How a Tier-1 Auto Supplier Cut Production Cycle Time by 23% in Six Months Using Process Mining

When a tier-1 automotive supplier's operations team discovered they were spending 18 percent of shift time on non-value activities, nobody wanted to believe it. The plants ran lean. The metrics looked tight. The problem was invisible because it lived in the gaps between systems, the handoffs between departments, the waiting periods that ERP logs never flag as waste. Process mining found it. In six months, the company cut production cycle time by 23 percent and recovered enough capacity to handle a $180 million new contract without capital expenditure. The stock moved 3.8 percent on the earnings call. The market understood the math immediately.

Challenge

The company manufactures transmission housings and electronic control modules for three major OEMs. Like most suppliers, it operates under margin compression that would make you wince: automotive contracts locked in at 2 to 4 percent net margin, with annual cost reduction requirements of 3 to 5 percent. Efficiency is not a nice-to-have. It is an existential requirement.

The plants had already gone through the obvious cuts. Lean consultants had been through twice. Six Sigma projects had eliminated textbook waste. Yet when management looked at lead times, throughput, and first-pass quality data, something did not add up. The time from order entry to shipment clock showed 47 days. The sum of all documented process steps showed 28 days. That 19-day gap lived in a blind spot: unofficial wait times, undocumented rework loops, bottleneck shifts between departments that nobody measured because they happened off the formal production schedule.

A traditional root cause analysis would have taken months and required hiring expensive consulting. Instead, the operations director green-lit a process mining pilot at the largest plant: 680 people, 42 production lines, roughly 120,000 SKUs per month.

Solution

Process mining software ingested logs from the MES, ERP, quality system, and scheduling software. The tool reconstructed every production instance: not the textbook workflow, but the actual workflow. What emerged was a map of reality.

The bottleneck was not where the gemba walks had assumed. A particular subassembly process triggered exception handling 34 percent of the time. When dimensions were marginal, parts bounced back for rework instead of moving forward. The rework path was off-system; shop floor workers handled it manually. A second pattern showed that shift changes correlated with a 2.4-hour average delay in line startups. A third revealed that one particular customer order type required manual quote-to-order conversion because an ERP configuration nobody remembered how to change forced workarounds.

None of these problems was hidden from the people doing the work. The difference was that process mining made them visible to the people making budget decisions. Quantified. Monetized. Impossible to rationalize as "just how things work."

The company addressed the three largest anomalies. It recalibrated the subassembly tolerance stack and tightened incoming inspection to reduce exceptions. It automated shift changeover checklist verification through the MES. It fixed the ERP configuration (a two-hour IT ticket, apparently, once somebody prioritized it). The tail of smaller inefficiencies got scheduled into a continuous improvement roadmap.

Results

Six months after the pilot launch, cycle time dropped 23 percent. Weekly capacity recovered amounted to 340 hours across the plant fleet. Rework rate fell from 4.2 percent to 2.8 percent. Quality improvement was a side effect, not the goal, but the company's PPM score tightened enough to unlock a price increase conversation with one major customer.

The company deployed process mining across all 12 manufacturing locations. The software operates continuously now, flagging new anomalies as they emerge. The second wave of optimization is targeting dock-to-dock time and supplier quality variance.

Wall Street valued this as a structural margin expansion. The $180 million new contract lands at higher utilization levels than the old run rate would have allowed. Gross margin on the incremental volume runs at 5.2 percent. That is not a miracle; that is what happens when you stop leaving money on the floor.

If you run a plant and you are not logging what your systems actually do, you are flying blind. Your competitors are not anymore.