Computer vision · Nameplates
The plate on the motorbecomes a data sheet.
Nameplate reading is the automatic extraction of data from the nameplate of a motor, pump or other equipment, from a photo, into structured fields such as manufacturer, model, serial number, power or voltage. We build workflows in which the technician photographs the plate, the system reads and organises the data, a person confirms it, and the result reaches your inventory, maintenance system or ERP.
Events
- 11:03:09Plate found · motor M-14
- 11:03:10Fields read and mapped to the schema
- 11:03:101 uncertain field → human check
- 11:03:41Exported to the equipment record
The data is on the plate, but not in the system
Every motor or piece of equipment has a plate with its technical data, but that data only reaches your records if someone copies it by hand. With hundreds of assets, an inventory takes weeks, and a wrong figure for power or serial number is only discovered when you order a part.
- Nameplate data is transcribed by hand into notebooks or spreadsheets, with copying mistakes
- Each manufacturer lays out and abbreviates the fields differently, which slows transcription down
- Photos taken in the field stay on phones, with no link to the asset in the records
- The maintenance system or ERP holds incomplete records, and spare parts are hard to find
Who it is for
- Factories and production sites taking inventory of their motors, pumps and machinery
- Maintenance teams that want complete equipment records without manual transcription
- Service and repair companies that receive equipment from several customers every day
- Facility management companies taking over buildings with hundreds of undocumented technical assets
Step by step
How the system reads a nameplate, step by step
The five steps on the animation show what happens to the photo of a motor nameplate: the shot, finding the plate, reading and sorting into fields, checking the values and the technician's confirmation.
Events
Events
- 11:03:09Plate found · motor M-14
Events
- 11:03:09Plate found · motor M-14
- 11:03:10Fields read and mapped to the schema
Events
- 11:03:09Plate found · motor M-14
- 11:03:10Fields read and mapped to the schema
- 11:03:101 uncertain field → human check
- 11:03:41Exported to the equipment record
Events
- 11:03:09Plate found · motor M-14
- 11:03:10Fields read and mapped to the schema
- 11:03:101 uncertain field → human check
- 11:03:41Exported to the equipment record
Step 1 of 5
The right photo
The technician photographs the plate from the front, with the whole plate in the frame and as close as possible, without the flash leaving a bright spot on the metal. On shiny plates, side lighting or a slight tilt of the phone helps. If the image comes out blurred or cropped, the app asks for a retake on the spot, while the technician is still next to the equipment.
How it can be used
Field app
The technician photographs and confirms the record on the phone, next to the equipment, and can retake a poor photo straight away. It suits new inventories and workshop intake.
Processing existing photos
Photos already taken, gathered from phones or folders, are processed in batches and the records are checked at a desk. This is the right option when you already have image archives with no data extracted from them.
Manufacturer-specific rules
When most of the equipment comes from a few manufacturers, the layout of their plates is learnt separately and fields are sorted more reliably. For very mixed fleets, general label-to-value extraction remains, refined over time.
What you receive at hand-over
- The reading and structuring workflow configured on your fields, on the server or cloud of your choice
- The field app or the photo upload interface, depending on the option chosen
- Rules for checking units, ranges and mandatory fields
- The confirmation interface, with the nameplate photo next to the record
- Export or integration with Excel, the maintenance system or the ERP, typically via import or API
- Documentation, technician training and a measurement report on the nameplates of your own equipment
No strings attached: tell us what the cameras should see and we'll tell you honestly whether it can be done.
Let's talkWhat the reading system can do
- 01
Finds and straightens the plate
The model locates the plate in the photo and corrects the perspective, so the text can be read as if from a flat sheet.
- 02
Reads the text on the plate
OCR extracts engraved, stamped or printed text, including common symbols and units of measurement.
- 03
Sorts the data into fields
Values are placed in your fields: manufacturer, model, serial number, power, voltage, current, speed, frequency, protection rating.
- 04
Checks the values
The system compares units and usual ranges and flags values that look wrong or are missing.
- 05
Reads barcodes and QR codes
Where the plate or label has a barcode, QR code or Data Matrix code, it is read and linked to the record.
- 06
Exports to your systems
Confirmed records go to Excel, the maintenance system or the ERP, together with the original photo.
How it works
How a nameplate gets from a photo into your records
The steps marked in yellow are approved by a person.
- Input
- AI
- Action
- Human in the loop
- Result
- 01
Input
Photo of the plate
The technician photographs the plate with a phone or tablet, from the app.
- 02
AI
Reading and structuring
The model reads the text and places the values in the defined fields.
- 03
Action
Automatic checks
Units, formats and unusual values are flagged for attention.
- 04
Human in the loop
Technician confirmation
A person compares the record with the photo and corrects anything wrong.
- 05
Action
Export to systems
The confirmed record reaches Excel, the maintenance system or the ERP.
- 06
Result
Equipment record
You have organised technical data, with the photo of the plate attached.
