Inspectors took the photo, and super.AI sent the make, model and serial number back to the company's system in hours, not days.
When this company ran asset inventories and maintenance work for its clients, inspectors photographed the nameplate on each piece of equipment. That's the small plate, sometimes called a data plate, with the make, model and serial number on it.
Getting those details from the photo into a system was done by hand. Someone looked at each image and typed it up.
The company had tried an OCR tool, but inspectors still had to select each piece of data in every image. That's a lot of clicking, and it slowed teams down between site visits.
“We capture the photo and send to super.AI and it's pulled and viewed and captured and sent back to our system with a turnaround of a couple of hours, not more than a day.”
An inspector took a photo of the nameplate and sent it to super.AI. Our AI pulled out the make, model and serial number, and for some programs the year, asset tag or work order too.
Before anything went back, human reviewers checked the results. Then a structured record went into the company's system through an API.
Inspectors no longer had to click on fields inside the image. They took the picture and moved on.
We worked with this company for 6 years, from 2020 to 2026, on programs for their clients in the U.S. and Europe.
Our counts here start in January 2024, so earlier years aren't included. We extracted about 226,000 nameplates in 2024, about 262,000 in 2025 and about 96,000 from January to August 2026, more than 580,000 over that stretch.
If your inspectors photograph nameplates and someone back at the office types up what's on them, that's the work we did for this company. Here's how our equipment data plate extraction works.
