Livestream 2/18
Introducing Control, for quality shops.
Learn More ->
The AI model behind auto-bubbling has been improved to detect significantly more characteristics across engineering drawings, including GD&T callouts, dimensional tolerances, and surface finish requirements that previous versions missed. We also redesigned the UI: detected characteristics are now highlighted on the drawing but not bubbled until you explicitly click on them.
In the future, AI will learn from your selection patterns to mirror your bubble preferences, making creating inspection plans not only fast but specific to your shop.
AI scans your drawing and highlights every detected characteristic. Your team clicks to confirm which callouts get bubbled and in what order. Each confirmed bubble auto-populates a full annotation — type, target value, tolerance, datum reference, grid location, and inspection settings — and lets you assign the operation, FAI flag, and sampling plan.
The new foundational learning layer is what sets this apart from conventional auto-bubbling. The AI will track your team's preferences, which callouts they bubble, which they skip, and the sequence they prefer, and adapt over time. Each drawing will take less manual input than the last.
Auto-bubbling isn't new. But most implementations are static; they detect what they detect and your team cleans up the rest every time. Fulcrum's AI actually improves with use, learning your team's patterns until it's drafting quality plans the way your best inspector would.
For shops scaling quality operations without adding headcount, this is the difference between software that saves time once and software that saves more time the longer you use it. AI handles the repetitive work. Your team handles the judgment calls. And the gap between those two keeps closing.
If you’re an existing Fulcrum user interested in Control, let us know here and your Account Manager or Executive will reach out to you.
New to Fulcrum? Talk to Sales about how forward-thinking technology can improve your shop.
