This report presents Bridge Data Virtual Validator, which is a prototype web-based tool developed to support automated quality assurance (QA) of INDOT bridge project deliverables. The project focuses on Final Tracings documents, where important bridge data are distributed across bridge plans, cost estimates, and other documents/elements. Two validation components were developed. The first component supports bridge plan format compliance checking by using optical character recognition (OCR), computer vision, and rule-based reasoning to identify missing or inconsistent letting control information such as file numbers, dates, stamps, signatures, and sheet numbers. The second component performs comparative checking of pay items between bridge plans and cost estimates with OCR-based information extraction, rule-based structuring, hybrid matching, and deterministic consistency checking. The comparative-checking algorithms were refined through a data-driven development process and with manually reviewed discrepancy records. Both components preserve source evidence and generate traceable reports so that reviewers can locate, verify, and resolve potential issues. The project demonstrates how existing PDF-based bridge deliverables can be converted into structured validation records without changing current submission formats, providing a practical foundation for future pilot deployment, reviewer feedback integration, and broader bridge data QA automation.

Infrastructure Assets: Bridge, Highway Assets
Resource Types: Research Report
Capabilities: Data & Information Systems
Management Processes: Performance Based Planning & Programming