By
Every Maximo implementation starts in the same place: a stack of OEM manuals. Before a single PM schedule, job plan, or spare parts list exists in Maximo, someone has to read that documentation and turn it into structured data. On a capital project, that can mean thousands of documents across hundreds of asset types.
Strategic Maintenance Solutions (SMS) partnered with Meibel to take that work from days to minutes, without giving up the accuracy Maximo data demands. Read the full case study on meibel.ai.
The challenge: thousand-page manuals, zero margin for error
SMS reliability engineers have long done this work by hand, and the source material doesn’t make it easy:
- Massive documents. A single OEM manual can run past 1,000 pages.
- Mixed formats and languages. PDFs, Word files, Excel sheets, and scanned images, often in more than one language.
- Data buried everywhere. Part numbers, serial numbers, model IDs, manufacturers, maintenance intervals, and task steps sit in prose, tables, and diagrams.
- Conflicting values. The same component can appear with different values in different parts of one manual, and each conflict has to be resolved.
SMS had tried open-source RAG tools and third-party parsers, but they fell short on accuracy, speed, and scale, especially with tables embedded in prose and multilingual content.
“A wrong part number in Maximo means the wrong part gets ordered, so close enough isn’t good enough for this work.”
Dan Holzwarth, Senior Consultant / Field Engineer, SMS
The solution: extraction agents built around Maximo
Using Meibel’s Document Intelligence and purpose-built extraction agents, SMS built a five-step pipeline:
- Ingest OEM manuals, maintenance guides, and parts catalogs.
- Extract structured data against output schemas that match Maximo’s import requirements.
- Resolve conflicts by surfacing every candidate value along with where it appears in the source.
- Score and route results. High-confidence data goes straight into Maximo-ready CSV or JSON files; low-confidence data is flagged for consultant review.
- Scale through batch processing, with the same agent definition and no code changes.
In one example, a 132-page Caterpillar generator set manual was processed with its tables and maintenance schedules intact.
The results
- Higher accuracy: extraction accuracy rose from a 70–95% baseline to 90–100%.
- Days to minutes: capital project setup that once took days now takes minutes.
- Faster delivery where it counts: PM schedule configuration, storeroom lockdowns, and parts list generation all move faster.
- Built to scale: the same pipeline handles 10 documents or 10,000.
What it means for Maximo clients
SMS is building this capability into its service delivery as a reusable part of every Maximo implementation. For clients, that means a faster path from equipment documentation to a working asset hierarchy, PM program, and inventory setup, with SMS consultants spending their time on judgment calls and maintenance strategy instead of retyping data from manuals.
Planning a Maximo implementation, upgrade, or capital project?
Talk to SMS about putting AI-assisted document processing to work on your equipment library: sales@sms-inc.net · +1 207-854-9400 · www.sms-inc.net
About Strategic Maintenance Solutions
Founded in 1999, Strategic Maintenance Solutions (SMS) is a trusted adviser and global leader in Enterprise Asset Management, Industrial IoT, and reliability consulting. As an IBM Platinum Partner, SMS delivers innovative solutions to Fortune 1000 companies across oil & gas, life sciences, healthcare, aerospace, and facilities management, driving sustainable operational improvements.
