Endfield Factory Guide — Build an Industrial Chain That Runs Itself

Why the Factory is the system that defines Endfield

If you came from Arknights mainline, the Factory is the part of Endfield with no real precedent. Mainline gave you base management — a static set of rooms you assign operators to and check once a day. The Factory is a live industrial chain: extraction nodes pull raw resources out of the map, belts move them, machines refine them, and the whole thing keeps producing while you're offline. That offline output is the point. It's the difference between a daily routine that takes ten minutes and one that takes forty.

The mistake nearly everyone makes is treating the Factory like a side feature you'll "get to later." It isn't. By the back half of patch 1.0, the Factory is producing the majority of your weekly materials — the iron for character ascension, the chemical fluids for weapon tuning, the rare resources you'd otherwise be grinding stamina for. A player with a stable factory and a player without one are running two different games by week three.

One honest caveat up front: this is written against the CN test server and what we know of Arknights mainline's economy design. The throughput numbers below are planning estimates, not verified production figures. Hatsune-precise belt rates and exact node yields will shift before global launch. What won't shift is the structure — extract, buffer, refine, deliver — and that's what this guide is really about.

When to start building (hint: not yet)

Do not place a single belt before you unlock the Industrial Blueprint. As the beginner guide covers, that unlock comes around main story chapter 3-4. Anything you build before it is throwaway — you'll be working with placeholder tools, limited node access, and a tiny build budget, and you'll end up tearing it all down.

Even after the unlock, hold back. The first version of your factory should be deliberately small:

• Chapter 3-4 territory: build only enough to learn the interface. One short iron line, end to end. Treat it as a tutorial you're paying for in materials.
• Don't expand to fill the map. Early node placement is constrained, and you don't yet know which resources you'll bottleneck on.
• Resist the urge to "optimize." You have no idea what your real demand curve looks like until you've been ascending characters for a week. Build for learning first, output second.

The reason for the patience is on the other end of the timeline: chapter 5 introduces advanced extraction nodes that change everything about how you'd lay out a factory. We'll come back to that — but plan now knowing the early build is temporary.

The one principle: stability beats throughput

This is the rule that separates a factory that quietly funds your account from one that needs babysitting. A line that runs 22 of 24 hours without jamming beats a line that hits peak efficiency for six hours and then chokes. Peak throughput is a vanity number. The number that matters is output-while-you-sleep.

Stability comes from one thing more than any other: buffer chests. A buffer chest sits between a producer and a consumer and absorbs mismatches in timing. When your extraction node briefly outpaces your refiner, the buffer fills instead of backing up the belt. When the refiner briefly outpaces the node, it draws down the buffer instead of starving. Without buffers, every tiny rate mismatch propagates down the whole line and eventually locks it.

The temptation is to spend your build budget on faster belts and more machines. Resist it. A slower line with generous buffering will out-produce a faster line that jams at hour six, because the slow line is still running when you log in the next morning. Build wide margins, not tight ones.

Your first three pipelines

Once you've cleared chapter 4 and have a real build budget, set up three independent lines. Keep them separate — don't try to feed three outputs off one shared spine. Independent lines fail independently; a shared spine fails all at once.

1. The iron pipeline (your bread and butter)

Iron ingots are the highest-volume material you'll need, mostly for character ascension. This line should be your largest and most over-buffered.

1. Extraction node on the richest iron deposit you've scouted.
2. Belt straight into a buffer chest — don't connect the node directly to the smelter.
3. Smelter draws from the buffer, outputs ingots into a second buffer.
4. Output buffer feeds your central storage / Factory output.

As a planning target, aim for a passive iron line in the ballpark of ~50 ingots/hour for your first stable build. Treat that as an order-of-magnitude goal, not a promise — the real rate depends on node richness and machine tier, both of which are still moving on the test server.

2. The rare-resource pipeline (low volume, high value)

Rare resources — the materials gating module upgrades and higher ascension tiers — come out slowly and you'll never have enough. This line is small but you want it running 24/7 because its output is the genuine bottleneck on your account's ceiling.

• Single node, single refiner, heavily buffered. A target of ~5-10 units/hour is realistic for an early build.
• Because volume is low, the buffer on this line doubles as long-term storage. Size it generously so it never overflows and stalls the node while you're away for a day.
• This is the line you'll prioritize first when you get advanced nodes in chapter 5.

3. The chemical pipeline (feeds weapon tuning)

Chemical processing produces the weapon tuning fluids that the combat layer eats constantly. As the combat guide explains, tuned weapons out-damage raw character levels, so this line directly drives your damage ceiling.

• Chemical lines usually have two-stage refining — a raw extract, then a conversion into usable fluid. That means two buffer points, one after each stage. This is the line most likely to jam if you under-buffer, precisely because of the extra stage.
• Match your fluid output to roughly how fast you're actually tuning weapons. Overproducing fluid you can't use just backs up the line and wastes node uptime that could be on rare resources instead.

