MicrosoftProductionEvidence: Medium50/100

Uddeholm transforms steel manufacturing for sustainability and efficiency

Uddeholm, a historic Swedish steel manufacturer, partnered with CGI to enhance sustainability and profitability using artificial intelligence on Microsoft Azure. By analyzing vast amounts of production data, the company developed AI and machine learning models to predict and prevent the formation of cracks in steel. This enabled earlier detection and improved process efficiency, reducing material waste and energy consumption. Uddeholm established a data governance model and built a central AI-IT infrastructure hub to support ongoing and future data-driven initiatives. Additional AI projects have helped the company with predictive maintenance and real-time detection of leaks, contributing to operational efficiency and reduced costs. The transformation puts data-driven decision-making at the heart of Uddeholm’s operations, aligning business growth with ambitious sustainability objectives. The initiative showcases how collaborative efforts and enabling technology can support industry leaders tackling key environmental and economic challenges. The project featured a comprehensive approach including high-quality data gathering, interdisciplinary information sharing, and dedicated organizational structures such as an AI-IT hub. The partnership continues to identify new areas for AI applications within production and supply chains.

Organization
Uddeholm
Location
Sweden

Reported outcomes

87%

quantified impactRisk, reliability & safety

Strategic outcomes

New product / capabilityPredicted and prevented steel crack formationScale & capacityBuilt a central AI-IT infrastructure hubSpeed & agilityEnabled real-time manufacturing monitoringSustainability & ESGReduced waste and energy consumption
Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Quantified impact: 87%

cgi.comVideo or webinarInferred claimMedium evidence strength

Enabled 87% risk prediction for crack formation, allowing earlier intervention.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
Customer
Uddeholm
Provider
Microsoft
Maturity
Production
Linked source
cgi.com

Additional AI projects have helped the company with predictive maintenance and real-time detection of leaks, contributing to operational efficiency and reduced costs

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 4

  • 1Predictive Quality Control in Steel Production
  • 2Real-Time Defect Detection Using AI
  • 3Data-Driven Sustainability Optimization for Manufacturing
  • Cracks in steel were only detected at late production stages, causing costly reprocessing.
  • High material waste and increased energy use due to late defect discovery.
  • Lack of real-time insight hampered process efficiency and decision making.
  • Sustainability and profitability seen as competing priorities in steel manufacturing.
  • Implemented machine learning on Azure to analyze production process data.
  • Developed AI models to predict and prevent crack formation in steel early in the process.
  • Introduced data governance and built an integrated AI-IT infrastructure hub.
  • Applied predictive analytics for equipment and maintenance optimization.
  • Real-time analysis to detect leaks and monitor manufacturing parameters.
  • Enabled 87% risk prediction for crack formation, allowing earlier intervention.
  • Reduced waste and saved energy by scrapping defective material pre-emptively.
  • Improved production efficiency and economic profitability.
  • Lowered maintenance costs through predictive maintenance projects.
  • Enhanced overall sustainability by decreasing resource and energy usage.
Architecture

Uddeholm collects production data, then integrates it into an Azure-based cloud platform for storage and analysis. Machine learning models are trained on this data to predict potential crack formation and other defects. The system delivers early warnings and recommended process changes to production teams, allowing real-time intervention. A governance model assures data quality and enables department-wide collaboration via an AI-IT infrastructure hub.

Implementation partners1
Sources & evidence1
Evidence: Medium50/100Evidence strength
  • Customer explicitly identified
  • Deployment status explicitly supported
  • Quantified outcome available
  • Technical implementation details available
Type: Video Or WebinarPublisher: cgi.com

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Was this useful?

Community

Comments

No published comments yet.