Past Performance

Challenge

Program offices fielding aviation, missile, and missile defense software need independent evidence that mission software is safe, secure, and ready for test and fielding, without slowing delivery.

Approach

• Independent verification and validation on systems from PAE Fires, PAE Maneuver Air, and the Missile Defense Agency.
• Static, dynamic, and hybrid software testing.
• Automated DevSecOps pipeline engineering, integrated with existing customer toolchains (CI/CD, scanning, ticketing, and reporting).
• Software safety and airworthiness assessments.
• Configuration and cybersecurity management across multiple integration labs.

Outcome

Weapon system software assurance for the Missile Defense Agency and U.S. Army partner organizations, delivered inside customer toolchains so findings reach engineering teams without a parallel process.

Challenge

AI and machine learning now appear both in adversary weapon systems and in U.S. aviation and missile systems. Evaluating AI threats has been labor-intensive, costly, and slow, and the Department needs a repeatable way to assess both.

Approach

• One repeatable workflow to assess U.S. and partner AI (blue) and to characterize adversary AI (red).
• Provenance and evidence custody from intake, with untrusted artifacts run only in isolated environments.
• Threat-informed scenarios that separate the adversary's plausible access from the evaluation team's access.
• Independent retest of mitigations by someone other than the implementer.
• A fixed and mobile evaluation capability.

Outcome

In progress. E3DICT (Efficient Engineering Evaluation and Development of Counter-AI in Tactical Missile and Aviation Systems) is building a repeatable AI assurance and counter-AI evaluation capability within ARISE (Artificial Intelligence Risk, Integrity, Security, and Evaluation).