If your data can’t move, your factory won’t either - The state of Historians in 2025

Too many factories still run on historians that can’t scale, connect or share. That slows everything down, from reporting and root cause analysis to AI and enterprise integration. Our CEO Jeroen Coussement breaks down what modern manufacturers should expect from a historian today: structured data, built-in context, real-time access, and zero friction between OT and IT.

New demands require new data strategies

Modern manufacturing generates huge amounts of data, from every process, sensor and operator. Data drives innovation, but that potential goes nowhere if data is not easily accessible or organized.

Meanwhile, the bar keeps rising: integrate production data with enterprise systems, support evolving use cases, deliver insights faster… while getting ready for AI, and doing each data project more efficiently than the last.

Those goals don’t add up without a proper foundation. Data needs to be available, structured, contextualized and reusable for any application.

How traditional historians became bottlenecks

Traditional historians, like AVEVA PI or GE Proficy Historian, have long been trusted tools for collecting and archiving time-series data from control systems, especially for engineering and operations teams. For decades, they’ve helped manufacturers monitor production, troubleshoot issues, and keep historical records for compliance.

But these systems were built for static environments, fewer integrations, and specialist users. As factories evolve, those assumptions no longer hold. Data needs to move across departments, connect seamlessly to cloud platforms, and support everything from real-time dashboards to AI models.

That’s where traditional historians start to show their limits. They weren’t necessarily designed to connect easily with other tools or systems or serve a broader range of users, which slows down every new project before it even starts. Despite the best efforts of IT teams, these systems often require specialized knowledge to extract and transform data, creating delays and bottlenecks for other departments.

From Storage to Strategy

The expectations for the modern historian are changing. While collecting and storing time-series data has always been a core strength, manufacturers now need systems that support structuring, sharing and applying data across tools, teams, and technologies.

Next-generation historians are built with exactly that in mind. They make production data usable. Not just by a small team of experts, but across the organization.

And as more teams rely on that data to make decisions, the need for ownership and control only grows. Manufacturers increasingly request more flexibility and visibility into where data is stored, who can access it, and how it integrates with other tools. That’s a shift modern historians must be ready to support.

Requirements of Historians in 2025

1. A data platform, not a storage vault

Next-generation historians are designed to make production data accessible, structured, and ready to move. They break away from vendor lock-in by supporting open formats like Parquet and modern protocols like MQTT and OPC-UA out of the box. That means data can be easily forwarded to other systems, whether it’s needed in the cloud, for AI applications, or for BI tools. Without the friction of proprietary connectors or restrictive licenses.

With a modern, open historian, you avoid the usual barriers: chasing IT for access, waiting weeks for a basic report, or hitting a wall every time a new data request comes in. Instead, data flows freely. It’s structured, accessible, and ready to be used across systems, teams, and tools.

A historian shouldn’t just store and display data. It should redistribute it, clean, structured, and wherever it’s needed.

Jeroen Coussement

Founder & CEO at Factry

2. Designed for the citizen data scientist (not just engineers)

You can’t drive factory-wide improvement if only one team sees the data. Modern historians should put insights in everyone’s hands, with intuitive interfaces that don’t require query builders, SQL skills or IT support.

This way, operational teams become independent data explorers who can spot downtime, waste, or process deviations and act right away. For example, operators visualizing hourly water consumption at the machine level can immediately identify waste and take action. Similarly, maintenance teams can correlate temperature patterns with equipment failures, while quality managers can trace batch anomalies across multiple process variables, all without IT intervention or complex training.

You can’t give people just static dashboards anymore. They need the tools to dig into data themselves. When people explore the data themselves, they don’t just find answers, they trust them. That sense of ownership leads to faster action, stronger follow-through, and continuous improvement from the ground up.

Jeroen Coussement

Founder & CEO at Factry

3. Data with built-in context, ready to use

Although raw timestamped data has value on its own, adding context significantly increases its usefulness. To achieve that, modern historians should automatically detect and calculate production events, such as batch completions, machine stops or material overfills, based on incoming signal patterns. Having this event logic built in, rather than added later, makes it easier to get consistent, usable insights from the start.

At the same time, data should be organized to reflect how production actually works, structured around assets, lines or processes that teams recognize. That way, analytics tools and people work with meaningful input instead of raw, anonymous time series data.

Reprocessing and relabelling become the exception, not the rule. With structure and context built in, teams can move faster from data to decisions.

4. Foundation for AI and advanced analytics

AI and analytics projects are only as good as the data they’re built on. The reality is that many AI initiatives fail not because of the algorithms, but because of poor data foundations. Without a consistent, well-structured source, results are unreliable, and every new use case means starting over.

A modern historian should act as a structured, contextualized layer that keeps data consistent across assets, systems and time, making it possible to build once and reuse across use cases. Without this foundation, each data project becomes a custom, throwaway exercise. With it, teams can move from isolated wins to sustainable rollouts that deliver actual business value.

Don’t build AI proof-of-concepts without clean, structured data. It won’t scale, and it won’t stick.

Jeroen Coussement

Founder & CEO at Factry

5. Ready to scale as factories evolve

Modern factories change constantly. New sensors, new lines, new tools, new people, new legislation. Today’s historians should handle those changes by design without needing to rebuild everything from scratch.

When data is modelled, structured and shared constantly, it can be reused across projects. Scaling to new sites or additional use cases becomes incremental, not a ground-up effort.

Modern historians need fine-grained access control, audit trails for traceability, and clear ownership over data sources and flows. This ensures secure, compliant use of data as systems, teams, and requirements evolve. This becomes even more critical with regulations like CSRD, which requires detailed tracking and reporting of resource consumption, emissions, and energy usage across manufacturing operations.

Factry Historian: the backbone of data-driven production

Factry Historian brings together the five key traits of a modern historian in one open, production-ready platform. It collects high-resolution data directly from PLCs, SCADA systems and sensors, structures it into asset models that reflect the real production layout, and detects key events such as batch completions, downtime or anomalies as they happen. The result is clean, contextualised data that’s ready to use across roles, systems and sites, significantly reducing the need for scripting or reprocessing.

From raw data to fast decisions

Engineers and operators need to act quickly, but too often the data isn’t there when it’s needed. Factry Historian changes that by making well-organised data available instantly. No tickets, no exports, no delay.

Let’s say a process engineer wants to understand why a batch from the night shift had a quality dip. With a few clicks, they open high-resolution trends, compare them with previous runs, and visually overlay changes like parameter tweaks or maintenance actions. If something went wrong at 2:07 AM, they isolate that moment and immediately see correlated values like pressure, temperature and flow rate. This turns process optimisation into a fast, continuous routine.

Meanwhile, data teams work from the same reliable dataset. Factry Historian outputs process data in open formats like Parquet, ready to feed into Python notebooks, cloud pipelines or machine learning models. The foundation is consistent, reusable, and ready for AI or whatever comes next.

Accessible for every role, not just experts

What sets Factry Historian apart is how broadly the data can be used. From operators and line managers to process engineers and data scientists, everyone gets the tools to explore and act. Information flows without friction between shop floor and analytics teams, helping each role move faster, with better input.

When data drives action, change follows

Once non-technical users can explore data on their own, the effect is immediate. That’s when energy savings happen. That’s when line performance improves. That’s when unexpected wins show up, often within weeks.

As operations evolve, Factry Historian scales with them. Adding new sensors, new lines or new sites doesn’t require rebuilding anything. That structure is already in place, so insight becomes repeatable and part of daily operations.

See it for yourself