Return to search results
TEAMER: Field Demonstration of MarineSitu Marine Energy Monitoring Tools (Public)
This submission contains data collected during a four-month field demonstration (July-October 2024) of MarineSitu's Adaptable Monitoring Package (AMP) at the Pacific Northwest National Laboratory (PNNL) Marine and Coastal Research Laboratory (MCRL) in Sequim, WA. The project evaluated the platform's survivability in a tidal channel and the iterative development methodology for real-time AI-driven environmental monitoring.
The dataset includes:
Sample Optical Imagery: High-resolution greyscale images from a modular camera systems capturing confirmed or suspected marine life interactions using automated object detection-activated acquisition.
Machine Learning Artifacts: Object detection model weights (.pt) for three iterations each of optical and acoustic models, validation datasets with manually labeled annotations (.txt), and model-generated detection predictions (.json).
Analysis Software: Python-based tools for model performance validation (Precision, Recall, mAP50, false positive rate analysis) and synchronized multi-instrument data review.
Prerequisites/Assumptions: Use of the included Python scripts requires a Python 3.10+ environment. Users should refer to the provided README.md files within the "Data Viewer" and "Model Validation" directories for installation and usage instructions.
Note: This submission includes a representative sample of optical imagery. A larger dataset of optical and sonar data will be made available on July 1, 2029, in the following repository: https://mhkdr.openei.org/submissions/689 (linked below).
Resources
6 resources available
-
TEAMER Post Access Report Version 1.2.docx
DOCX -
README.md
MD -
Data Viewer.zip
ZIP -
Model Validation.zip
ZIP -
Sample Data.zip
ZIP -
TEAMER Field Demonstration of MarineSitu Marine Energy Monitoring Tools
HTML
Find Related Datasets
Search by Tags
Click any tag below to search for similar datasets
Complete Metadata
| @type | dcat:Dataset |
|---|---|
| accessLevel | public |
| bureauCode |
[ "019:20" ] |
| contactPoint |
{ "fn": "Paul Murphy", "@type": "vcard:Contact", "hasEmail": "mailto:paul@marinesitu.com" } |
| dataQuality |
true
|
| description | This submission contains data collected during a four-month field demonstration (July-October 2024) of MarineSitu's Adaptable Monitoring Package (AMP) at the Pacific Northwest National Laboratory (PNNL) Marine and Coastal Research Laboratory (MCRL) in Sequim, WA. The project evaluated the platform's survivability in a tidal channel and the iterative development methodology for real-time AI-driven environmental monitoring. The dataset includes: Sample Optical Imagery: High-resolution greyscale images from a modular camera systems capturing confirmed or suspected marine life interactions using automated object detection-activated acquisition. Machine Learning Artifacts: Object detection model weights (.pt) for three iterations each of optical and acoustic models, validation datasets with manually labeled annotations (.txt), and model-generated detection predictions (.json). Analysis Software: Python-based tools for model performance validation (Precision, Recall, mAP50, false positive rate analysis) and synchronized multi-instrument data review. Prerequisites/Assumptions: Use of the included Python scripts requires a Python 3.10+ environment. Users should refer to the provided README.md files within the "Data Viewer" and "Model Validation" directories for installation and usage instructions. Note: This submission includes a representative sample of optical imagery. A larger dataset of optical and sonar data will be made available on July 1, 2029, in the following repository: https://mhkdr.openei.org/submissions/689 (linked below). |
| distribution |
[ { "@type": "dcat:Distribution", "title": "TEAMER Post Access Report Version 1.2.docx", "format": "docx", "accessURL": "https://mhkdr.openei.org/files/709/OW_RFTS_MarineSItu_PNNL_Post_Access_Report_Version_1.2.docx", "mediaType": "application/vnd.openxmlformats-officedocument.wordprocessingml.document", "description": "The project's public post-access report. Covers the deployment at PNNL's Marine and Coastal Research Laboratory, the instrumentation and platform, the data acquisition and management approach, the iterative object detection model development carried out during the deployment, and the resulting performance and reliability outcomes. Section 7.5.1 and the Test Plan Deviation discussion in Section 8.2 describe how project data are divided