Sun Ho Ro
← All projects
Document AIPDF AutomationEngineering QA/QCPlan Review

Bridge Plan Review Automation

An AI-enabled PDF review workflow that prepared large annotated bridge plan sets and improved visualization of comment and reply relationships.

Client
New Jersey Turnpike Authority
Location
Piscataway, NJ
Year
2026
Role
GeoAI Engineer

Context

The NJTA Interchanges 1 to 4 Capacity Enhancements Program involved planning, design, and technical review work across bridge structures, ramps, profiles, and roadway geometry. The project team needed to review large annotated plan sets with extensive comments and threaded markups.

Challenge

The review environment did not provide the same annotation management capabilities as the tools often used by engineering teams. Comments, replies, page layouts, and markup relationships had to be made readable in a print-based and non-Bluebeam workflow.

My Role

I led the development of an automated PDF preparation workflow for internal project teams. My work focused on extracting annotation data, cleaning comment structures, improving page layouts, and making comment relationships easier to interpret.

Approach

The workflow processed annotated blueprint files, extracted PDF annotation metadata, cleaned and organized comment and reply data, and generated improved visual layouts for review. The solution was designed around the actual review process, where engineers needed clear visibility into threaded comments without manually rebuilding markup context.

Output

The project produced enhanced PDF review packages with improved layout, cleaner annotation handling, and clearer visualization of comment and reply relationships.

Impact

The workflow reduced manual preparation effort and made dense plan review materials easier to navigate. It helped technical reviewers focus on engineering decisions instead of fighting with document formatting.

Tech & Techniques

PythonPyMuPDFPDF Annotation ProcessingBlueprint ReviewData ExtractionVisualization