Authors: Sylvain Gouttebroze, Senior Research Scientist, SINTEF Industry; Treesa Rose Joseph, Software Developer, SINTEF Ocean; Sridevi Krishnamurthi, Research Scientist, SINTEF Industry

Imagine if every piece of factory gear came with a digital passport. Specifically, such passport tells the battery production manager what it is, what it measures, and how to communicate with it. In BATMACHINE, this passport is the equipment wrapper. The wrapper is a compact set of files and software bridges that slots any machine—regardless of age or manufacturer—into the project’s data space (i.e. its data ecosystem).

In this article, BATMACHINE partner SINTEF will talk about wrappers and their potential role in machinery for battery manufacturing. Specifically, SINTEF will discuss the role they play in machinery interconnectivity and communication, process improvement, and data collection.

What the wrapper does

First, the wrapper harvests data. Online machines list their live OPC UA channels, which flow directly into a timeseries database. Offline gear— like lab testers—is given structured data models, so their output looks like any other sensor stream. This ensures that all data, whether live or batch, is captured in a consistent, usable format. Second, it makes data searchable. Obscure tags such as “T_SetPnt_03” are mapped to clear terms in a shared battery manufacturing ontology. Engineers can simply search for “Dryer bottom temperature”, and fetch relevant data from any plant. Finally, it adds context. The wrapper stores extra clues (i.e. metadata), such as machine function, subcomponents, and physical dimensions like dryer length, so this ensures that the data arrive with meaning, not in isolation. This enables smarter analysis and better decisions.

How we build a Wrapper

Creating a wrapper is a collaborative process. At first, a machine expert begins by listing the key parameters that define the machine’s operation. Then, a semantic expert hooks these terms into existing ontologies, or extends them when battery production additional descriptive terms. The output is a handful of JSON/CSV files, ready for deployment. Each wrapper runs in its own lightweight container, and visual dashboards are layered on top to provide partners with the information they need for process tuning and optimisation.

Test drive: The Mathis Slot Die Coater

To test the work, SINTEF piloted the method on a Roll-to-Roll coater machine that lays down 200mm electrode films. From hundreds of available signals, the Team picked the vital few set points (SP) like web speed, and global readings such as dryer temperature. These signals were split into “control” and “state” groups, and mapped to EMMO compliant terms. As a result, this pilot revealed gaps in the domain ontology, so SINTEF provided to its expansion to support future developments.

Why it matters

Wrappers allow both online and offline machines to speak the same digital language. This makes data traceable and reusable, opening the door to automated recipe tuning and, ultimately, faster ramp up of greener European battery lines. In the next steps, the SINTEF Team will feed offline test results back into the loop, closing the circle between lab insights and factory control.
Standardising machine parameters using semantic technologies is essential for enabling consistent comparison of results across manufacturers and our software (equipment wrapper), makes this process easier by increasing usability, helping us achieve exactly that.

Fig.1 SINTEF Machinery Wrapper structure