Autoneum reached optimal acoustic performance for an advanced material using structured data and a repeatable workflow

Why Acoustic Material Selection Is a Data Problem

Autoneum is the global market and technology leader in acoustic and thermal management solutions for vehicles, developing innovative lightweight components for light vehicle OEMs worldwide. Selecting the right material for a high-performance automotive component requires evaluating acoustic performance across multiple properties simultaneously - including diffuse field absorption, a critical KPI that determines how effectively a material manages sound energy under real-world conditions.

‍

For a hood panel development programme, the available data came from time-intensive physical testing but contained gaps: missing values for diffuse field absorption made the dataset incomplete. Without those values, analysis could not proceed and product decisions were delayed. Re-running physical tests to fill the gaps was both costly and impractical within the programme timeline. The team needed to work with the data they had - structure it, identify gaps explicitly, and build a model that could predict missing acoustic property values from the available evidence.

‍

An Automated Workflow That Predicts What Testing Cannot Provide

Autoneum partnered with Key Ward to build an automated workflow that ingested and standardised the incomplete foam measurement data, learned the relationship between known material properties and absorption values, identified data quality gaps, trained and optimised machine learning models on the available measurements, and predicted the missing diffuse field absorption values. The workflow automatically generated a fully auditable report - with every prediction traceable, validated against the known data range, and flagged if inputs were incomplete or anomalous.

‍

What had been a fragmented, manual process became a repeatable and scalable engineering workflow. The team can now identify optimal materials in minutes rather than days, instantly predict acoustic KPIs, reuse the same workflow intelligence to fast-track subsequent studies, and accelerate RFQ response time for new material programmes.

‍

What the Full Case Study Documents

The full case study documents the complete workflow architecture - data ingestion, ML model training and selection, built-in safeguards for prediction validation, and the auditable reporting structure. The headline result: predictions cut from days to minutes compared to conventional testing and solver simulation. If your engineering team is working with incomplete simulation or test datasets blocking material selection decisions, the approach documented here is directly applicable.