Skip to content
FerruleTech

Insights

Engineering notes, not thought leadership

Each of these answers one question properly — the kind an automation engineer actually searches for at eleven at night. Written by the people who do the work, and published whether or not it leads to an enquiry.

In the pipeline

Coming next

We would rather show the plan than pad this page with posts that were not written by an engineer. Two per month. If one of these is a question you need answered now, ask us directly and we will answer it before it is published.

  • How to connect an existing PLC to a cloud dashboard without changing the control program

    The non-intrusive path from an installed PLC to a live dashboard: which protocols to read, how to size poll rates, and where the genuine limits are.

  • What OEE actually measures, and why most factory OEE numbers are wrong

    Availability, performance and quality — where the assumptions hide, and how to produce an OEE figure that survives a review meeting.

  • Retrofitting machine monitoring onto equipment with no PLC

    Current transformers, proximity counters and remote I/O: getting real state and real counts from machines that expose nothing.

  • MQTT Sparkplug B explained for automation engineers

    Why plain MQTT is not enough, what birth and death certificates give you, and how state awareness changes alarm design.

  • Edge vs cloud: what should be computed at the plant

    A decision framework based on latency, bandwidth, availability and cost — with worked examples from vibration and OEE.

  • A realistic architecture for monitoring 20 machines across two plants

    A full worked architecture: gateways, protocols, tag model, storage, retention and what it costs to run.

  • Why your SCADA data never reaches your ERP

    The boundary problem, the person in the middle with a spreadsheet, and how to design the interface properly.

  • Reading energy consumption per machine without rewiring the panel

    Clamp-on CTs, meter selection, where to install, and how to attribute consumption to machine, shift and product.

  • Downtime reason codes: designing a system operators will actually use

    Why most reason-code systems are abandoned within a month, and the design constraints that prevent it.

  • Migrating a legacy PLC without a full line shutdown

    Proving functional equivalence, pre-testing off-line, staged cut-over, and keeping a real way back.

  • Industrial network segmentation basics — the Purdue model in practice

    What the levels mean on a real shop floor, where the boundary device sits, and what your IT team will ask for.

Ask the question directly

An engineer will answer it, whether or not there is a project in it.

CallWhatsAppEstimate