Lithium-ion battery manufacturing runs in three stages, electrode manufacturing, cell assembly, and cell finishing, turning powders and foils into charged cells. Formation and aging alone can take one to three weeks, and because dozens of steps each carry a defect risk, overall yield compounds downward, making yield and traceability the decisive metrics.

A gigafactory looks like an assembly plant but behaves like a chemistry plant. The product is not fully "made" until it has been charged, rested, and proven electrochemically stable, and a mistake at any of dozens of steps can quietly poison a cell that only fails weeks later. This guide walks the three stages, explains why formation and aging dominate the timeline and the cost, and shows why yield, not raw throughput, is the number that decides whether a battery plant makes money. It is an educational companion to broader process guides like plastics manufacturing and traceability in manufacturing.

What are the stages of battery cell manufacturing?

There are three: electrode manufacturing, cell assembly, and cell finishing. Electrode manufacturing makes the coated foils that store and release charge. Cell assembly combines those electrodes with a separator and electrolyte into a sealed cell. Cell finishing wakes the cell up, charging it for the first time, resting it, and testing it, before it can ship. The three stages are physically and culturally different: the front is a coating and chemistry operation, the middle is precision assembly, and the back is a slow, capital-intensive electrical process.

Three stages of cell manufacturingOne cell, three very different factoriesELECTRODE MFGmix slurry - coat foildry - calendercoating + chemistryCELL ASSEMBLYwind/stack - weld tabsenclose - fill - sealdry room, precisionCELL FINISHINGformation - agingtest - gradeslowest, most capital
The front is coating and chemistry, the middle is dry-room precision assembly, and the back is a slow electrical process. Each has its own failure modes.

How are battery electrodes made?

Electrode manufacturing has four steps: mixing, coating, drying, and calendering. First, active material, conductive additive, and binder are blended with a solvent into a uniform slurry with tightly controlled viscosity and particle dispersion, a bad mix cannot be fixed downstream. The slurry is then coated onto metal foil, usually with a slot-die that lays down a precise, even film; coating uniformity directly sets how consistent the finished cells will be. The coated foil passes through a drying oven that removes the solvent and leaves a solid electrode layer.

Finally, calendering presses the dried electrode between rollers to compact it. Reducing the porosity of the coating raises its density and electronic conductivity, which improves the cell's energy and power, but the window is narrow. Under-compact and you leave performance on the table; over-compact and you can damage the coating or hurt the electrode's ability to absorb electrolyte later. Calendering is a good example of the whole battery problem: a physically simple step with a tight process window where being slightly off quietly costs yield two stages downstream.

What happens in cell assembly?

Cell assembly turns coated electrodes into a sealed, filled cell, and it happens in a dry room because lithium chemistry reacts with moisture. The electrodes and separator are wound into a jelly roll or stacked in layers, depending on the cell format. The current-carrying tabs are welded, the stack is inserted into its housing, cylindrical, prismatic, or pouch, and the cell is filled with liquid electrolyte and sealed. Every one of these steps is a precision operation where a particle of metal, a weld defect, or a trace of moisture can create a latent flaw.

The defining constraint here is contamination control. Moisture, airborne particles, and metal fines are all cell-killers, so assembly runs under low humidity and tight particle control, and the process fights contamination at every handoff. This is why a battery plant cannot simply be run faster to make more cells: pushing throughput without holding contamination and process control just converts speed into scrap that may not reveal itself until formation.

Why do formation and aging dominate the timeline?

Because they are governed by chemistry that cannot be rushed. Formation is the cell's first, carefully controlled charge and discharge, during which a protective layer forms on the electrode surface, the step that makes the cell a working battery rather than an assembled one. Aging is the rest period that follows, where the cell stabilizes and any weak or defective cells reveal themselves through abnormal voltage behavior. Together, formation and aging commonly take one to three weeks, which makes cell finishing the slowest stage on the line and one of the largest shares of total manufacturing cost.

