AI for Manufacturing is a curated database of real AI implementations in manufacturing. Every case study describes an actual deployment at a named (or, where the source withheld it, described) organization — not a hypothetical, a vendor demo, or an invented outcome. This page explains where the data comes from and how it is checked.
Each case study originates from a published source — a vendor case study, press release, regulatory filing, conference talk, or reputable news report documenting the deployment. Entries cite that source and link back to it, and record the source's publication date. New entries are not accepted without a traceable source.
Candidate implementations pass through an automated extraction and scoring gate before they are eligible to publish. The gate checks that an entry names a real organization, describes a concrete problem and solution, and carries a verifiable source. Entries that are too thin, low-confidence, or unsourced are held for review or excluded rather than published to pad a count. Metrics are reported as stated by the source; we do not manufacture ROI figures.
Every entry records the publication date of its underlying source, so you can see how recent the reported outcome is. Any aggregate figure shown across the site — counts, medians, ranges — is computed directly from the underlying database at build time, never hand-entered or rounded up. If the data doesn't support a number, we don't show one.
Publishing an entry isn't the end of the check. On a recurring weekly cycle we re-verify that each published entry's source link is still live, and record the date of the most recent successful check as that entry's last-verified date — shown on the case study itself. Entries whose source has gone offline are flagged and removed rather than left standing on a dead link, so the last-verified date reflects a real, repeated check and not a one-time import.
The directory is maintained by a named editor with a public professional profile (see the byline above), backed by the automated sourcing and quality pipeline described here. If you find an inaccuracy — a wrong metric, a misattributed company, a stale source — please tell us and we will correct or remove the entry. We would rather have a smaller, accurate directory than a larger, unreliable one.
AI Roadmap — Eight workflows in your plant where AI would pay back this quarter. We know which. 12 pages, in 72 hours.