ISA-95 Five Levels Map the Path From Shop Floor Sensors to ERP

The ISA-95 reference model defines five levels that organize manufacturing data from the physical process up to business planning, and it is technology-neutral: plants can apply it across different vendors, equipment generations, and automation levels. That is the working map Minitab lays out in its guide to connecting MES, ERP, and shop floor data. The article's central claim is concrete: integration fails or succeeds on whether measurement context — timestamp, unit of measure, asset identifier, and product or process context — survives the trip between systems.

Where the data starts

At Level 1, sensors, smart devices, valves, and other equipment measure temperature, pressure, flow, vibration, and speed. A raw measurement becomes useful only when it carries a timestamp, a unit of measure, an asset identifier, and enough process context to interpret it. This is also the level where material is cut, mixed, heated, filled, or assembled.

The takeaway: most shop floor data collection begins at the process itself, not in the database.

What the control layer adds

PLCs and distributed control systems give raw sensor values operational context by relating them to equipment states, setpoints, alarms, and production cycles. The article warns against sending every high-frequency tag directly to enterprise systems. A well-designed connection filters and aggregates events, measurements, and summaries without compromising performance or security.

SCADA systems sit in this layer, letting operators monitor equipment status, alarms, and process trends while exposing relevant data to operations systems. The design rule is selective aggregation, not wholesale piping.

How quality records connect to production

MES and broader MOM systems link machine activity and SCADA data to work orders, lots, recipes, labor, and work in process. Quality data from gages, laboratory systems, and QMS records should share the same product, batch, serial number, asset, and time references as production data.

With that linkage in place, teams can determine whether a process change, alarm, material lot, or machine condition preceded a defect — rather than comparing disconnected reports after the fact. That is the traceability chain buyers of SPC software should test before purchase.

What moves to ERP

ERP handles business planning and logistics. It sends demand forecasts, production orders, material requirements, and schedules down to operations; MES returns actual output, material consumption, yield, and production status. Planners then work from current plant performance while production teams see the context behind changing priorities. Defining this enterprise-to-manufacturing exchange is a central focus of ISA-95 itself.

Which records to integrate first

Legacy machines, spreadsheets, custom databases, and inconsistent naming conventions make integration hard. Minitab's answer is not one giant repository. The recommended sequence:

• Start with a decision that currently needs manual reconciliation — tracing a quality issue, explaining downtime, or comparing planned versus actual output.
• Identify the few records across levels needed to answer that question.
• For a quality use case, build the chain: work order → product → material lot → machine → process conditions → inspection result.
• Agree on identifiers, timestamps, units, and definitions before building the data flow.
• Preserve the authoritative system for each record; integration should connect records and standardize their meaning, not duplicate them.

What the connected factory looks like in practice

A fully connected factory is difficult to achieve and harder to maintain, because equipment, systems, and priorities keep changing — the article treats integration as ongoing work, not a project with an end date. As reliable relationships form across production and business data, manufacturers can automate data collection, prepare data for analysis, monitor performance in real time, and apply statistical process control with less manual effort.

Minitab names two tools for that stage: Minitab Connect and Prolink, which bring data from multiple sources into governed, repeatable pipelines.

The open question for any plant running this playbook is whether its identifiers and timestamps are consistent enough today to support the audit, traceability, or SPC deployment it plans to fund next year.