Datadory notebook
National Bridge Inventory Download Data: Dataset Structure and Field Coverage
1,744 datasets. Pick your catch. Every guide here is built on what the catalog can actually prove.
1,744 datasets. Pick your catch.
What fields does each NBI record include?
Each row carries location, structure type, age, traffic exposure, condition ratings, appraisal results and inspection history for exactly one structure. Datadory's verification pass checked all 26 documented fields against FHWA's Record Format page, including latitude/longitude, structure type, year built and reconstructed, average daily traffic, and the full condition-and-appraisal block.
Two definitional breaks matter if you trend across editions. STRUCTURALLY DEFICIENT was redefined starting with 2018 data, and FUNCTIONALLY OBSOLETE was discontinued after 2016. And because inspections recur roughly every 24 months while publication is annual, any year's file mixes fresh inspections with records up to about two years old - which is why FHWA itself cautions the data is not suitable for route clearance purposes.
Which datasets complement the NBI - and when should you use them instead?
The NBI answers "what condition is this specific bridge in?" No other cataloged source carries per-span deck or substructure ratings. But three neighbors cover what it leaves out:
| Dataset | Grain | Strength vs the NBI |
|---|---|---|
| USDOT Open Data Portal - National Transportation Datasets | Individual crossings, incidents, road segments; state-level aggregates | Socrata SODA API over about 176 highway- and rail-related tables, headlined by 438,762 grade-crossing rows and 251,149 Form 57 incident rows |
| Overture Maps Transportation Theme | LineString segments worldwide | Monthly GeoParquet with about 300 million road segments - 128,200,419 residential alone - plus 1,476,435 standard-gauge rail segments under commercial delivery terms |
A practical stack joins them: Overture supplies geometry to plot spans, the USDOT portal adds crossing and safety context, and the NBI supplies the condition codes neither contains. The full tradeoff sits on the NBI vs Overture Maps Transportation Theme comparison.
What can you build with 34 years of annual editions?
Annual datasets run from 1992 through 2025 - 34 editions keyed on ITEM 8 Structure Number, which is stable across decades. That continuity is why deterioration modeling leads the use cases: training condition-rating trajectories over roughly 624,000 structures needs exactly this panel shape, and it is the first thing data scientists working on highways and railtracks reach for.
Screening is the other bulk win. One national extract replaces assembling fifty state spreadsheets: filter CAT10 = poor across all 624,193 records in a single pass, or join ITEM 113 scour-critical flags against your own route network. Developers automating the ingestion get a documented character-position dictionary for every year, which is what the developers builders use cases page builds on.
Pick up where this leaves off
Every one of these ships with sample rows before you commit to anything.
National Bridge Inventory (NBI) ASCII Files
USDOT Open Data Portal - National Transportation Datasets
crossingid · railroad · typexing …+1 more
Overture Maps Transportation Theme
OpenStreetMap Overpass API
Want rows instead of a pitch? Name the datasets.
API, files, or your warehouse. Daily, weekly, or hourly.
Get a sample