Cross-department Facilitation
A recurring leak solved by linking
material properties with assembly conditions.
When multiple departments are involved, investigations quickly lose structure and direction. A hydraulic seal failure had divided the engineering, quality, and procurement teams for months. Nobody owned the problem. Causetec gave them a shared process.
The situation
Three departments. Three different explanations. No resolution.
A manufacturer of industrial hydraulic assemblies had been dealing with field returns for the same failure mode for seven months: hydraulic fluid leaking from the main cylinder seal on units returned from customers in the 8–14 week usage window. The failure rate was 3.2% of shipped units well above the 0.5% warranty threshold and generating significant field service costs.
The engineering team believed the seal specification was correct and pointed to assembly as the cause. The assembly team had followed the SOP and pointed to the seal material. Procurement had received no quality alerts from the seal supplier and pointed back to engineering. Each department had a plausible explanation that placed the problem somewhere else. Seven months later, nothing had changed.
A cross-functional review meeting had been scheduled four times and cancelled twice. When it did run, it produced no agreed conclusion and no assigned actions. The investigation had no facilitator, no shared framework, and no method for evaluating competing hypotheses objectively.
Define
The failure had a consistent profile that nobody had documented together.
A quality engineer was tasked with facilitating a structured investigation using Causetec. Before the first cross-department session, she worked through the Data Wizard to build a precise problem definition. The leak was occurring exclusively at the upper seal land on the primary cylinder rod not at the lower seal, not at the port connections. Failures were clustered in units from production weeks 31 through 38. Units produced before and after that window showed normal warranty rates.
That production window immediately narrowed the investigation to a specific period. Something had changed or something had been different in weeks 31 to 38. The problem definition alone gave the cross-department session a shared starting point that no previous meeting had established.
Identify
The problem solver gave everyone a structured space to contribute without blame.
The 6M Ishikawa session was run as a facilitated workshop with representatives from engineering, assembly, and procurement. Because the framework was neutral and causes linked to relevant facts, participants contributed causes without the defensiveness that had characterised previous discussions.
Material surfaced three candidates: seal durometer variance between batches, seal storage temperature compliance, and dimensional tolerance on the seal OD. Method surfaced two: rod surface finish spec at assembly, and torque sequence for the gland nut. Machine surfaced one that had not been raised before: assembly press ram parallelism drift the possibility that the press used to seat the seal had developed a misalignment, causing uneven seating load across the seal face.
For the first time, all three departments were working from the same list of candidates. The question of who was "to blame" became secondary to the question of which cause was most likely.
Evaluate
Two causes were linked. Together, they explained everything.
AI likelihood scoring ranked two candidates as high probability: seal OD dimensional variance and assembly press ram parallelism. The team pulled the incoming inspection records for seal batches received during weeks 31 to 38. Three batches from the same supplier production run showed OD measurements at the upper end of the tolerance band still within spec, but at the limit.
The assembly team checked the press ram. Parallelism had drifted 0.18mm from centre within the machine's service interval but sufficient to generate a non-uniform seating load. Neither factor alone would have caused the failure. Together, they did. A seal seated at maximum OD tolerance, seated by a press with 0.18mm ram drift, produced a seal face that was slightly cocked creating a micro-leak path that only opened under sustained hydraulic pressure after several weeks of operation.
Root cause confirmed: interaction between upper-tolerance seal OD and press ram parallelism drift during the weeks 31–38 production window.
Resolve
Four actions. Three department owners. All closed within two weeks.
The press ram was realigned and a monthly parallelism check added to the PM schedule. The seal incoming inspection procedure was tightened to flag batches where more than 10% of measured ODs exceed the midpoint tolerance, triggering a hold and review before release to assembly. Engineering revised the seal seating SOP to include a post-press visual inspection of seal face contact. Procurement raised a supplier quality notification requesting tighter process control on OD variation.
The results
Rapidly solved investigations without backlog, without blame.
The recurring seal failure has not reappeared in the four months since the corrective actions were closed. Warranty return rate on the affected assembly has dropped from 3.2% to 0.3% below the 0.5% threshold for the first time in over a year. The annual warranty cost saving is estimated at €180,000.
The quality engineer who facilitated the investigation has since used Causetec to run two further cross-department investigations at the same facility. Both were resolved within a week. The operations director has referenced the structured approach as the key difference not in the technical skill of the teams, but in having a neutral shared framework that kept the investigation moving and kept departments aligned.
AI-guided Root Cause Analysis for quality and operations teams. Structured investigations, consistent quality, fewer repeat incidents.
Causetec OY 2026© | All rights reserved.