How Epiroc Reduced Product Returns by 30% and Accelerated Manufacturing Innovation with AI

By: GoBeyond Team

July 3, 2025

3 min read

AI case study for Epiroc – 30% fewer product returns with Azure ML

Quick Overview

Epiroc used Microsoft Azure Machine Learning and AI governance software to automate and optimize steel manufacturing processes, ensuring consistent quality and reducing customer rejections and product returns.

Epiroc

Company Size

~14,000 employees (estimated)

Revenue Range

$4B-$5B annual revenue (estimated)

Primary Challenge

Fragmented data and inconsistent manufacturing processes causing quality issues and product returns

Key Metrics

\- 30% reduction in product returns and customer rejections\- Created an “AI Factory” in 60 hours\- Developed ML models for heat treatment process in 6 weeks\- 11 analytical teams using AI models globally

The Problem

Manufacturing facilities worldwide had siloed data and lacked a unified system to share best practices and optimize complex steel heat treatment processes, impacting product quality and efficiency.

The Solution

Implemented Azure Machine Learning, Azure Data Factory, Azure Databricks, and AI governance software from Sogeti to create a scalable “AI Factory” platform for data sharing, process automation, and predictive modeling.

Results

\- Consistent steel quality with better control over density, hardness, and flexibility\- Reduced waste and improved operational efficiency\- Faster innovation cycles and model deployment\- Enhanced compliance with regulations