2025-PGCwHatfield-Single Source_Redacted.pdf

1 of2

Version 2.0 19Aug2020

Polar Geospatial Center DETERMINATION USING

SIMPLIFIED ACQUISITION PROCEDURES (SAP)

FOR AN ACTION NOT EXCEEDING THE SIMPLIFIED ACQUISITION THRESHOLD

(AUTHORITY: FAR 13.106-1(b)(1))

Requisition #143021-25-0309 / Requirements-25-4062

1. Agency and contracting activity. Department of Commerce, Fisheries 0MB Field Delegates and

Northeast Fisheries Science Center I Protected Species Division

2. Description of supplies or services required to meet agency needs (including the estimated value). The National Marine Fisheries Service (NMFS), Northeast Fisheries Science Center (NEFSC), requires specialized satellite imagery preprocessing and machine learning services to support the Geospatial Artificial Intelligence for Animals (GA/A) project. The work will enhance object detection capabilities for identifying whales in high-resolution satellite imagery and expand existing preprocessing tools to accommodate new imagery sources and marine settings. This requirement is sole source based on the unique expertise of , and their prior development of the ortho.py script at the Polar Geospatial Center (PGCJ, which is foundational to the GA/A preprocessing workflow. Their deep experience in geospatial data science, remote sensing, and Al model development, specifically in marine and Maxar imagery contexts, makes them uniquely qualified to perform the work. The planned contract is firm-fixed-price, with a total estimated value of and a performance period from September 1, 2025, to September 1, 2026. No portion of the requirement is for a brand-name item.

3. Identification of the single source or the brand name to be solicited.

Polar Geospatial Center University of Minnesota 90 Church Street SE Twin Cities, MN 55455 https:/lwww.pgc.umn.edu Phone: (612) 624-5457 (PGC main line)

4. Supporting rationale. Only one source or brand name is reasonably available as detailed below:

Compatibility to exjstjng systems or egujpment - The required service must be compatible in all aspects (form, fit, and function) with existing systems, and the source is uniquely qualified to meet the requirement. NMFS' Geospatial Artificial Intelligence for Animals (GAIA) preprocessing workflow relies heavily on the 'ortho.py' script originally developed by the

(Insert PRIRFQ number, as applicable)

Polar Geospatial Center (PGC), which performs core tasks such as orthorectification, projection, and image correction. This script has already been integrated into GAIA's cloud-based architecture and customized for processing Maxar WorldView-2 and WorldView-3 imagery for whale detection. The proposed work involves expanding'ortho.py' to support additional satellite sensors (e.g., Maxar Legion) and to enhance spectral processing tailored for marine environments. These modifications require deep familiarity with the existing codebase and preprocessing logic to ensure seamless integration and continuity of operations. PGC is uniquely positioned to perform this work, as their team authored the original tool and have long-standing experience with GAIA-specific workflows and image sources. Other vendors may offer image preprocessing capabilities, but they do not possess the required access to, or understanding of, GAIA's tailored preprocessing system, nor can they ensure interoperability without significant reengineering.

Attempting to replace or adapt external solutions would risk delays, data inconsistency, and substantial cost increases, as well as potentially break compatibility with research to date on visualization for marine mammals.

5. Market Research.

Market research was conducted to identify vendors capable of supporting GA/A's satellite imagery preprocessing and machine learning needs, with a focus on compatibility with the existing 'ortho.py' script developed by the Polar Geospatial Center (PGC). Reviews of government contract vehicles and vendors such as Radiant Earth and Descartes Labs found no sources with the required expertise in GA/A workflows, Maxar imagery, or Azure-cloud. Only PGC, as the original developer of key tools and long-time collaborator with NOAA, can meet the technical requirements without major reengineering or added risk. A sole source award is justified based on this finding.