Outpost that sells blueprints and modules for credits. This is also where the Quantum Science Lab becomes worth its 10 Superconductor Coil build cost — at 5 points per item it is the fastest science in the game once fed.
4. The Relic Destroyer, then the Interstellar Warp Drive (200 × Material, Computation, Electromagnetic, Quantum). Kill the Destroyer to unlock the interstellar drive, leave the system, and the last two sciences — Astronomic (Encoded Simulation Matrix, which needs Lonsdaleite plus a Supercomputer) and Xenobiologic (Biolattice, which needs three alien microbe cultures and a Growth Chamber) — unlock the endgame modules and personal equipment.

Multiplayer, peace mode and creative mode

Co-op is live in Early Access after an experimental playtest round. Same rules, shared scale — the developer frames it as a way to expand the industrial operation rather than a separate mode. Peaceful mode removes the combat pressure, and the stated direction of Early Access is more of these dials (combat, research and automation requirement scaling) rather than a single difficulty setting. Creative mode is the main-menu designer: free building, save and share blueprints, no progression.

How this was made, and what is still unverified

This page was not played through. It is compiled from the Steam store page, the developer's own update posts (launch announcement, price announcement, Next Fest demo notes, the September playtest briefing and the playtest patch notes up to 0.1.80), and a fan database whose numbers were extracted from the demo build 26097974 on 2 November 2026. Treat every figure as demo-build, not launch-build. Costs and rates that moved between the demo and launch are exactly the ones this page cannot catch for you.

What is still unknown, in order of how much it will cost you:

1. Science Lab consumption rate. Points per item are published; items per minute per lab are not. The production rates in the science table are what the upstream machine can make, which is the ceiling, not the throughput.
2. Ore distribution. Which bodies hold cobalt, graphite, iridium and lonsdaleite is not in any source we found — the developer only says other sectors of the solar system, then other star systems.
3. Enemy numbers. The Relic Destroyer's health, the wave timing and the damage model are not published. Weapon stats above are yours, not theirs.
4. Conveyor Belt Mk. 3. The extracted record lists it at 120/min, identical to Mk. 2, while the Underground Belt Mk. 3 is listed at 180/min. That looks like a data artefact — do not design a ratio on it until you have built one.
5. The mission chain beyond the tutorial. Objectives exist (the demo surfaced named ones for Station 2×2 C and Computation Science) but the full list is not exposed anywhere.
6. Balance drift. Playtest 2 Patch 3 and 4 already moved build costs (Container 5 Titanium Plates → 2 Lightweight Frames; Station Plasma Turret 10 Lightweight Frames → 10 Capacitors) and module mass. Expect the same in the launch window.

If you find a number that is wrong against the live build, say so in the Steam hub thread — the page gets corrected with a dated entry rather than quietly edited.