← All Side Quests
Oct 12, 2026Data

What Holds Us Up

See how the age, material, and recorded condition of Indiana's bridges fit together.

Try the project ↓
How to use this project

Help

How to explore

The National Bridge Inventory records bridges on public roads. This project describes Indiana's 2025 inventory snapshot; it does not replace a current inspection, engineering review, closure notice, or route authority.

  1. Begin with older bridges

    Show the 2,014 records with a year built before 1940, then read their good, fair, and poor shares.

  2. Choose another generation

    Pick a broad period or a decade column. Every later view follows that choice.

  3. Narrow by material

    Choose a main span material and switch between bridge count and total recorded deck area.

  4. Read the parts

    Compare deck, superstructure, substructure, and culvert ratings, then check the reconstruction-year split without treating it as cause and effect.

Inventory bridge
A structure included in the National Bridge Inventory for a public road; the count is records, not crossings or routes.
Bridge condition
Good, fair, or poor based on the lowest applicable deck, superstructure, substructure, or culvert rating.
Deck area
The recorded bridge length multiplied by its deck or approach width, shown in square metres.
Reconstructed year
A reported year of major work; a blank does not prove that no repair or maintenance occurred.
Side Quest / DataStart with bridges built before 1940, switch between bridge count and deck area, choose a decade or main span material, compare applicable component ratings, and inspect the reconstruction-year split and individual records.
Indiana bridges · 2025 inventory

What holds
us up

See how the age, material, and recorded condition of Indiana's bridges fit together.

The inventory covers bridges on public roads. It records inspection ratings, not a promise that a bridge is safe or unsafe today.

01 · Choose a period

Begin with a generation of bridges.

Choose when bridges were built. Then narrow the view by the main span material. Every section below follows the same choices.

Measure bars by
Selected groupEvery yearall materials
Records19,542of 19,542
Good8,21342.0% of records
Fair10,28652.6% of records
Poor1,0435.3% of records
Good 42.0%Fair 52.6%Poor 5.3%

The three labels come from the lowest rating among the deck, superstructure, substructure, or culvert. A poor label means at least one applicable part was rated 4 or lower.

02 · Follow the years

Each decade leaves a different span.

Each column groups bridges by year built. Height shows the number of records; the three fills show condition. Choose a decade to carry it through the page.

GoodFairPoor

The oldest recorded year built is 1838. Age and condition appear together here, but the inventory does not show that age caused any rating.

Material view

What carries the main span?

These are the source's broad material groups. Continuous and non-continuous versions are combined. Choose one row to filter the other views.

03 · Compare the parts

One low part sets the whole label.

Each row counts available component ratings in the selected group. Good is 7–9, fair is 5–6, and poor is 0–4. Some parts do not apply to every record.

PartRating mixAvailable ratings
Deckthe surface and supporting deck
16,898
Superstructurethe beams or truss above the supports
17,143
Substructurethe piers and parts below the span
17,141
Culverta buried water passage used instead of a bridge span
2,399
Reconstruction record

A date is not a treatment score.

The inventory gives a reconstructed year for 5,526 Indiana records. It does not say that every change was the same kind of work. Compare the labels, but do not read cause into the difference.

No reconstructed year recorded14,016 selected records
Reconstructed year recorded5,526 selected records
04 · Read the record

A bridge row is still only a record.

These examples show what the source keeps for a few long or heavily traveled structures. Names, traffic estimates, and dates can be abbreviated or older than the 2025 inventory year.

01 · Structure 080777

SR 912

Crosses ind h. canal rr rds.

Year built
2020
Material
Prestressed concrete
Condition
Good
Length
1,900.7 m
Daily traffic
10,000 (2020)
02 · Structure 014204

US 41 SB

Crosses ohio river access road.

Year built
1965
Material
Steel
Condition
Fair
Length
1,654 m
Daily traffic
19,723 (2019)
03 · Structure 014202

US 41 NB

Crosses ohio river access road.

Year built
1932
Material
Steel
Condition
Fair
Length
1,644.3 m
Daily traffic
18,907 (2019)
04 · Structure 008868

US 31

Crosses ohio riv. i-64 streets.

Year built
1929
Material
Steel
Condition
Poor
Length
1,501.3 m
Daily traffic
15,200 (2004)
How to read this

Inspection data is a starting point.

The page describes the Federal Highway Administration's 2025 inventory snapshot. It does not replace a current inspection, engineering review, closure notice, or route authority.

