About the Manufacturing Operations Matrix

Taxonomy, priority methodology, and data sources

What This Is

This matrix catalogs 210 manufacturing unit operations — every major way humans shape, join, cut, coat, and treat materials — and evaluates each one for integration into Speculative Technologies' compiled manufacturing cells.

Each operation is assessed across four dimensions: simulation readiness (can we model it before atoms move?), software-defined machines (do programmable machines exist?), cost scaling (how does unit cost change from 1 to 1000 parts?), and startup transient (NRE, capital, and time to first part).

Compiled Manufacturing = software-defined, closed-loop, modular manufacturing cells that can produce anything from a software spec. The cell IS the production line. Scale by adding cells, not building factories.

Taxonomy

Operations are organized primarily by DIN 8580, the German standard that classifies manufacturing processes into six groups by how they change material cohesion:

GroupGermanEnglishDescription
1UrformenPrimary shapingCreating shape from formless material (casting, sintering, AM)
2UmformenFormingChanging shape without removing material (forging, bending, rolling)
3TrennenSeparatingRemoving material (machining, cutting, grinding, EDM)
4FugenJoiningConnecting parts (welding, brazing, fastening, adhesive)
5BeschichtenCoatingApplying material layers (plating, painting, PVD, thermal spray)
6Stoffeigenschaft ändernChanging propertiesAltering material properties (heat treatment, hardening)

The 210 operations span 18 categories (A through R), each with a letter code. Supplementary standards — ASTM/ISO 52900 for additive manufacturing, ISO 4063 for welding — provide finer classification where DIN 8580 is coarse.

CodeCategoryDIN GroupCount
ACasting1 - Urformen14
BPowder Metallurgy / Sintering1 - Urformen5
CBulk Deformation / Forming2 - Umformen16
DSheet Metal Forming2 - Umformen15
EConventional Machining3 - Trennen (3.2)18
FAbrasive Processes3 - Trennen (3.3)10
GNon-Traditional Material Removal3 - Trennen (3.4)12
HAdditive Manufacturing1 / 515
IFusion Welding4 - Fugen (4.6)12
JSolid-State Joining / Brazing / Adhesive4 - Fugen10
KMechanical Joining / Fastening4 - Fugen (4.3)5
LPolymer and Composite Processing1 / 419
MSurface Treatment and Finishing5 - Beschichten14
NHeat Treatment6 - Stoffeigenschaft ändern10
OElectronics ManufacturingMixed10
PInspection and Testing (NDT)N/A (quality)10
QAssembly and Material Handling4 - Fugen (4.1)5
RElectric Motor ManufacturingMixed10

Cell Integration Tiers

Each operation is assigned a cell integration tier indicating how naturally it fits inside a shipping-container-scale manufacturing cell:

1
Robot End-Effector
Tool mounts directly on a robotic arm. Fully software-defined, fast changeover between operations. Examples: laser cutting head, FSW tool, deburring spindle.
2
Dedicated Station
Standalone machine inside the cell with robotic loading/unloading. Examples: CNC mill, press brake, powder bed fusion printer.
3
Shared Resource
Too large, complex, or slow for per-cell integration. Shared across multiple cells. Examples: HIP unit, electroplating line, large vacuum furnace.
4
Not Cell-Compatible
Requires factory-scale infrastructure, high volume, or heavy manual operation. Examples: continuous casting, blast furnace, manual grinding.

Priority Methodology

Each operation is assigned a priority — high, medium, or low — based on a composite of two factors: how much NRE cost compiled manufacturing can eliminate, and how well the operation fits into a modular cell.

Composite Scoring

Priority = NRE score (0–3) + Cell-fit score (0–3), for a total range of 0–6.

NRE score — higher tooling/setup cost = higher value from compiled manufacturing:

NRE costScore
≥ $20,0003
≥ $5,0002
≥ $1,0001
< $1,0000

Cell-fit score — how readily the operation integrates into a software-defined cell:

ConditionScore
Tier 1 + fully programmable3
Tier 2 + fully programmable2.5
Tier 2 + partially programmable2
Tier 3 + fully/partially programmable1
Tier 3–4 + manual/batch, or not cell-applicable0

Priority thresholds:

Composite scorePriority
≥ 4high
≥ 2medium
< 2low

Distribution

High
72 ops (34%)
Medium
99 ops (47%)
Low
39 ops (19%)

Rationale

The previous scoring weighted cell-fit almost exclusively. The composite score rebalances to give equal weight to NRE elimination — the thesis that compiled manufacturing's biggest economic value is eliminating per-geometry tooling costs. Operations with expensive tooling (dies, molds, fixtures) that can also fit in a cell score highest. Operations that are great cell fits but already tool-less (e.g., ISF, CNC milling) score medium — easy to integrate but lower economic leverage.