What nameplate reading needs
- Real photos of the plates on your equipment, from several manufacturers and in different states of wear
- The list of fields you want to capture and the format in which they must reach your records
- Examples of correctly completed records, used for testing and for the validation rules
- A phone or tablet with a good camera and, in dark spaces, a light source
- Access to the target system: an Excel file, the maintenance system or the ERP, via import or API
Possible integrations
- Microsoft Excel and CSV files
- Google Sheets
- ERP systems, for example SAP
- Maintenance management systems (CMMS)
- Microsoft SharePoint
- Web and mobile field applications
- REST API and webhooks
- n8n or Make for workflows
- PostgreSQL databases
- Microsoft Excel and CSV files
- Google Sheets
- ERP systems, for example SAP
- Maintenance management systems (CMMS)
- Microsoft SharePoint
- Web and mobile field applications
- REST API and webhooks
- n8n or Make for workflows
- PostgreSQL databases
Each integration's compatibility is checked during the analysis phase, for the systems and versions you use.
Control and transparency
What stays with people
- Confirming each record before it is saved, especially values flagged as uncertain
- Setting the mandatory fields and the rules for when a value is considered plausible
- Technical decisions made on the basis of the data, for example ordering a part or replacing a motor
- Filling in missing or unreadable data from the manufacturer's documentation
Limits and operating conditions
- Heavily worn, rusty, painted-over or damaged plates may have characters that can no longer be read correctly
- Reflections on metal, blurry photos or photos taken at a steep angle reduce reading quality
- Manufacturers use different layouts and abbreviations; an entirely new type of plate may need additional examples
- The system reads what is written on the plate; it does not check whether the equipment runs at the stated values
How we evaluate nameplate reading
We agree the indicators together before launch and track them afterwards. The values depend on the process, the data and the volume.
- Fields read correctly
- The share of correct fields per record, compared with an independent manual transcription on a sample
- Corrections by technicians
- How many fields are changed at confirmation, tracked by field and by manufacturer
- Unflagged wrong values
- Errors found after saving that the automatic checks did not flag, analysed by cause
- Results by plate type
- Reading quality compared separately on engraved, stamped and printed plates, new and worn
How we build nameplate reading, in 4 steps
- 01
We define the fields and the destination
We agree which data matters, what it is called in your records and where it needs to go.
- 02
We gather real photos
We collect plates from your equipment, including hard-to-read ones, and prepare the correct records for testing.
- 03
We build and test the workflow
We combine reading, structuring and validation rules, then test on plates the system has not seen.
- 04
We integrate and monitor
We connect the export to your systems and track technicians' corrections to improve reading over time.
Use cases
The scenarios below are conceptual examples, designed to illustrate realistic applications. They are not implementations delivered by WEB-AZ.
- Manufacturing
Inventory of a factory's motors
The maintenance team photographs the nameplates of motors on the lines, and the confirmed records form a complete inventory in the maintenance system.
Conceptual example - Equipment servicing
Equipment intake at the workshop
When a pump arrives for repair, the photo of its plate fills in the service record, and the receptionist only checks the flagged fields.
Conceptual example - Facility management
Taking over a building
When a building is taken over, the equipment in the plant rooms is documented quickly, and the list goes into a spreadsheet checked by the responsible engineer.
Conceptual example
Useful guides
Before you decide, read
Frequently asked questions
What is automatic nameplate reading?
It is the extraction of data from the nameplate of a motor or piece of equipment, from a photo, into structured fields such as manufacturer, model, serial number, power or voltage. The data is checked by a person and then sent to the inventory, the maintenance system or the ERP.
What types of plates can be read?
Plates on electric motors, pumps, compressors, transformers or other industrial equipment, whether engraved, stamped or printed. We first test on photos of your plates, because their layout and condition vary a lot from one manufacturer to another.
Do we need to buy special equipment?
Usually not. A phone or tablet with a good camera is enough for most plates, and a torch helps in dark spaces. For very small or very worn plates, we look together at whether anything else is needed.
What happens if the system misreads a value?
Every record goes through human confirmation, and uncertain or unusual values are flagged. The technician sees the photo next to the record and corrects anything needed before saving.
Can the data be exported to Excel or an ERP?
Yes. Confirmed records can be exported to Excel or CSV or sent directly to the ERP or maintenance system, via import or API. The exact method depends on the interfaces your system offers, and we check this at the start.
Can labels with barcodes or QR codes be read too?
Yes, barcodes, QR codes and Data Matrix codes on plates or labels can be read in the same step and linked to the equipment record. They complement the text that is read, especially for the serial number or product code.
Can’t find the answer? Write to us and we’ll reply directly.
Do you have equipment whose data only lives on its nameplate?
Send us a few photos of nameplates and the list of fields you need. We will tell you what can be read automatically and how the data reaches your system.
Related solutions
- Intelligent document processingThe engine that extracts, classifies and summarises documents in batches and delivers structured data, ready to review and import.
- Industrial visual inspectionCameras on the production line flag parts with visible defects, and the operator confirms every rejection.
- AI agent for operationsTracks internal requests, tickets and stock thresholds, prepares tasks for teams and coordinates handovers between departments, with a person responsible for every decision.