Why your factory jams — and how to clear it

A stalled factory is almost always one of four problems. Diagnose in this order:

Symptom	Likely cause	Fix
Belt full, machine idle	Output backed up — downstream consumer can't keep pace or storage is full	Add a buffer downstream, or raise consumer throughput
Machine idle, belt empty	Starved — upstream node or belt can't supply fast enough	Add a node, widen the belt, or accept the lower rate
Line stalls only overnight	Buffer too small — fine while you watch, overflows while away	Bigger buffer chests; size for 12+ hours unattended
Wrong material in the buffer	Node mis-assigned — pulling the wrong deposit	Re-check node assignment; rare-resource nodes are easy to misclick

The single most common cause is the third row: insufficient buffering. The line looks perfect when you're staring at it, then you log in twelve hours later and it's been stalled for ten of them. If a line ever works in front of you but is jammed when you return, the answer is almost never "faster belts" — it's "bigger buffers."

The second most common is a node mis-match on a multi-line layout: a belt routed to the wrong refiner, or a node assigned to a deposit that doesn't feed the chain you think it does. Walk the line end to end before you start rebuilding machines — the break is usually a routing error, not a capacity problem.

The chapter 5 reset: what to tear down, what to keep

Chapter 5's advanced extraction nodes are a genuine power jump — substantially higher yield per node than the early ones. The catch is that they change the math of your whole layout, and your chapter-4 factory will feel cramped and slow the moment you have access. This is the planned obsolescence the beginner guide warns about.

The trap is the demolish penalty. Tearing down a built structure refunds only about 50% of its materials (a figure consistent with mainline-style sink design — confirm at launch). Rip out your whole factory in a panic and you've thrown away half your build investment. So plan the upgrade on paper before you touch anything in-game:

• Keep: belts and buffer chests on lines whose routes aren't changing. Layout is layout; if the path is still good, the advanced node drops into the same spot.
• Keep: your rare-resource line's footprint. Upgrade the node in place; it's your highest-value, lowest-volume line and rarely needs re-routing.
• Tear down: any iron line you built tight against an early node's low ceiling. Advanced nodes will starve a belt network that was sized for the old throughput — that one's worth the 50% hit to rebuild wider.
• Tear down last, build first: where possible, stand up the new line beside the old one, confirm it runs, then demolish the old one. You eat less downtime and you don't strand yourself with zero production mid-upgrade.

Do the math with a sketch before spending a single demolish. Half your materials are riding on getting the plan right the first time.

Advanced: point your factory at what you're actually leveling

A mature factory shouldn't produce a flat mix of everything. It should be pointed at your current progression loop. If you're ascending Perlica this week and tuning her weapon, your factory should be biased toward the iron and the specific chemical fluids that feed that work — not churning out materials for a character you won't touch for a month.

Practically:

• Read your demand off your goals, not off the wiki. Before you set the week's chains, look at what you're actually upgrading and produce for that.
• Idle a line you don't need. If you're not tuning weapons this week, throttle the chemical line and put that node uptime into rare resources — the one thing you're always short on.
• Re-point after every banner. A new character means a new demand profile. The first thing to do after pulling is decide what your factory should be feeding to get them combat-ready.

This is the same lesson the resource planner applies to stamina: a fixed routine wastes output on things you don't need yet. The factory is just a stamina-free version of the same idea — so steer it.

The minimum-maintenance daily checklist

Once your three lines are stable, the entire daily factory routine should fit in under a minute:

• Claim Factory output first thing — empties storage so nothing backs up.
• Glance at each line for a stall. Belt full + machine idle, or machine idle + belt empty, means a jam. Everything moving means you're fine.
• Check buffer levels. A buffer pinned at 100% means downstream can't keep up; a buffer at 0% means upstream is starving. A healthy line sits somewhere in the middle.
• Re-point if your goals changed (new character, new weapon to tune).
• Set the next chain last, so it runs the full day while you're offline.

That's it. A well-built factory is one you barely look at — which is exactly the point.

FAQ

How long until the Factory "pays back" what I sink into it? Roughly a couple of weeks of passive output to cover its build cost, going by mainline-style economy pacing — then everything after is pure profit. The real return isn't materials, though; it's the daily time you stop spending grinding stamina for things the factory makes while you sleep. Treat the exact payback window as an estimate that'll firm up at launch.

What chapter do I unlock it? The Industrial Blueprint unlocks around main story chapter 3-4, and chapter 5 brings the advanced extraction nodes that make a serious build worth it. Don't go big before chapter 5 — you'll just be rebuilding.

How much do I lose if I tear it all down? Demolishing refunds only about half the materials (subject to launch tuning), so a full panic-rebuild throws away roughly 50% of your investment. Plan the upgrade on paper, keep what doesn't need re-routing, and build new lines beside old ones before demolishing.

How do I maximize offline output? Buffer everything, don't overbuild. The factory only produces for the hours it runs unjammed, so the highest offline output comes from the most boring, most over-buffered layout — not the fastest one. Size your buffers for at least 12 hours unattended and you'll wake up to full storage instead of a stalled line.