Relynk started with a specific frustration: the vibration data showing the bearing failure was sitting in a SCADA historian for 11 days. No one had a way to see it.
In 2021, Simen Bakken was heading reliability engineering at a mid-size manufacturer outside Oslo. The plant had 200+ sensors across 14 production lines. Process history lived in three separate SCADA historians — OSIsoft PI on one line, AVEVA Wonderware on another, and GE Proficy on the third. None of them talked to each other. Anomaly review happened weekly, in Excel, by one person.
A bearing failure on a press machine cost 38 hours of unplanned downtime and a missed delivery commitment. The vibration data showing the drift had been sitting in the historian for 11 days. No alert fired. No one looked.
Simen spent six months building the first version of Relynk — not to replace the SCADA systems, but to aggregate their historian streams, run a continuous baseline comparison, and route anomalies to the right person before a drift became a shutdown. The three protocols were Modbus TCP, OPC UA, and direct historian API. That same combination still covers most of Relynk’s deployments today.
Relynk was incorporated in Oslo in 2022 and received seed funding from TRK Group in 2023. The platform is used by operations teams in paper milling, chemical processing, food & beverage continuous production, and water treatment across Europe and North America.
Most industrial monitoring tools are built for data scientists or require months of OT consultant time. We build for operations engineers running SCADA who have 40 assets to watch and a four-hour maintenance window every other Saturday.
Relynk never writes to your SCADA, historian, or PLC. The platform reads via OSIsoft PI Web API, AVEVA Wonderware REST, GE Proficy connector, or direct OPC UA — and only reads. Your OT team reviews one outbound read connection. No new control path.
A 90-day rolling baseline with z-score deviation scoring gives operators a result they can explain to their maintenance manager. Relynk builds those baselines from your existing historian data on day one — no 12-month training run, no data-science team required.
We integrate with SAP PM, IBM Maximo, and Fiix because that’s where the work actually gets scheduled. An anomaly alert that stays in a dashboard doesn’t fix anything. A work order in your CMMS with the sensor context attached does.
The Relynk team has a combined 40+ years in industrial operations, reliability engineering, and industrial software development. Several team members have prior backgrounds at OSIsoft and Inductive Automation. We are not a general-purpose IoT platform that stumbled into manufacturing.
Founded in Oslo
People on the team
Industries served
Customers in Europe and North America