That long tail has real operational consequences. It means enormous amounts of capital tied up in cells that are made but not yet sellable, huge floor space filled with formation and aging racks, and a long feedback delay between a process problem upstream and its discovery downstream. A coating or contamination defect introduced on Monday might only surface as an aging failure two weeks later, by which point the line has made a great deal more of the same defect. Shortening that feedback loop is one of the highest-value problems in the whole plant.

Formation and aging dominate the timelineWhere the calendar goesELECASSYFORMATION + AGINGoften 1 to 3 weeks of controlled charge, rest, and gradinghourshours-dayweeksA defect made upstream may not surface until aging, weeks and many cells later
Formation and aging dwarf the steps before them. That long tail ties up capital and creates a slow feedback loop between an upstream defect and its discovery.

Why is yield the number that decides a gigafactory?

Because yield compounds across dozens of sequential steps, and small per-step losses combine into a large scrap rate. If a cell must clear fifty steps and each is 99 percent good, the compounded yield is roughly 0.99 to the fiftieth power, around 60 percent, not 99. The math is unforgiving: to reach a healthy overall yield, every step has to be excellent, because one weak step drags the whole product down. This is why first-pass yield and its compounding, not headline throughput, is the metric battery operators obsess over.

The practical consequence is that improving a gigafactory means finding and fixing the defect sources at every stage, and doing it fast, before the long formation-and-aging delay lets a problem run for weeks. That takes contamination control, tight process windows on coating and calendering, and cell-level traceability that ties each cell back to its electrode lots and machine parameters, so when an aging failure appears, engineers can trace it to a root cause instead of guessing. Structured defect tracking and statistical process control across the line are the tools that make that possible.

Yield compounds downward across many steps99% good per step is not 99% good overall100%0%start 100%~60% after 50 stepsEvery step has to be excellent, because one weak step drags the whole product down
Yield multiplies, so a step that looks nearly perfect still costs real output when there are dozens of them. Raising overall yield means attacking defects everywhere, not just the obvious losses.

How do you run a battery plant well as it scales?

Scale-up is where most of the pain lives, because a process that works at pilot scale rarely transfers cleanly to volume. Here is the operating sequence that separates a smooth ramp from a scrap generator.

  1. Lock down contamination control first. Hold dry-room humidity and particle limits as non-negotiable. Moisture and metal fines are silent yield-killers that surface only at formation.
  2. Tighten the front-end process windows. Coating uniformity and calendering density set cell consistency; control them tightly, because errors here compound downstream.
  3. Capture cell-level traceability at the source. Tie each cell to its electrode lots, coating and calendering parameters, and assembly conditions as it is made, so an aging failure can be traced to a cause.
  4. Shorten the feedback loop from finishing to the front. Connect formation and aging results back to the upstream process so a defect source is caught in days, not weeks.
  5. Manage yield as the primary metric. Track first-pass yield by step and attack the biggest losses, not just the most visible stops.
  6. Attack downtime on the pacing equipment. Coaters, dry rooms, and formation systems are expensive; unplanned stops there are costly, so track machine downtime and reliability deliberately.
  7. Standardize and preserve process knowledge. Keep the current process settings and recipes in front of every crew so a good ramp is repeatable across shifts and lines; see standard work.

What do the sources say?

Where does an operational layer fit in a gigafactory?

Right in the long, expensive gap between an upstream defect and its downstream discovery. A battery plant rarely lacks advanced equipment; it loses yield because process data, traceability, and defect signals live in separate systems, so when an aging failure appears two weeks after the coating that caused it, the trail is cold. An operational layer that captures process parameters, contamination checks, and cell-level traceability as the work happens, and connects formation results back to the front of the line, turns that slow, cold trail into a fast root-cause loop. That is the honest value: not replacing the chemistry or the equipment, but connecting them so yield problems are found in days instead of weeks. It is the same real-time capture CLS used to replace paper logging with live floor data (the CLS case study). For the systems picture, see what is a manufacturing operating system and how Harmony connects the floor. No rip-and-replace, connect the machines and systems you already run.