Definitions used on this page
Inventory bridge
A structure included in the National Bridge Inventory for a public road; the count is records, not crossings or routes.
Bridge condition
Good, fair, or poor based on the lowest applicable deck, superstructure, substructure, or culvert rating.
Deck area
The recorded bridge length multiplied by its deck or approach width, shown in square metres.
Reconstructed year
A reported year of major work; a blank does not prove that no repair or maintenance occurred.

I selected the Federal Highway Administration's National Bridge Inventory through the allowlisted Data.gov catalog, then opened the annual 2025 source, official ArcGIS layer, coding guide, condition definitions, and public-domain notice on October 12, 2026.

Why this exists

The question

How do year built, main span material, reconstruction records, and inspection component ratings appear together across Indiana bridges in the Federal Highway Administration's 2025 National Bridge Inventory?

What to try

Start with bridges built before 1940, switch between bridge count and deck area, choose a decade or main span material, compare applicable component ratings, and inspect the reconstruction-year split and individual records.

Method and limits

About the data

I found the National Bridge Inventory through the allowlisted Data.gov catalog and used the Federal Highway Administration's official 2025 layer, which describes the data as current on June 20, 2025. The layer identifies Federal Highway Administration as publisher and says the federal dataset is available for unrestricted public use. The Data.gov record links the annual 2025 file and the U.S. government works notice. I also checked the agency's coding guide and condition definitions rather than interpreting field codes from the catalog summary.

On October 12, 2026, a standard-library Python fetcher made one count query and ten ordered data queries for Indiana state code 18. It retained 19 fields needed for identification, age, material, size, traffic, reconstruction year, inspection date, four component ratings, and the published condition class. The compact compressed source is 524,192 bytes with SHA-256 `f968e5c3ecca31281ecde13c46abba40300357ad8871527d991c1c270ee42961`. Its metadata stores each exact query URL, response byte count, and response SHA-256.

A second standard-library Python script verifies the compressed source's byte count, checksum, schema, 19,542-row total, unique sorted object IDs, condition codes, and every row's lowest applicable component rating. It combines the source's continuous and non-continuous material codes into seven plain-language groups while retaining each bridge's year built, reconstructed year, applicable component ratings, published condition, length, deck area, traffic estimate, traffic year, and inspection year. Repeated reduction produced the same 880,758-byte browser snapshot with SHA-256 `ec4caa9d33f5c63676a296008b6de8a35ae99ed1ac53df4c3dd1fa2d5952925e`. The page imports that snapshot and makes no runtime request.

Good means the lowest applicable deck, superstructure, substructure, or culvert rating is 7–9. Fair means it is 5–6. Poor means it is 0–4. Counts treat every inventory row equally. The deck-area switch sums the published deck-area field, so larger structures carry more weight. The reconstruction comparison separates rows with and without a reported reconstructed year; it does not assume that a blank means no work occurred or that reconstruction caused the current rating.

The snapshot is an administrative record, not live monitoring. Inspection and traffic years differ by structure, abbreviated names can be unclear, and the inventory can be corrected after publication. Condition is a broad federal reporting class, not a safety verdict. Design, environment, traffic, maintenance, inspection timing, rehabilitation, reporting practice, and many other factors can move together. These views describe associations in one snapshot and cannot establish why any bridge received its rating.

Skills used

  • Data analysis
  • Data engineering
  • Statistical reasoning
  • Visualization
  • Information design
  • Interaction design
  • Visual design
  • Creative coding
  • Product thinking
  • Editorial storytelling
  • Accessibility
  • Performance
  • Systems design

What the data shows

  1. The 2025 layer contains 19,542 Indiana bridge records: 8,213 good, 10,286 fair, and 1,043 poor. The label comes from the lowest applicable component rating, so one lower-rated part can set the whole bridge's class.
  2. The 2,014 records built before 1940 include 168 good, 1,436 fair, and 410 poor classifications. A reconstructed year appears on 1,328 of those records, but the snapshot cannot show what work was done or whether it caused the present rating.
  3. Among 4,508 records built in 2000 or later, 3,863 are good, 634 are fair, and 11 are poor. This age pattern is descriptive; inspection timing, exposure, use, design, and maintenance are not held constant.
  4. Prestressed concrete is the largest combined main span material group at 7,761 records, followed by concrete at 5,535 and steel at 5,058. Material and condition can be compared by count or deck area without treating either view as a causal test.