Field Notes
What we think about delivery, mission systems, open source, and the work that reaches the flight line.
DAF CMSO MS Enterprise Governance Is the Hard Problem
The DAF Chief Modeling and Simulation Office oversees a large M&S enterprise with many organizations and major investments. The hard engineering risk is not one model; it is whether shared data, scenarios, and assumptions stay governed.
The DoW AI Strategy Is an Integration Problem
The Department of War AI strategy emphasizes AI-first warfighting and pace-setting projects. The engineering problem underneath is integration: data, authority, verification, monitoring, and mission-thread behavior.
Software Acquisition Policy Now Points to the Harder Layer
The 2025 software acquisition policy push makes the Software Acquisition Pathway the preferred option for software components in business and weapon systems. The harder work sits below policy: architecture, interfaces, and evidence.
X62A VISTA Shows the AI Verification Gap
Have Remy used the X62A VISTA to test autonomous missile-evasion agents trained in simulation. The hard part is not only the AI model; it is proving that simulated behavior remains valid in live flight.
DARPA Resilient Software Work on MQ-9 Needs Scope Discipline
DARPA's resilient software work applies formal methods to MQ-9 operational software. The credible path is not proving every legacy layer at once, but choosing high-risk boundaries, partitioning trusted cores, and producing usable evidence.
The B-21 Refueling Receptacle Shows Interface Discipline
The first upper-surface photos of the B-21 show more than a visible refueling detail. A receptacle is an interface contract, and the same interface discipline applies to software, mission systems, data links, and test evidence.
Analytics Summit 2026: Model Validity Is the Hard Question
The 2026 DAF Modeling, Simulation and Analytics Summit points toward digital training, multi-domain environments, and decision support. The hard engineering question is still model validity: why should a program trust the simulation?
Software Modernization Still Comes Down to Legacy Programs
The FY2025-2026 DoD software modernization plan pushes cloud, software factories, and resilience, but the hardest work remains inside legacy programs. Moving those systems takes interface mapping, test evidence, and migration judgment.
Iron Bank DevSecOps Helps, But It Is Not the Architecture
Iron Bank DevSecOps practices give DoD teams hardened container images, scans, SBOMs, and risk evidence. That reduces supply chain friction, but a vetted image is still a building block rather than a mission-system architecture.
Distributed EW Is an Architecture Problem
Distributed EW sounds clean in a briefing, but the real problem is architecture. Electromagnetic effects have to be coordinated across platforms, bands, timing windows, control paths, and vendors that were not designed as one system.
DoD Sherpa Embedded Expertise Is the Point
The Air Force sherpa model puts software development experts inside mission teams so secure delivery decisions happen where the work is done. Embedded expertise is not overhead; it is how difficult programs avoid late technical surprises.
AWS DevOps Agent Changes the On-Call Rotation
AWS DevOps Agent is generally available with preview customers reporting lower MTTR, faster investigations, and stronger root-cause accuracy. If those numbers hold in mixed environments, on-call economics start to change.