Why Does Digitalisation Matter?
- OEE MEASUREMENT — put a number on your production efficiency (an hourly percentage for every machine)
- DOWNTIME ANALYSIS — which machine stands idle, for how long, and why?
- MAINTENANCE PREDICTION — detection from sensor data before a breakdown
- PRODUCTION TRACKING — the order → workshop → delivery chain becomes visible
- QUALITY TRACKING — the defect rate plus root cause analysis
- OPERATOR PERFORMANCE — comparison on a shift-by-shift basis
- SPARE PARTS MANAGEMENT — the consumption trend captured automatically
CNC Data Logging — What Is Collected?
- MACHINE STATE — on / off / idle / running / fault
- RUNNING HOURS — spindle running time plus idle waiting time
- ALARM RECORDS — code plus time plus frequency
- PRODUCTION COUNT — finished parts (program completions)
- PROGRAM EXECUTION — which G-code, and for how long
- TOOL USAGE — which tool, and for how long (with an ATC)
- TEMPERATURE AND VIBRATION — spindle and axis sensors
- POWER CONSUMPTION — instantaneous kVA plus the daily total
OEE — Overall Equipment Effectiveness
OEE is the gold standard of industrial efficiency measurement. It is the product of 3 factors:
- AVAILABILITY — the time the machine was on / the time planned
- PERFORMANCE — the actual production rate / the theoretical maximum rate
- QUALITY — accepted parts / total parts produced
- OEE = Availability × Performance × Quality
- The world benchmark: above 85% is excellent, 60% is average, below 40% is poor
A Worked Example
Take a CNC bridge saw — 8 hours were planned, it was on for 7 hours (7/8 = 87.5% availability), it actually cut for 5 hours (5/7 = 71% performance), and of 10 parts produced 9 were accepted (9/10 = 90% quality). OEE = 0.875 × 0.71 × 0.90 = 56%. There is room for improvement.
Data Collection — The Hardware
- THE CNC OUTPUT — modern CNC controls (TEX, Fagor, Beckhoff) provide a standard Ethernet/OPC output
- IoT SENSORS — can be retrofitted to an older machine (vibration, temperature, current)
- BARCODE/RFID — work order plus product tracking
- AN OPERATOR PANEL — a tablet or touchscreen (start of shift plus entry of downtime reasons)
- A GATEWAY — transfers all the data to a cloud or local server
The Software Ecosystem
- MES (manufacturing execution system) — workshop management software
- ERP — orders plus stock plus invoicing, integrated
- SCADA — visualisation of CNC and IoT data
- BI (business intelligence) — reporting with Power BI or Tableau
- CLOUD-BASED — modern solutions (PTC ThingWorx, Siemens MindSphere)
- Türkiye: Logo, Mikro, Eta — local ERP integration
Predictive Maintenance
Detecting from sensor data that a machine is "becoming unwell" before it breaks down:
- VIBRATION SENSOR — spindle bearing wear → vibration rises
- TEMPERATURE SENSOR — a bearing in distress → the temperature trend moves upwards
- CURRENT SENSOR — a motor under strain → the current fluctuates
- OIL ANALYSIS — periodic laboratory checks (metallic particles)
- AI/ML — pattern detection from historical failure data
- The result: 40-60% of breakdowns are detected in advance and dealt with through planned maintenance
Digitalising the Workshop — A Phased Approach
Phase 1 — Data Collection
- Setting up the Ethernet connection on the CNC control
- An OPC server — to receive the data
- A simple dashboard — visualising on / running / stopped
- A 6-12 month period — building the habit plus basic measurement
Phase 2 — Analysis
- OEE calculation plus reporting
- Downtime analysis — the reasons categorised
- An operator panel — the operator enters the downtime reasons
- Monthly improvement meetings
- A 12-18 month period
Phase 3 — Advanced
- IoT sensor retrofit
- Predictive maintenance (the critical machine first)
- Full ERP integration — order → workshop → delivery
- A customer portal — seeing the status of an order live
- AI/ML projects
- A 2-3 year period
The Approach to Investment
- TO START — Excel plus manual records (free, but not sustainable)
- MID-RANGE — an off-the-shelf SaaS solution (an annual subscription), for a small workshop
- LARGE — a bespoke ERP plus MES plus SCADA installation (a large workshop)
- CLOUD — low capital outlay, subscription-based, practical for most workshops
- EU grants and the digital transformation support programmes in Türkiye are worth following
Practical First Steps in an SME Workshop
- Check the Ethernet output on the CNC machine
- A simple OEE spreadsheet template — the operator fills it in at the end of the shift
- A monthly downtime analysis report
- After 6 months — visualisation with simple BI software
- After 1 year — the decision on a sensor retrofit
- Activation of remote diagnostics with the authorised service provider