Production Downtime Analysis
Repeated line stops caused by a
worn guide roller nobody had checked.
When downtime keeps coming back, it signals a loss of control over the process. A machine parts supplier was losing 4–6 hours of production per week to the same recurring stoppage. Causetec found the cause in a single shift.
The situation
The line was stopping two or three times a week. Nobody could explain it.
An machine parts supplier was running a high-output press line producing stamped steel components. For eleven weeks, the line had been experiencing unplanned stoppages on average two to three times per week, each lasting 90 minutes to two hours. Total lost production: roughly 4–6 hours per week.
Maintenance had replaced the press tooling twice, recalibrated the feed system, and run a full inspection of the hydraulic circuit. Each time, the line ran cleanly for three to five days before the stoppages returned. The maintenance manager had escalated to the plant director. A customer delivery schedule was already at risk.
The team was firefighting responding to each event as it happened, resetting the line, and moving on. Nobody had stepped back to investigate the pattern. Nobody had mapped all the possible causes. The same three or four hypotheses kept circulating at every shift handover meeting, and none of them had ever been properly validated.
Define
The stoppages had a pattern - it just hadn't been written down.
The production engineer opened Causetec and worked through the Data Wizard. The structured questions forced the team to document what they actually knew rather than what they assumed. The stoppages were occurring on press station 3 only. They happened most frequently during the second half of a production run, rarely at the start. The fault code logged each time was the same: vibration threshold exceeded emergency stop triggered.
This was the first time those three facts had been written down together. Station-specific, run-duration-dependent, vibration-triggered. That was enough to change the entire direction of the investigation.
Identify
The investigation pointed to parts of the machine nobody had examined.
The team had previously focused almost entirely on tooling, hydraulics, and feed calibration. The AI sparring parner surfaced additional candidates the team had not considered: guide roller wear state, roller bearing lubrication interval, mounting bracket torque values, and vibration sensor calibration status.
Critically, it also surfaced a Mother-Nature candidate: thermal expansion of the press frame during extended runs. At operating temperature, the frame expanded by a measurable amount. If guide roller clearances were already at the upper tolerance limit, thermal expansion pushed them beyond it, generating vibration peaks that tripped the emergency stop sensor. This only occurred after the line had been running for 60–90 minutes, which explained the run-duration dependency perfectly.
Evaluate
Fact based likelihood evaluation confirmed what a physical inspection then proved.
Likelihood scoring ranked two candidates as high probability: guide roller wear state and thermal frame expansion. The maintenance engineer pulled the guide roller on station 3 for inspection. The roller had worn unevenly the contact surface on the drive side was 0.4mm below nominal. Within specification at cold start, but once the press frame reached operating temperature and expanded, the effective clearance exceeded the vibration threshold.
The worn roller had passed every standard maintenance check because those checks were performed cold, at the start of a shift exactly when the problem never occurred. It had been in this condition for an estimated six to eight weeks. Eleven weeks of stoppages. The cause had been present the entire time.
Resolve
Guide roller replaced. Maintenance schedule revised. Line running clean.
The worn guide roller on station 3 was replaced immediately. A secondary check of stations 1, 2, and 4 found one further roller approaching the wear limit replaced as a precaution. The maintenance SOP was updated to include a warm-state inspection of all guide rollers at operating temperature, not just cold-state checks at shift start. Roller wear was added to the scheduled PM checklist with a tighter inspection interval of every 500 production hours.
The results
Improve your OEE by eliminating the downtime you keep accepting as normal.
The line has run without a single unplanned stoppage on station 3 in the three months since the corrective actions were closed. The 4–6 hours of weekly lost production have been fully recovered.
The customer delivery risk was resolved before any shipment was affected. The plant director has since mandated that all recurring stoppages across the facility defined as any fault occurring more than twice in a rolling 30-day period are formally investigated using Causetec before a maintenance work order is raised.
AI-guided Root Cause Analysis for quality and operations teams. Structured investigations, consistent quality, fewer repeat incidents.
Causetec OY 2026© | All rights reserved.