Key Takeaways: OEE data quality depends entirely on how it's collected. Manual entry produces numbers that are 10–20 percentage points more optimistic than reality. PLC-based automatic collection produces accurate, real-time data that reveals the true production performance. Fabrico connects to virtually every PLC brand in manufacturing — Siemens, Allen-Bradley, Fanuc, Mitsubishi, Beckhoff, Omron — through OPC-UA, Modbus, and proprietary protocols, plus IoT sensor options for legacy machines without digital outputs.
The most common OEE implementation question is not about software features or pricing — it's about machine connectivity. "Can your platform connect to our Siemens S7-300s?" "We have a mix of old and new machines — how do you handle that?" "Our oldest press doesn't have any digital outputs at all — what do we do?"
These questions are the right ones to ask, because connectivity quality determines OEE data quality, and OEE data quality determines whether the numbers you see in the dashboard reflect what's actually happening on your production floor.
Manual OEE data entry — operators recording production counts and downtime at the end of each shift — produces numbers that are consistently 10–20 percentage points more optimistic than the actual production performance. Operators report approximate numbers; supervisors round to the nearest 5 minutes; stops under 5 minutes often aren't reported at all. The resulting OEE number provides false comfort and drives improvement investment toward the wrong problems.
Automatic PLC-based OEE data collection eliminates these errors. The machine reports its own state, count, and cycle time through its controller signals. No human estimates, no rounding, no selective reporting. The OEE number reflects what actually happened, not what someone remembers or chooses to report.
Modern PLCs with OPC-UA (post-2010 equipment):
OPC-UA is the industrial standard communication protocol that most modern PLCs support natively. Siemens S7-1500, Allen-Bradley CompactLogix and ControlLogix (with EtherNet/IP), Beckhoff TwinCAT, and many others publish OPC-UA servers that Fabrico reads directly via a standard network connection. No additional hardware, no proprietary software, no license fees for connectivity. This is the cleanest connectivity path — configure the OPC-UA server address and tag names in Fabrico, and real-time machine data begins flowing.
The practical limitation of OPC-UA: the specific OPC-UA server tags that expose machine state, production count, and cycle time vary by PLC brand, model, and application programmer. Fabrico's implementation team has configured hundreds of PLC types and maintains a tag library that covers most standard machine types. For non-standard configurations, a 1–4 hour PLC programming engagement from a qualified OT engineer typically exposes the required signals.
Legacy PLCs without OPC-UA (pre-2010 equipment):
Siemens S7-300 and S7-400 series PLCs use proprietary S7comm protocol rather than OPC-UA. Fabrico supports S7comm connectivity for these legacy Siemens controllers — the same real-time data collection, different protocol. Allen-Bradley ControlLogix and MicroLogix older series use EtherNet/IP or DF1 serial protocols; Fabrico supports both. Fanuc CNC controllers use FOCAS2 protocol for machine data; Fabrico's FOCAS2 adapter provides spindle utilization, program runtime, and production count data from Fanuc CNCs without requiring CNC programming modification.
Machines without digital outputs (pre-digital or purpose-built equipment):
Not every machine has a PLC or digital signal that cleanly indicates run/stop state and production count. For these machines, Fabrico offers three IoT sensor approaches:
Fabrico's edge gateway processes signals from all sensor types and forwards the standardized machine state and production count data to the Fabrico cloud — enabling the same OEE dashboard and analysis for sensor-connected legacy machines as for PLC-connected modern machines.
The connectivity options above handle machines with electronic control systems. A significant portion of manufacturing — particularly in high-mix, labor-intensive assembly operations — involves manual stations, semi-automated stations, and hybrid lines where human operators perform critical steps between automated processes.
PLC connectivity and IoT sensors don't capture what happens at these stations. They detect machine state changes but miss operator handling time, manual inspection steps, and the workflow inefficiencies that accumulate in human-machine interface zones.
Fabrico's computer vision — Inefficiencies Zoom-In — provides connectivity for these environments:
Camera-based production monitoring: Cameras positioned at manual or hybrid stations detect production events (part arrivals, operator interactions, product departures) and calculate actual cycle time per operation. No sensor installation, no electrical work, no machine modification — the camera observes and the computer vision algorithms interpret.
Anomaly detection and micro-stop capture: Beyond cycle time measurement, Fabrico's computer vision detects deviations from normal operating patterns — micro-stops, operator reaching into a machine to clear a jam, parts arriving in wrong orientation, quality inspection delays. These events are captured as video clips, categorized by type, and integrated into the OEE calculation as performance rate losses.
Operator efficiency analysis: For assembly stations where operator technique varies, computer vision captures the specific motions and time allocations that differentiate high-performing operators from average performers — identifying training opportunities and standardization targets that PLC data cannot surface.
The combination of PLC connectivity for automated equipment, IoT sensors for legacy machines, and computer vision for manual stations gives Fabrico full coverage across the entire production floor — every machine, every station, every process type, in a single OEE dashboard.
Connecting production machines to cloud-based OEE software introduces OT network security considerations that IT and OT teams appropriately scrutinize. Fabrico's connectivity architecture is designed to meet these security requirements without compromising the OEE data quality that makes the investment valuable.
Edge gateway architecture: Fabrico uses a purpose-built edge gateway device that sits in the DMZ between the OT network and the IT/internet network. The edge gateway reads machine signals from PLCs via OPC-UA, Modbus, or other protocols on the OT network side. All connections from the edge gateway to Fabrico's cloud platform are outbound-only HTTPS — no inbound connections from the internet to the OT network are required or permitted.
This architecture satisfies the ISA/IEC 62443 requirement for OT network isolation: the OT network boundary is not exposed to internet connectivity, and the edge gateway cannot receive commands from outside that could affect machine control systems.
Data handling: The edge gateway processes raw machine signals into production metrics (machine state, production count, cycle time) before forwarding to Fabrico's cloud. Raw PLC data — which may contain proprietary process parameters — remains on the OT network; only the derived production metrics leave the plant network.
Offline capability: The edge gateway stores up to 72 hours of production data locally and forwards to Fabrico's cloud when connectivity is available. Network interruptions — planned maintenance windows, temporary connectivity outages — don't create OEE data gaps. When connectivity restores, the edge gateway uploads the stored data automatically.
IT and OT security review of Fabrico's connectivity architecture is a standard part of the implementation process. Fabrico provides the network architecture diagrams, data flow documentation, and security specification that IT teams require before approving OT network connectivity. Most security reviews complete within 1–2 weeks for standard manufacturing environments.
Login
Blog
