UMH Historian

Ask your site anything.

Keep your UNS data long-term, in the model you already defined. No second system needs to be built.
Talk to our team
See the documentation

Trusted by global enterprises

The Problem

You have the data. Getting an answer is a project.

Did last week's setting change actually help to improve OEE? Did this process run the way it was supposed to?
To answer these questions, you need the history in place.
Rigid and inflexible
The historian owns its own asset model, separate from your UNS. The data is there, but reaching it takes time and manual effort.
Hard to maintain
A gateway, a database, a hand-written schema, queries rebuilt per project. Four things to keep in sync, and one person who knows how.
Doesn’t scale
Adding a new line means starting from scratch. A view built for one line doesn't work on the next, and the cost grows with every tag you connect.
Our solution

The model you already defined carries into history.

Dashboards, alerts and any statistical analysis all need to compare now against before. The UNS only carries the present. You need data from the past. That’s where UMH Historian comes in.
The Historian subscribes to the sources you’ve already defined in UMH Core. Dozens of standard connectors, with monitoring included.
Connect
Describe and model your site once. The data in the Historian mirrors the UNS structure exactly, so no separate configuration is needed.
Model
Downsampling is available in UMH Core. Compression and retention are configurable. Choose what data you want to keep and for how long.
Store
An automatic pipeline is generated into Timescale. Standard views and tables are available, so Grafana queries the hierarchy rather than raw tables.
Consume

Use Cases

OEE and Production Visibility
What is the impact of last week’s change? Look at the data across any time frame, identify the trends and make comparisons. All from the model you defined once.
Learn more
Energy and Resource Management
Is this line using more than it should? Consumption tracked over any time frame, compared across lines and products, so you see the change before the invoice comes.
Learn more
Monitoring and Alerting
Is this normal for this machine? Trends over weeks and years tell you what normal looks like, so you can set the threshold and spot the deviation early enough.
Learn more
Quality and Compliance
Did this process run the way it was supposed to? Heating and cooling steps, the duration of the process, held against a batch. For any record you need to keep.
Webinar

UMH Historian: Store your UNS data and query it from Grafana

Why gateway-to-database setups break, and a live walkthrough from connecting a machine to your first Grafana panel.
Watch now >

Frequently asked questions.

How long is data stored, and can we control that?

Retention and compression settings are configurable. You set what to keep at full detail, how long data is kept and when it is deleted.

Can we use our own Timescale and Grafana?

Yes. UMH provides the integration between them. Managed services such as Timescale Cloud and Grafana Cloud also work.

Do we have to use Grafana?

No. Data is stored in Postgres and queried with standard SQL, so any tool that can connect to Postgres will work. Grafana is included in the recommended UMH stack. Note that SQL is the query interface, so teams without SQL knowledge will need support for ad hoc queries.

Who is this for, and who is it not for?

It fits companies with no historian today, discrete manufacturers with historian use cases, and process companies running it alongside an existing historian. If your environment requires a long validated track record before adoption, talk to us first so we can look at your specific situation.

We already have AVEVA PI. Can we still use UMH Historian?

Yes. UMH runs alongside AVEVA PI rather than replacing it. Data is collected from your PLCs and sent to PI as it is today, while the same data also becomes available for new use cases. Existing Asset Framework models can be migrated into UMH data models when you want to move something over. There is no need to migrate historical data or revalidate the processes built on PI.

UMH Historian

Start using the data you already have

Talk to our team about your specific use case, or explore the platform yourself, fully open source.
Talk to our team
Explore the documentation