Case study · Semiconductors & electronics

Micron scales vision, acoustic and thermal AI across its memory fabs

Wafers pass through some 1,500 steps over months; people reviewing 30-40 images per wafer miss defects, and equipment wear, leaks and overheating are costly and dangerous if caught late.

Manufacturer
Micron Technology
Use cases
3
Evidence
Manufacturer-reported
Year
2025

Manufacturer-reported  This result comes from the manufacturer's own publications, such as newsroom posts, earnings materials or engineering blogs. It has not been independently verified.

The challenge

Wafers pass through some 1,500 steps over months; people reviewing 30-40 images per wafer miss defects, and equipment wear, leaks and overheating are costly and dangerous if caught late.

The approach

Micron runs AI on data from 590,000 sensors (100 million wafer images per week): computer vision and deep-learning automatic defect classification, acoustic anomaly detection near pumps and actuators, and thermal imaging of transformers, plus yield platforms used by 14,000+ people.

Reported improvements

Percent change as reported by the source; hover for the exact wording

Product scrap↓ 50%
Time to market for new products↓ 50%
Manufacturing tool availability↑ 4%

3 use cases in this implementation

  1. Visual quality inspection

    Computer vision and deep-learning automatic defect classification scan wafer and assembly/test images and video for scratches, holes and film bubbles and cluster defects by type.

    • 50% reduction in product scrap
    • 50% reduction in time to market for new products
    • 1 million work hours annually in labor productivity gains
  2. Predictive maintenance

    Microphones near robotic actuators and pumps learn normal sound signatures and alert on new frequencies; thermal imaging watches transformers for overheating.

    • 4% improvement in manufacturing tool availability
  3. Energy & utilities optimization

    AI sensors provide granular energy metering across fab equipment.

    • They deliver granular metering of energy, leading to significant usage and energy savings

Outcomes as reported by the source. We do not adjust, combine or extrapolate reported figures.

Related case studies

Put AI to work

Find the AI opportunities hiding in your plant.

Genvo, our implementation partner, works with manufacturers to pick the right first project, connect the systems you already run and put AI agents and tools into daily use, with people approving what goes out.

Assessments and consultations are provided by Genvo. How our partnership works.