between this repository and the embargoed archive, and why the complete imagery and prediction archives are held under a data use moratorium. The analysis scripts in model_validation reproduce the performance figures reported in Section 8." }, { "@type": "dcat:Distribution", "title": "README.md", "format": "md", "accessURL": "https://mhkdr.openei.org/files/709/README.md", "mediaType": "application/octet-stream", "description": "Top-level overview of the submission. Describes the four components, how to set them up, and how to run the analysis and playback tools against the included sample data. Read this first: the tool commands assume all three archives are extracted into the same directory, and the paths will not resolve otherwise." }, { "@type": "dcat:Distribution", "title": "Data Viewer.zip", "format": "zip", "accessURL": "https://mhkdr.openei.org/files/709/data_viewer.zip", "mediaType": "application/zip", "description": "A playback utility for reviewing synchronized instrument data with object detection predictions overlaid on the imagery. Plays multiple instrument feeds side by side against a common timeline, holding each feed's last frame where its capture rate differs, and can skip frames whose detections fall below a confidence threshold. Requires Python 3.8 or newer, OpenCV, and NumPy. Includes a setup script, requirements file, and usage documentation. Licensed under AGPL-3.0." }, { "@type": "dcat:Distribution", "title": "Model Validation.zip", "format": "zip", "accessURL": "https://mhkdr.openei.org/files/709/model_validation.zip", "mediaType": "application/zip", "description": "Trained object detection models and the tools to reproduce the performance figures reported for the project. Contains weights (.pt) for three iterations each of the optical and acoustic models, the manually annotated validation datasets, and per-run validation outputs including plots and confusion matrices. Two scripts are included: one evaluating each model against a shared validation set to give Precision, Recall, and mAP50, and one repeating the deployment false positive analysis from a reviewed dataset of 91,662 detections. Requires Python 3.10 or newer, Ultralytics, Polars, and Pandas. Licensed under AGPL-3.0; the models were developed under MarineSitu's commercial license with Ultralytics." }, { "@type": "dcat:Distribution", "title": "Sample Data.zip", "format": "zip", "accessURL": "https://mhkdr.openei.org/files/709/sample_data.zip", "mediaType": "application/zip", "description": "A representative selection of verified detections from the deployment, structured for the included data viewer. 924 grayscale frames, each paired with the object detection prediction the edge system recorded for it, across twelve time windows in July, August, and October 2024. Organized into four datasets by subject: fish (eight windows), seals on imaging sonar (two windows), crab, and a bird recorded simultaneously on two cameras. Instruments are two optical cameras and a Tritech Gemini 1200ik imaging sonar. Every window was reviewed by a MarineSitu engineer before inclusion. The included README documents each window and explains how to read the annotations." }, { "@type": "dcat:Distribution", "title": "TEAMER Field Demonstration of MarineSitu Marine Energy Monitoring Tools", "format": "HTML", "accessURL": "https://mhkdr.openei.org/submissions/689", "mediaType": "text/html", "description": "The delayed release submission for this project including the TEAMER post access report, README, data viewer, model validation, and sample data." } ] |
| identifier | https://data.openei.org/submissions/8760 |
| issued | 2026-08-21T06:00:00Z |
| keyword |
[ "AMP", "Adaptable Monitoring Package", "Environmental Monitoring", "Hydrokinetic", "MHK", "ML", "Machine Learning", "Marine", "Object Detection", "Python", "Sequim Bay", "TEAMER", "acoustic", "acoustic imagery", "code", "data", "energy", "imagery", "miniAMP", "optical", "optical imagery", "power", "raw data", "sonar" ] |
| landingPage | https://mhkdr.openei.org/submissions/709 |
| license | https://creativecommons.org/licenses/by/4.0/ |
| modified | 2026-08-28T18:08:12Z |
| programCode |
[ "019:009" ] |
| projectLead | Lauren Ruedy |
| projectNumber | EE0008895 |
| projectTitle | Testing Expertise and Access for Marine Energy Research |
| publisher |
{ "name": "MarineSitu", "@type": "org:Organization" } |
| spatial |
"{"type":"Polygon","coordinates":[[[-180,-83],[180,-83],[180,83],[-180,83],[-180,-83]]]}"
|
| title | TEAMER: Field Demonstration of MarineSitu Marine Energy Monitoring Tools (Public) |
