data collection

Core meaning

Data collection is the systematic capture of raw data from people, machines, and software systems into a defined structure so it can be stored, processed, and analyzed.

In industrial and regulated manufacturing environments, data collection commonly refers to the ongoing recording of production, quality, maintenance, and environmental data from shop-floor activities and related IT/OT systems.

How data collection is used in manufacturing

Typical uses in manufacturing and industrial operations include:

– **Production execution data**: start/stop times, quantities produced, scrap counts, downtimes, changeovers, and operator IDs.
– **Quality and inspection data**: measurements, test results, pass/fail decisions, sampling results, and inspection sign-offs.
– **Equipment and process data**: temperatures, pressures, speeds, setpoints, alarms, and events from PLCs, SCADA, historians, and sensors.
– **Material and genealogy data**: lot/batch IDs, serial numbers, material consumption, and traceability links between inputs and finished goods.
– **Compliance and log data**: electronic signatures, access logs, audit trails, and records needed to support internal and external audits.

Data may be collected:

– **Manually** (operators entering values on MES terminals, checklists, or forms)
– **Automatically** (direct acquisition from machines, instruments, and control systems)
– **Via integration** (exchanging data between MES, ERP, LIMS, CMMS, WMS, and other enterprise systems)

Boundaries and scope

Data collection:

– **Includes** the processes, configurations, and mechanisms used to capture and store raw data at the point of occurrence.
– **Includes** the definition of what data is captured (data model, fields, timestamps, identifiers) and at what frequency or trigger.
– **Does not inherently include** analysis, reporting, or decision-making; those are downstream activities (such as analytics, reporting, or operations intelligence) that use collected data.
– **Does not necessarily imply enforcement** of process steps or rules; it can record what happened without preventing non-compliant actions.

Common methods and system roles

In OT/IT and MES contexts, data collection commonly involves:

– **MES and electronic batch records (EBR)** capturing operator actions, inspections, and process parameters during execution.
– **Data historians** storing high-frequency time-series data from control systems and sensors.
– **SCADA/HMI systems** logging alarms, events, and process values.
– **Edge devices and gateways** aggregating and normalizing data from heterogeneous equipment.
– **Integration middleware or APIs** transferring data into central repositories, data lakes, or enterprise applications.

Configuration typically specifies:

– Trigger conditions (e.g., at operation start/end, at defined time intervals, at alarm, on measurement)
– Required versus optional fields
– Data validation rules and formats
– Time synchronization and source identification

Relation to MES enforcement (site context)

When discussing whether an MES can enforce that specific steps or inspections are completed, **data collection** refers to recording the evidence that those steps or inspections were carried out (e.g., entering inspection results, capturing electronic signatures, logging machine parameters).

In this context:

– Data collection provides the **record** that an operation, test, or check occurred.
– Enforcement logic in MES uses required data collection points, routing rules, and status checks to **block progression** until the needed data has been entered or captured.
– Audit trails and access logs are collected data that demonstrate who performed or approved each action.

However, the mere presence of data collection screens or fields does not guarantee enforcement; enforcement depends on how the MES and related systems use collected data in workflow and authorization rules.

Common confusions and related terms

– **Data collection vs. data logging**: Data logging often implies automated, continuous recording (e.g., from sensors), whereas data collection is broader and includes manual and event-based capture.
– **Data collection vs. data acquisition (DAQ)**: Data acquisition usually refers to the technical process of reading signals from hardware; data collection is wider and includes human inputs, system integrations, and structured storage.
– **Data collection vs. data analysis**: Collection is about capturing and storing; analysis is about interpreting and using that data.

Using these terms precisely helps distinguish between:

– The act of recording what happened (data collection)
– The mechanisms that read physical signals (data acquisition)
– The processing and interpretation of information (analytics, reporting, operations intelligence)