Dam Inspection Report Document Processing

Overview

Development stage

DeployedSource: c) Deployed – The use case is being actively authorized or utilized to support the functions or mission of an agency.

High-impact designation

Presumed high-impact, determined notSource: b) Presumed high-impact, but determined not high impact

Impact justification

The AI is used for information extraction that informs dam portfolio decisions. However, the principal basis for decisions is made by humans for the repairs, and is vetted through aerial imagery analysis, the Oklahoma Conservation Commission, and associated dam sponsors.

Start date

2023-12-20 (day precision)Source: 12/20/2023

Withheld from public reporting

Not reported

Mission

Topic area

Administrative Functions

Operational functions

Document processing and extraction Derived by keyword rules; see methodology

Problem the AI is intended to solve

The AI use case allows our staff to identify the biggest issues and trends across 2,100+ dams.

Expected benefits

The benefit and positive outcome from the AI is a wholistic understanding of dam issues across 2,100+ dams. This allows us to better analyze the data and build a risk-prioritization system to rank sites and projects to work on, reducing labor hours required to complete the task manually, and reducing the dependence on specific staff to maintain knowledge of each dam's inspection response.

System outputs

The model outputs text and checkbox responses, including dam metadata, inspection issue responses (yes and no checkboxes), and further remarks on the issue regarding what has been or needs to be done on the dam.

Technology

AI classification

Computer visionSource: Computer Vision: AI that processes and interprets visual data (e.g., images and videos).

System name(s)

AI Builder

Custom-developed code

No

Public source code

Not reported

Sourcing

How it was built

Developed in-houseSource: b) Developed in-house

Vendor (as reported)

Not reported

Vendors (standardized)

None on the standard list

A vendor is the supplier named by the agency. It does not identify the underlying model or AI technology.

Data and privacy

Involves PII

No

Privacy Impact Assessment

Not reported

Authorization to Operate

Yes

Demographic features

k) None of the above

Training and evaluation data

The data is made up of dam inspection forms collected by dam owners (ex. conservation districts, cities) or Natural Resources Conservation Service (NRCS) staff. The data was validated through manual checks and is composed of 30+ dam inspection reports with different handwriting. The data contains 120+ variables that are read from each two-page inspection report. The data types are date, text, table (text variable types in the table), and checkbox binary outputs (checked: yes or no).

Federal Data Catalog

Not reported

Governance

0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.

Potential impacts and how they were identified

Not reported