Notable Examples

D15 Stamping / Progressive Die — NRE $150K (score 3) but Tier 4 / not cell-applicable (score 0) → composite 3 → medium. High NRE but poor cell fit.
G07 Laser Cutting — NRE $1K (score 1) + Tier 1 fully programmable (score 3) → composite 4 → high. Good cell fit plus some NRE.
D14 Incremental Sheet Forming — NRE $0 (score 0) + Tier 1 fully programmable (score 3) → composite 3 → medium. Excellent cell fit but no NRE to eliminate — already tool-less.
A04 Die Casting — NRE $50–80K (score 3) + Tier 3 partially programmable (score 1) → composite 4 → high. Huge NRE savings potential.
L09 Filament Winding — Core spectech process. Robot-mountable winding head, fully software-defined fiber paths, fast changeover via mandrel swap. The first demo part (pressure vessel) uses this.

Startup Transient

Each operation tracks four numbers that characterize how hard it is to get started — the ramp from "we decided to do this" to "we're producing good parts at steady state."

FieldUnitsWhat it measures
NRE CostUSDNon-recurring engineering cost: tooling design, mold/die fabrication, fixture builds, process development — everything you pay once regardless of how many parts you make. This is the cost compiled manufacturing aims to eliminate.
Capital CostUSDEquipment/capital expenditure to acquire the machine if you don't already own one. Ranges from ~$5K for a benchtop tool to $2M+ for a large CNC or HIP unit.
Time to First PartDaysCalendar days from the decision to produce a part to the first acceptable part off the process. Includes procurement, setup, tooling fabrication, and initial process tuning.
Learning CurveUnitsNumber of parts you need to produce before the process reaches steady-state quality and throughput. A learning curve of 5 means the first ~5 parts may have higher scrap rates, longer cycle times, or require operator adjustments. Processes with tight tolerances, novel materials, or sensitive parameters (e.g., LPBF, investment casting) have longer learning curves. Highly repeatable software-defined processes (CNC milling, laser cutting) tend toward 1–3 units.

The Startup Scatter Plot visualizes NRE vs. Capital Cost across all 210 operations. The Table View shows all four startup transient fields side by side.

Cost Estimates

All cost estimates are back-of-envelope approximations intended for rough comparison, not procurement decisions. They assume:

Cost scaling is shown at four quantities (1, 10, 100, 1000) to illustrate how different processes behave. Tooling-intensive processes (die casting, injection molding) show steep cost reduction with volume. Software-defined processes (CNC machining, laser cutting, AM) show flatter curves.

Cost Linearity

Two linearity metrics capture how per-unit cost scales with volume, and how much of that scaling is driven by NRE.

Production linearity (labeled "Lin." in the table) measures the per-unit cost curve as reported in the cost scaling data — production costs only, excluding startup NRE:

production linearity = per-unit cost at qty 1,000 ÷ per-unit cost at qty 1

NRE-loaded linearity (labeled "Lin.+NRE") shows the total cost curve a customer experiences when startup NRE (tooling, engineering, process development) is amortized across the production run:

NRE-loaded linearity = (NRE/1000 + qty-1000 cost) ÷ (NRE + qty-1 cost)

Interpretation: Both metrics range from 1.0 (perfectly flat cost curve) to near 0 (steep volume discount). The gap between them reveals how much NRE drives the cost curve. When the two values are similar, NRE is a small factor; when NRE-loaded linearity is much lower than production linearity, NRE dominates the economics.

OperationProduction Lin.NRE-loaded Lin.Ratio
D14 — ISF0.7500.7501.0x
I01 — SMAW0.5000.04610.8x
G07 — Laser Cutting0.1330.00527.5x
A04 — Die Casting0.0020.0011.5x

Why it matters for compiled manufacturing: Production linearity represents the compiled manufacturing scenario — the cell already has equipment, and there is no per-geometry tooling NRE. NRE-loaded linearity represents the traditional manufacturing scenario. The ratio between them quantifies compiled manufacturing's economic advantage: laser cutting becomes 27x more linear when NRE is eliminated; ISF, which is already tool-less, shows no change.

Note on capital costs: Equipment capital ($100K–$2M) is excluded from both metrics because it applies equally to both scenarios — you need machines whether they are in a cell or a factory. Capital costs swamp per-unit production costs and would compress all operations to near-identical linearity values (~0.001), obscuring the meaningful variation between processes.

Data Sources

Taxonomic Standards

Textbooks

Simulation Tools Referenced

Machine Vendors Referenced