Retrofit
Machine data without a PLC swap
Legacy machines deliver over Modbus, classic control protocols or added sensors. We lift that to OPC UA level without you touching a controller.
Follow-up · Integration · OPC UA
We connect your machines, model the data points and hand them to ERP, MES, time-series database or AI — vendor-independently and without replacing the controller. Delivered by a maintainer of the most used open-source OPC UA libraries for Node-RED.
Request connectivityRetrofit
Legacy machines deliver over Modbus, classic control protocols or added sensors. We lift that to OPC UA level without you touching a controller.
Servers
Your machine already ships an OPC UA server — but the address space is a data graveyard. We turn it into an information model your systems can work with.
Bridges
Where IT systems expect MQTT and the shop floor speaks OPC UA, we build the bridge — with semantics preserved rather than a flat topic tree.
Modelling
We model to ISA-95 and, where they exist, to your industry's companion specifications. That way your structure survives the next system change.
OPC UA is not a protocol you “switch on”. The difference between working connectivity and a data graveyard is the information model — and that is craft, not a configuration dialog.
Klaus Landsdorf has worked in the OPC Foundation working groups since 2013 and has maintained one of the most used open-source OPC UA libraries for Node-RED since 2017. The standards we deploy at your site are written where we contribute.
For you that means: no learning curve on your invoice, no vendor lock-in, and a model that still holds when the next machine comes from a different manufacturer.
In most cases yes. Where no interface exists we capture over Modbus, classic control protocols or, if need be, retrofitted sensors at the cabinet. Send us a list of controller types and we will tell you what works without intervention.
No. That is the point: connectivity happens alongside, through an edge gateway such as the SWIFT-BOX. The controller stays unchanged and the machine acceptance stays valid.
Yes, in clearly bounded roles: integration, protocol bridges and data preparation. We use it where it is the fastest and most maintainable answer, and containerise operation with Docker. Where a controller has to regulate deterministically, control belongs in the controller — not in Node-RED.
You, if you want to — we document the model and train your team on it. If you would rather not, we take over operation and evolution.
We recommend it. One machine, one clean model, one measurable result — after that the rest can be estimated reliably. Name the machine and we will quote for it.
An e-mail address is enough. We come back once with a short read on it — no newsletter.
Controller type, existing interfaces and the values you need — that is all it takes for a first assessment.
Request connectivity