NASA Logo, National Aeronautics and Space Administration

+NASA Home

+Ames Home

Robust Software Engineering-Led Group Wins Digital Transformation Hackathon Award for "Most Potential NASA Impact"
Intelligent Systems Division Banner

Robust Software Engineering-Led Group Wins Digital Transformation Hackathon Award for "Most Potential NASA Impact"

Robust Software Engineering (RSE) members led a multi-Center team in the NASA Digital Transformation (DT) Hackathon on the challenge topic “Natural Language Processing for Requirements Checking” and were awarded “Most Potential NASA Impact”. The team, Semantic Models for Analyzing RequiremenTs (SMART), surveyed and demonstrated tools currently being developed by NASA (including the RSE-developed Formalized Requirement Elicitation Toolset, or FRET) to extract and analyze NASA project requirements from unstructured governing documents to identify if project requirements are correct, consistent, and complete. The multi-Center team was led by Nathaniel Benz (RSE) with significant contributions from Anastasia Mavridou (RSE), as well as representatives from Glenn Research Center (GRC), Goddard Space Flight Center (GSFC), the Jet Propulsion Laboratory (JPL), Johnson Space Center (JSC), and Langley Research Center (LaRC).

The team has been invited to a one-hour Transformation Tuesday, where representatives from the four winners will briefly present their projects for greater awareness. The architecture design proposed by the SMART team will also be submitted as a proposal for follow-on DT funding for FY21.

BACKGROUND: Large-scale projects have a huge number of documents and a relatively small team to review them. Time constraints and average reading speeds make it difficult for humans to find duplicative and conflicting requirements. Missing requirements related to safety or critical interface requirements may not be found at all because different people review various documents. Applying Natural Language Processing (NLP) to review a collection (corpus) of requirements documents can solve this and does NOT require any changes to current practices or documents. An NLP system can read thousands of pages per second and extract an ontology from a corpus, which defines terminology and specifies taxonomies. The system developed will parse each requirement as it is entered into the system to determine from the corpus of knowledge that requirements are not duplicated, conflicting, possibly missing, are classified by subjects (safety-criticality, avionics, thermal… ), and can be traced. Requirements for projects drive the design and ensure projects achieve their objectives. An NLP system reduces labor for tracking requirements between disciplines, classifying safety-critical requirements as well as requirements by other subject categories, and standardizes information for developers and testers.

NASA has been in the strategy and planning stages of the Digital Transformation Initiative since late 2017. The overall goal of the Initiative is to enhance mission success by engaging digital culture and technologies to modernize our workplace, workforce, and work products. This hackathon is part of NASA's initial DT implementation plan. The goals of the DT hackathon are to surface as many DT ideas as possible, build NASA's DT community, provide hands-on DT education, and demonstrate rapid application of DT to NASA challenges. NASA@WORK is an internal, Agency-wide platform that provides NASA employees with an unconventional and inventive way to share knowledge and advance projects across Centers.

NASA PROGRAM FUNDING: Office of the Chief Technologist (OCT)

TEAM: Nathaniel Benz, Anastasia Mavridou, Johan Schuman, Maged Elaasar (JPL), Svetlana Hanson (JSC), Jitin Krishnan (GSFC), Karsten Look (GRC), and Ted Sidehammer (LARC)

POINT OF CONTACT: Nathaniel Benz,

First Gov logo
NASA Logo -