Sandia Labs to Deploy AI-Augmented Inspection for Nuclear Ceramics This Fall
Sandia Labs will deploy AI‑augmented inspection for nuclear ceramics by early fall, with operators verifying AI findings under NNSA funding.

Brighter Future
TL;DR: Sandia National Laboratories will deploy an AI‑augmented inspection system for nuclear ceramic components by early fall, with operators verifying AI findings. The project, funded by the National Nuclear Security Administration’s AI for Nuclear Security initiative, aims to catch defects earlier and save time and money.
At Sandia’s Albuquerque site, engineers are installing optical and acoustic imaging systems that scan ceramic billets—the raw starter pieces—before they are shaped into final parts. The images feed into a software tool that highlights potential flaws, allowing inspectors to review them on a workstation instead of peering through a manual microscope. This shift reduces the one‑to‑two‑year training period traditionally required for visual inspection.
The optical system captures high‑resolution surface images, while the acoustic system sends sound waves through the billet to reveal internal cracks or voids. Together they produce a digital record that the AI software scans for anomalies.
Operators will double‑check every AI‑flagged defect and will also catch any flaws the AI misses, ensuring human oversight remains central. Sandia Labs plans to have the new imaging and AI inspection system operational by early fall, according to the lab’s schedule. The National Nuclear Security Administration’s AI for Nuclear Security initiative finances the effort, which supports applying artificial intelligence to the nation’s nuclear deterrence mission.
By catching defects at the billet stage, the lab avoids investing further labor and material into parts that would later be rejected. Early estimates suggest the workflow could cut inspection time, though exact figures are pending. The approach also frees skilled technicians to focus on other tasks while components are being scanned, increasing overall throughput.
Looking ahead, watch for the system’s performance data after the fall rollout and any plans to extend the AI‑augmented workflow to other Sandia facilities and nuclear security partner sites.
Continue reading
More in this thread
Conversation
Reader notes
Loading comments...