Abstract
In the field of both traditional and Agile systems engineering, engineers manually perform a variety of tasks to derive, decompose, trace, rewrite, and evolve large numbers of engineering artifacts – work that is time-consuming and prone to human error. Mistakes made during the project lifecycle have compounding negative impacts on product efficacy and remediation costs grow exponentially problems go unchecked. Thus, improving requirements work through the adoption of modern technologies and methodologies is of significant importance to increase turnaround time and quality of products.
In 2012, Army Training and Doctrine Command (TRADOC) Combined Arms Center for Training (CAC-T) and the MITRE Corporation developed software to accelerate common and repeatable engineering tasks to improve the quality of capability development and acquisition working group efforts. Recent advances in artificial intelligence (AI) natural language processing (NLP) have modernized this software into a web application known as the Requirements Analysis using Artificial Intelligence for Mapping (RAAM), which provides capabilities for the automation of language-related tasks such as generating requirements, improving requirements quality, and performing traceability analysis. In the past two years, the tool has been used by various government organizations to accelerate initial requirements generation by 9,000%, reduce traceability analysis times by 84%, and to improve requirements quality based on International Counsel on Systems Engineering (INCOSE) and Institute of Electrical and Electronics Engineers (IEEE) writing standards.
This paper will discuss several distinct stages of the digital engineering workflow that involve high impact but tedious manual work. It will propose how certain NLP technologies and techniques can be used to accelerate those tasks and improve quality of the outcome through software automation. To further illustrate these capabilities, the paper will focus on the RAAM tool as a case study for real-world application of NLP to digital engineering.