A data historian is a specialized database that is an industrial plant’s record of what happened. It continuously stores timestamped readings from control systems, so engineers can replay how a process behaved, trace what caused a fault, and produce reports and compliance records.

What does a data historian do?

A data historian connects to the systems that run industrial equipment: programmable logic controllers (PLCs), distributed control systems (DCS), and supervisory control and data acquisition (SCADA) systems. It records each value with a timestamp, continuously, so the plant keeps an unbroken record of process behavior. It stores that record efficiently and serves it back in time order, so an engineer can pull up a tank level, a motor current, or a line pressure for any past interval.

A historian reads from control systems but doesn’t control equipment. That separation lets teams query, trend, and report on process data without touching the systems that keep the process running. Because the historian serves as the trusted record of what happened, it is built for stable, continuous capture.

Data historian use cases

How do operational technology teams use a data historian?

Operational technology (OT) teams at the plant rely on the historian for three kinds of work:

  1. Process visibility. Trends and historical views show how a process behaves over hours, shifts, or months, and whether it’s drifting.
  2. Troubleshooting and root-cause analysis. After a fault, trip, or excursion, engineers replay the signals around the event to see what changed first.
  3. Reporting and compliance. Time-aligned historical data supports production reports and regulatory records. In pharmaceutical manufacturing, for example, good manufacturing practice (GMP) requirements call for long-term retention of data about each batch produced.

These uses share a pattern. They look back at one plant’s process, at the rates control systems already report, and they value a trustworthy record of events over every raw reading a sensor could produce.

Uses across key industries

Industry Use case
Manufacturing Data historians can monitor machine efficiency and productivity, allowing for identification of bottlenecks and inefficiencies. Predictive maintenance use cases can foresee potential machine breakdowns, minimizing downtime.
Oil and gas Data historians help track the performance of drilling rigs, pipelines, and other equipment. Predictive analysis can identify potential issues, while historical data can assist in decision-making processes and improve safety measures.
Transportation and logistics Historians can track and analyze data from various sensors in vehicles and equipment, improving maintenance scheduling, fuel efficiency, and overall fleet management.
Utilities Data historians track and analyze energy consumption data, helping identify inefficiencies and opportunities for cost savings. Additionally, they can monitor equipment status in real-time, predict potential outages, and enable rapid response.

Where do traditional data historians fit, and where do they strain?

Historians emerged in the 1980s, when sensors, storage, and compute were all expensive and sized to run on-site. They’re strong at what they were built for: a dependable record of one plant’s process. The strain shows up as data volume, sites, and analytics demands grow.

Where historians excel Where they strain
Reliability and continuous capture Built for stable, long-term operation with an unbroken record Historians serve a single site so cross-site aggregation needs workarounds and data can end up siloed
Connecting to control systems Collect directly from PLCs, DCS, and SCADA without touching the equipment Harder to integrate with modern analytics and cloud tools
Compliance and audit Secure, auditable records with long-term retention Points within a tolerance are discarded, so brief events like vibration spikes or early fault signatures can't be recovered, resulting in lossy compression.
Cost as you grow Predictable for a stable, well-defined set of signals Every new sensor or asset is a licensing decision, even though sensors are now cheap, and per-tag licensing can escalate quickly.
Speed of insight Strong at looking back: trends, event replay, and reports Often oriented toward historical review rather than real-time analysis across teams

How is a data historian different from a time series database?

A data historian is an industrial data system designed to maintain a trusted record of plant operations for monitoring, troubleshooting, reporting, and compliance. By comparison, a time series database is a general-purpose data system for high-volume, timestamped data. It can serve as a historian while also supporting real-time monitoring, analytics, applications, predictive maintenance, and AI inference across systems and sites. See the full breakdown in Data Historians vs. Time Series Databases.

Frequently asked questions

Does a data historian control industrial equipment?
No. A data historian reads and records values from control systems such as PLCs, DCS, and SCADA, but it doesn't send commands to equipment. In the Purdue Model, control happens at Levels 1 and 2, while historians sit at Level 3 with other plant operations systems. Queries and reports run against the historian without affecting the control loop.

How long do data historians keep data?
Retention depends on the plant and its obligations. Historians are built for long-term storage, and regulated records, such as pharmaceutical batch data under GMP, may be kept for many years or indefinitely. Because long retention multiplies storage needs, traditional historians rely on compression to keep archives manageable, trading some signal detail for storage efficiency.

Can one data historian collect data from multiple plants?
Historian architectures are plant-centric: they were designed for one site's storage and reporting. Aggregating data across sites, fleets, or distributed assets usually requires workarounds that grow harder to maintain as sites are added. Many organizations add a central time series layer for cross-site analysis instead. See when to complement your data historian with a time series database.

Are time series databases priced per-tag like data historians?
In a historian, a tag is one measured point. Many time series databases model the same reading a bit differently. For example, in InfluxDB 3, a pump-pressure reading is stored with a timestamp, the pressure as a field, and the site, line, and asset as tags: metadata that describes the reading rather than a separate stream. Whether historian assumptions still fit a workload is the subject of how to choose between a data historian and a time series database.