A manufacturer came to us with a familiar problem: thirty machines on the floor, a rising energy bill, and no real visibility into where the time and the kilowatt-hours were going. Production knew the lines “felt busy”, but nobody could prove how busy, or where the losses hid. Here is how we built a SCADA energy-monitoring and OEE system that turned that fog into numbers.
⚡ The challenge
Thirty machines, several controller brands, and energy data that existed only on local meters nobody read. Management wanted three things: live energy consumption per machine, a permanent history they could analyse, and an honest OEE figure to find the real bottlenecks — not gut feeling.
🖥️ The SCADA solution
We deployed a SCADA layer connected to all thirty machines, reading from the existing PLCs and energy meters over the plant network. Each machine reports its state (running, idle, stopped, fault), its cycle counts and its live power draw. Operators see a real-time overview on the floor; supervisors see the whole plant from any browser.
- Live power (kW) and consumption (kWh) per machine and per line
- Machine states captured continuously — no manual logging
- One dashboard for thirty machines across mixed controller brands
🗄️ SQL historian — the data that stays
Every signal is logged to an MS SQL database, so nothing is lost when a shift ends. That historian is the backbone: it lets the plant compare this week to last month, this machine to its twin, this shift to the night shift — with real timestamps, not estimates. Historical trends, energy reports and downtime logs all come from the same single source of truth.
📊 OEE & piece-time calculations
On top of the historian we built the metrics that actually drive decisions:
- OEE — Availability × Performance × Quality, calculated per machine, line and shift
- Piece time / cycle time — actual time per part versus the target, so slow stations stand out
- Downtime analysis — stops categorised and ranked, so the biggest losses are obvious
- Energy per part — consumption tied to output, not just to the clock
Suddenly “the line feels slow” became “station 7 loses 9 minutes per shift to micro-stops” — the kind of statement you can act on.
✅ The result
The plant gained a single, honest picture of how thirty machines actually perform — live on the floor and in history for analysis. Energy is now measured per machine and per part, OEE is tracked rather than guessed, and the worst bottlenecks surfaced in the first weeks. The data pays for itself by pointing maintenance and improvement effort exactly where the losses are.
If you are running machines you can’t really see — energy, OEE, downtime — we can build the same kind of visibility on top of the equipment you already own.