Applying Generative-AI to NASA Documentation and ProcessesThis research and development project leverages generative-AI to assist in the generation of software process documentation based on NASA standards. By utilizing fine-tuned AI models, the proposed system will analyze NASA's software guidelines, helping to translate them into well-structured, compliant process documents. This assistance can reduce the manual effort required to produce such documentation, enhance consistency, and assure alignment with NASA's stringent software development and operational requirements.
In addition to assisting in the generation of software process documentation, the project explores how generative-AI can help create audit checklists as well as assess the compliance of NASA provider documentation against applicable NASA standards. This approach would support the compliance auditing process, providing real-time insights and assessments. The intended result will be a streamlined process, potentially including a Python-based tool and database, that improves audit efficiency, reduces human error, lowers manpower costs and required manhours, and assures continuous compliance with NASA and industry evolving standards for safety-critical software development. Future task might be to investigate the software industry approach and standards for potential collaboration.
Document ID
20240013678
Acquisition Source
Johnson Space Center
Document Type
Poster
Authors
Robert Benorden (Science Applications International Corporation (United States) McLean, Virginia, United States)
Date Acquired
October 29, 2024
Publication Date
December 10, 2024
Publication Information
Publisher: National Aeronautics and Space Administration
Subject Category
Cybernetics, Artificial Intelligence and RoboticsDocumentation and Information Science
Meeting Information
Meeting: JSC Innovation & Ugly Sweater Showcase
Location: Houston, TX
Country: US
Start Date: December 10, 2024
Sponsors: National Aeronautics and Space Administration