Software Engineering Institute (SEI)
This project focuses on two main research thrusts with known military applications: (1) Software Engineering, Systems Verification and Validation, and Mission Assurance (formerly Mission Assurance); and (2) Information Assurance. SEI research focuses on the most significant and pervasive software challenges within the DoW, such as the scalability and reliability of software assurance, supply chain risk management, validation of and trust in autonomous systems, human-computer and human-technology teaming and interaction, computing and communication at the tactical edge, and efficiency and performance of acquisition strategies and software development appropriate for a contested cyber environment.
FY2027 planned work Use hybridized AI methods to generate prototype system architectures and use automated graph matching verification and validation methods to check correctness of system design.
FY2026 to FY2027 change The decrease of $2.564 million between FY 2026 and FY 2027 reflects the Office of the Under Secretary of War for Research & Engineering (OUSW(R&E)) Internal Realignment to National Defense Education Program.
FY2026 plans — current year Leverage AI generated software methods to automate software systems composition. Validate the results of the systems composition using statistical and graph theoretic methods including model-based systems engineering.
FY2025 accomplishments Integrate techniques in automated learning for system measurement, software Development and Operations, and model-based systems engineering for an automated assessment, modeling, and software deployment process. Focus on strategies for resilience and mission assurance in large complex infrastructures and develop prototype systems that can be transition and tested into DoD applications from cloud to embedded systems.
FY2027 planned work Map AI methods for software infrastructure generation to optimal correctness and system performance regions so there is a taxonomy of the best AI strategies for generating differently system architecture designs.
FY2026 to FY2027 change The increase of $0.002 million between FY 2026 and FY 2027 reflect minor budget fluctuations.
FY2026 plans — current year Characterize the effect of training data sets on automated software system generation and develop methodologies to show how software integration and information service composition can be improved and validated through training data curation.
FY2025 accomplishments Enable combined distributional machine learning for risk analysis between software, machine learning, and cyber security to enable assessment and management of automated systems. These risk metrics will be introduced to a variety of DoD applications from system assessment, to enterprise cloud analytics, and legacy embedded systems.