DOI · DOI-0168 · record R2595
Machine Learning for Bat Acoustics
Department of the Interior · USGS
Overview
- Development stage
- PilotSource: b) Pilot – The use case has been deployed in a limited test or pilot capacity.
- High-impact designation
- Not high-impactSource: c) Not high-impact
- Impact justification
- Not reported
- Start date
- 2022-08-23 (day precision)Source: 8/23/2022
- Withheld from public reporting
- NoSource: a) No
Mission
- Topic area
- Science
- Operational functions
- No function tag matched Derived by keyword rules; see methodology
Problem the AI is intended to solve
Improving the accuracy and timeliness of species status assessments and science to support deregulation efforts.
Expected benefits
This effort will result in significant cost savings related to environmental review for permitting.
System outputs
This model will be refined and used for prediction and decisions.
Technology
- AI classification
- Other
- System name(s)
- Not reported
- Custom-developed code
- No
- Public source code
- code.usgs.gov ↗
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
- No
- Demographic features
- k) None of the above
- Training and evaluation data
- This dataset contains audio files of bat echolocation calls used to develop NABat ML algorithm V1.0, excluding the test (holdout) set. Recordings were collected by monitoring partners across North America using ultrasonic recorders for stationary and mobile surveys, then post-processed to remove noise and assign species labels. Labeling typically involves automated classification followed by manual review (see NABat guides). Files were submitted in WAV format and include 35 classes (34 species + noise), with 4 species excluded for low sample size. From this pool, recordings were randomly selected and split into training and validation sets; the test set is not included. Audio files are grouped by four-letter species codes, with a reference dataset providing Family, Genus, Species, and Common name definitions.
- Federal Data Catalog
- www.sciencebase.gov ↗
Governance
0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.
Pre-deployment testingNo answer reported
AI impact assessmentNo answer reported
Independent reviewNo answer reported
Ongoing monitoringNo answer reported
Operator trainingNo answer reported
Fail-safeNo answer reported
Appeal processNo answer reported
User and public consultationNo answer reported
Potential impacts and how they were identified
Not reported