TL;DR

Traditional MES manages top-down — when the network drops, the line stops. mes-line inverts this: each line runs independently, recipe-driven, offline-capable, add lines without touching the architecture.

The Fatal Flaw of Traditional MES

MES is the factory's brain. But when that brain sits in a server room 500 meters away and every command depends on the network — a network outage becomes a production outage. The line stops. The MES shows "offline." The production manager waits for IT.

This isn't hypothetical. It happens every day on factory floors.

mes-line's design logic is simple: each production line gets its own brain. No factory network approval needed for autonomous execution. When connected, data syncs to Online. When disconnected, the line keeps running.

mes-line module architecture

Recipe hub at the center, 9 modules around it. Change the recipe, everything adapts. No hard-coded dependencies between modules.

Recipe: Not a Feature. The Architectural Backbone.

In traditional MES, modules call each other directly. Work orders call equipment. Production calls quality. Change one module, the ripple effects break three others.

mes-line uses recipes as the single integration point. Work orders match engineering and equipment through recipes. Production pulls spec tables from recipes. OEE/CPK computes from production data. No module directly calls another module. Every relationship flows through the recipe table. Want to add a new module? Just read the recipe table. Zero intrusion.

Nine Modules. You Don't Have to Use All of Them.

ModuleIn One Sentence
Recipe MgmtLine/station hierarchy, JSON version history, rollback any change
Work OrdersEnter a product code → auto-match recipe → auto-fill engineering and equipment
ProductionAuto-acquisition or manual entry, per-item OK/NG verdict
EquipmentLine→Station→Device 3-tier, cascading cleanup across 18 tables on delete
OEEA×P×Q real-time, configurable takt time
CPKCp/Cpk/Ppk, device/line level, instant alert below threshold
Dashboard9 KPIs + device grid + alert ticker, big-screen standalone mode
SyncBidirectional with Online, resume-from-break
HMIPlugin-per-equipment repo model, independent versioning

Tsingley DAQ: Six Protocols, One Engine

Modbus, Mitsubishi MC, Siemens S7, OPC UA, UDP, Node-RED — no need to write a new driver for every project. Tsingley DAQ ships with six pre-built protocol adapters. Background polling, auto-reconnect, JSON persistence. MockDriver simulation mode means you can develop and test without a single PLC on hand.

DAQ protocol support

DAQ pipeline: Tag config → Background polling → Auto-reconnect → JSON storage → Rule engine trigger

Standalone or Line Edition?

Independent machines (CNC, cleaner, marker) → Standalone, 1 per machine. Multiple machines jointly producing the same product → Line Edition, 1 set for the whole line. The deciding question: same product item? Continuous process?

Standalone upgrades to Line Edition at any time — add stations, add equipment. No redeployment needed.

Standalone vs Line comparison

Standalone: 1 machine = 1 system, fully isolated. Line Edition: 10-15 machines = 1 system, divided by process station.

Should You Choose mes-line?

If your MES already works but you're tired of "network-down-line-down," "upgrade-means-plant-shutdown," and "add-a-line-means-rearchitect" — mes-line sits between your equipment and your MES, adding line-level autonomy without replacing anything. If you don't have a system yet, mes-line is the first one.