SATPC0036299 Tab 06 RDSS..pdf

Page 1 V2.0 3-15-2017

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION

NASA SHARED SERVICES CENTER

RECOMMENDATION AND DETERMINATION TO SOLICIT FROM ONE SOURCE

OR BRAND NAME

PICK ONE

Sole Source:

I recommend that NASA, NASA Shared Services Center negotiate with Georgia State University only for Solar Dynamics Observatory (SDO) AI Helio Foundation Model (HelioFM) for Improved Space Weather Forecasting. The total estimated cost of this effort is and the estimated period of performance or lead-time for delivery is August 15, 2024 through February 14, 2026.

OR

Brand Name Justification:

I recommend that NASA, NASA Shared Services Center negotiate with only with vendors who can provide the subject band name from Manufacture Name for Enter Supply or Service to be procured. The total estimated cost of this effort is Enter Dollar amount and the estimated period of performance or lead-time for delivery is Date.

This recommendation is for the acquisition of supplies or services determined to be reasonably available from only one source/only one manufacture. Competition is impractical for the following reasons (please answer the questions below):

1. What are the reasons for soliciting only the suggested source/suggested brand name? Please provide the unique characteristics of the items or services along with the supporting rationale and the market research performed to substantiate your reason for the suggested source/brand name.

The design and development of "AI Foundation Models for Science", led by the NASA Inter-agency Implementation and Advanced Concepts Team (IMPACT) at the Marshall Space Flight Center (MSFC), are being conducted as an open, collaborative effort grounded in NASA SMD's principles of Open Science [https://science.nasa.gov/open-science/]. As such, various organizations are participating in this effort in specific areas and at different levels. The Data Mining Lab (DMLab) within the Department of Computer Science at Georgia State University has been a part of this open collaboration, providing their specific science expertise to the team. This solicitation seeks to focus their current efforts to address specific IMPACT application needs. The DMLab supports research projects related to broad areas of data mining and knowledge discovery from solar big data, including computer vision, indexing, data management and integration, deep learning, frequent pattern mining, and time series mining dedicated to understanding solar phenomenon prediction. Due to the complex nature of the interactions associated with solar flares and coronal mass ejections, building a model to improve Space Weather Forecasting remains a significant scientific and technological challenge. To address these difficulties, the DMLab team at Georgia State University uses artificial intelligence (AI) and deep machine learning techniques, among others, to improve these models. AI Foundation models (FM), in particular, provide flexibility for users to pretrain a single model and then customize it for specific downstream tasks like Space Weather Forecasting. The DMLab team will provide research and science expertise to support the development of a Foundation Model in Space Weather Forecasting applications. Their highly specialized data science techniques in data mining and knowledge discovery from solar big data and deep machine learning techniques are required to meet the development and research requirements needed for the design of novel downstream science use cases to test and evaluate the model. Specific use cases related to Space Weather Forecasting are of significant value to their research project and science community.

2. What is the impact to the government if another vendor/brand is used? Include any dollar, schedule, or compatibility impacts if possible.

Page 2 V2.0 3-15-2017

Due to their current level and specific areas of involvement, no other vendor provides the subject matter expertise required to develop specialized Space Weather Forecasting use cases for the development and evaluation of a Foundation Model. Therefore, the schedule and fiscal ramifications of finding a suitable replacement for this requirement are not feasible.

ADDITIONAL COMMENTS:

Background/Rationale: Artificial intelligence (AI) model development is playing an important role in advancing our understanding of science. Foundation models (FM), in particular, provide flexibility for developers to pretrain a single model and then customize it for specific downstream tasks. The NASA IMPACT team is developing a foundation model for Space Weather Forecasting applications (commonly referred to as HelioFMs) and is seeking support in designing, building and evaluating the model to advance and explore a wide range of tasks within the field of Space Weather (i.e., Forecasting).

8/7/2024

Technical User Name Date

Title