Logo of Acerta
Contact us

Case Study

Solving multi-plant quality issues at Dana

Objectives

Improve NVH performance and reduce scrap by enabling Toledo and Fort Wayne teams to use shared data to pinpoint and resolve the true sources of a recurring quality issue.

Challenge

A recurring subcomponent assembly failure at Toledo could not be isolated locally, requiring visibility into upstream production data from Fort Wayne.

Key Results

  • 8% throughput increase on final assembly
  • 98% FTT on subcomponent assembly

8%

Throughput increase

98%

Sub-assembly FTT

8%

Throughput increase

98%

Sub-assembly FTT

Background

Dana operates a network of manufacturing plants that depend on each other for high-precision components. When Dana’s Toledo driveline facility experienced severe swings in throughput at their backlash validation stations, sometimes dropping FTT by 30-40%, local teams exhausted all internal checks. Despite validating test machines, confirming sub-component inputs, and stabilizing assembly processes, the issue kept returning. To resolve the issue, the Corporate Senior Manufacturing Data Manager engaged with local teams at Toledo and their sub-component feeder plant, Dana’s Fort Wayne Manufacturing Center. He then used LinePulse to analyze data spanning the operations at both facilities. This multi-plant visibility enabled the teams to trace the root causes upstream, eliminating the quality variations and stabilizing operations at the Toledo plant.

Problem

Unpredictable NVH failures at Toledo

The Toledo plant saw NVH results that fluctuated dramatically over time, a pattern described as “wave-like.” FTT could swing from 85–90% down to nearly 50%, with no single model, or line, consistently responsible. Internal checks confirmed that inputs such as bearing preload(s) and gear contact pattern were stable. Internal checks also verified that NVH and gear backlash results were highly correlated. After further verification with LinePulse of all upstream stations to backlash, the team was able to say definitively that this issue was not being created at Toledo.

Gaps in upstream visibility

The Toledo teams lacked real-time visibility into Fort Wayne’s gear-level data, even though gear machining is a major contributor to backlash and NVH performance. Fort Wayne produced rich gear test data, but it had never been digitized or analyzed at scale. Without a unified data layer, neither plant could easily compare processes, detect tolerance drift, or correlate component behavior to final NVH fallout.

Solution

1. Digitalizing Fort Wayne gear data for cross-plant access

The corporate data leader worked with Fort Wayne to unlock years of untapped diagnostic information from legacy Hypoid gear testers. These files were digitized and fed into LinePulse as structured, searchable process data.

This allowed Toledo engineers to directly compare upstream machining behavior with downstream NVH outcomes for the first time.

2. Corporate-led root cause analysis using full birth history

Once data was streaming into LinePulse, the corporate manufacturing data manager used the platform to trace failed NVH results back through center-section assembly data at Toledo and all the way upstream to the individual gear machining signals at Fort Wayne. This cross-plant genealogy allowed the team to identify precisely which gear machining features were contributing to NVH spikes and implement new controls at Fort Wayne.

3. Shared run charts, limits, and monitors across both plants

With the leading indicators of problems identified, both plants deployed live run charts and in-platform process monitors to detect when Fort Wayne processes drifted toward print tolerance limits. Single-variant monitors and SPC-style alerts flagged changes within minutes instead of weeks, enabling immediate intervention before parts were shipped.

Results

The cross-functional engineering team in the Toledo plant and at Dana corporate eliminated the recurring subcomponent volatility by stabilizing the gear machining process and making upstream variation fully visible downstream.

  • Center section sub-assembly throughput FTT increased from a median rate of 80% to a daily average of 98%, with only isolated single-digit failures instead of hundreds per wave.
  • Final axle assembly throughput increased by 8%.
  • Toledo no longer experienced waves of failures, reducing rework, weekend shifts, and operational stress.
  • Both plants now use shared dashboards to compare performance, trace genealogy, and spot cross-plant issues in minutes rather than months.

This case demonstrates how unified analytics and multi-plant visibility allowed Dana to solve a complex, intermittent quality issue that no single plant could diagnose alone.

It’s pure impact to the Dana bottom line.

“Toledo producing closer to 98% or 99% first-time-through—it’s pure impact to the Dana bottom line. We’ve reduced scrap rates and cut the number of components being thrown out or recycled.”
Brian Longartner
Senior Manager, Manufacturing Data Science

Case Studies

Proven Impact