Siger is an electrical engineering company that designs and produces custom switchboards and electrical panels for industrial clients. When Nicola Vaccari, CEO, engaged with donia®, his team was dealing with a manual-intensive process for extracting bills of materials from engineering documents — primarily electrical diagrams produced in various formats by internal designers and external partners.
The problem was not a lack of data. It was the time and effort required to turn that data into production-ready information. Engineers spent hours reading through diagrams, extracting component lists by hand, and entering them into the warehouse and procurement system. Errors were common. Turnaround times were unpredictable. And the margin for last-minute changes was effectively zero.
The challenge: documents as bottlenecks
In custom manufacturing environments like Siger's, every project is different. Switchboard configurations change with each client specification. That variability makes standardized ERP templates difficult to maintain and forces engineers to re-do extraction work for each new order. With project volumes growing and lead times tightening, the manual BOM extraction process had become a bottleneck that no amount of additional staffing could reliably solve.
Vaccari's goal was clear: extract bills of materials from electrical diagrams automatically, with enough accuracy to feed directly into the warehouse allocation and procurement workflow — without manual review for the majority of cases.
How donia® was deployed
donia® was configured to receive electrical diagram files — including PDFs with embedded schematic data and scanned diagrams requiring vision-based analysis — and extract structured component lists from them automatically. The Document Intelligence module handled the extraction, field normalization, and output formatting. The Workflow Builder was used to add validation rules, route edge cases for manual review, and push confirmed BOMs directly to the warehouse management system.
Within the first two weeks, the team processed a backlog of over 40 projects through the automated pipeline. The accuracy rate on clean digital diagrams was high enough to bypass manual review in the large majority of cases. For scanned or complex documents, the workflow routed only flagged items to an engineer — reducing review time by more than 80% compared to full manual extraction.
Results after 60 days
- 30% reduction in warehouse labour costs linked to BOM processing and inventory preparation
- BOM extraction time reduced from hours to minutes per project
- Elimination of manual component re-entry for standard digital diagrams
- Faster engineering document turnaround, enabling shorter project lead times
- Improved accuracy in inventory reservation and procurement requests
What Vaccari says
"With donia®, we extract bills of materials from electrical diagrams automatically. Less manual effort, better accuracy, and engineering documents turned into production data faster."