Why there's no single price

"How much does OEE cost" is the most common question we get, and the honest answer is that it depends almost entirely on what you're measuring and what's already on the line. A single, modern packaging line with a PLC that already speaks OPC-UA can be reading live OEE in days for a fraction of what it costs to instrument six ageing lines that have no controller at all. So instead of a headline figure, it helps to understand the cost drivers — then you can place your own plant on the scale.

Rule of thumb: price OEE per line, and start with your single worst bottleneck. The line that loses you the most production is also the one where the system pays back fastest — so it's both the cheapest place to start and the easiest to justify.

What actually drives the cost

  • How you capture machine state. If the line has a PLC exposing run/stop, counts and reject signals (OPC-UA, Modbus), we read those directly and the hardware cost is low. If it doesn't, we add a small edge device with sensors — a current sensor on the motor, a proximity or photo-eye for counts, a stack-light tap for state. That's the single biggest swing in price.
  • Number of lines. Cost scales roughly per line, not per factory. Ten lines isn't ten times harder than one, but it's not free either — each needs its own pick-ups and configuration.
  • Downtime reason capture. Counting good vs bad parts is cheap. Capturing why a line stopped — via an operator touchscreen for downtime reason codes — adds a terminal per line but is where most of the improvement actually comes from.
  • Connectivity. A line on factory Wi-Fi or wired Ethernet is cheap to connect. A remote or isolated site needing LoRaWAN, NB-IoT or 4G adds gateway and airtime cost (see our connectivity guide).
  • Integration depth. A standalone OEE dashboard is one price. Feeding OEE into your ERP/MES, or pulling production orders back the other way, is engineering time on top.

Indicative ZAR ranges

addanode.com publishes no fixed price list, so the bands below are indicative and exist to help you budget — not to quote you. Real numbers come from a scoped walk of your lines.

  • Single modern line (PLC already exposes data): roughly R40,000–R70,000 once-off to connect, configure dashboards and train, plus a monthly platform fee.
  • Single old line (no usable PLC, add sensors + operator terminal): roughly R70,000–R120,000 once-off, because you're adding the signal capture the line never had.
  • Multi-line rollout: per-line cost drops as you standardise hardware and config across similar lines — the first line carries the setup learning; lines two onward are cheaper.
  • Ongoing: a monthly platform and support fee covering hosting, updates and support, sized to the number of lines.

The point of these bands is the order of magnitude: an OEE pilot is a five-figure decision per line, not a six- or seven-figure ERP-style project. That matters, because it means you can prove it on one line before committing to a plant-wide rollout.

What inflates a quote (and how to avoid it)

Most of the times we see an OEE quote balloon, it's one of these:

  • Instrumenting everything at once. You don't need every line on day one. Start with the bottleneck; the gain there funds the rest.
  • Rip-and-replace controllers. Some vendors price in a new PLC. You rarely need one — a good edge device reads the signals an old machine already produces. We deliberately design around your existing equipment.
  • Per-seat or per-tag software licensing. Licensing models that charge per user or per data tag punish you for using the system. Look for flat, line-based pricing.
  • Cloud-only architectures that fall over during load shedding — then need extra spend to make resilient. Edge buffering should be in from the start, not a bolt-on.

How to think about payback

The cost only makes sense next to what poor OEE is already costing you. Take your bottleneck line: estimate the value of one hour of lost production (the margin on the output it would have made). Most OEE programmes recover several percentage points of availability in the first quarter simply by making downtime visible and reason-coded — losses that were previously invisible suddenly have an owner. On a line where an hour of downtime is worth tens of thousands of rand, recovering even a few hours a month covers the entire system in well under a year.

That's the real answer to "how much does it cost": less than what the bottleneck is quietly losing you every month it stays invisible.

This is exactly how we scope OEE management at addanode — per line, decision-led, and built to read the equipment you already own. It runs on the same addaNet platform as your condition and energy data, so production, reliability and cost sit in one picture. And because we build both the hardware and software in-house, we can instrument an old line without forcing a controller upgrade you don't need.

Frequently asked questions

How much does an OEE system cost per line in South Africa?

As an indicative guide, roughly R40,000–R70,000 once-off for a modern line whose PLC already exposes data, and R70,000–R120,000 for an old line that needs sensors and an operator terminal added — plus a monthly platform and support fee. These are indicative ZAR bands; a scoped quote depends on what's on your lines.

Do I need to buy a new PLC to measure OEE?

Almost never. A small edge device can read the signals an old machine already produces — motor current, a count sensor, a stack-light tap — so you measure OEE without a controller upgrade. We design around your existing equipment on purpose.

Is there a monthly fee, or just an upfront cost?

Both, typically. There's a once-off cost to instrument and configure each line, then a modest monthly platform and support fee covering hosting, updates and support. Avoid models that charge per user or per data tag — flat, line-based pricing scales far better.

How quickly does an OEE system pay back?

Usually within months, not years. Most programmes recover several points of availability in the first quarter just by making downtime visible and reason-coded. On a bottleneck line where an hour of lost production is worth tens of thousands of rand, recovering a few hours a month covers the system well inside a year.

Can I start with just one line?

Yes — and you should. Start with your worst bottleneck: it's the cheapest place to begin and the easiest to justify, because it loses you the most production. Prove the gain there, then standardise the same setup across your other lines.