Preserving critical-infrastructure knowledge
A policy package · Brad M. Lindsey — Master Electrician (DoD/DHA) · Master HVAC Tech · independent engineer
A tradesman’s case for provenance-governed AI to keep critical-infrastructure knowledge from walking out the door at every retirement and contract turnover.
The package
The full argument: federal assets deteriorating while the workforce retires, the invisible knowledge loss, why governed AI is the trustworthy answer, and a low-cost federal pilot. The ask is permission to demonstrate — not an appropriation.
The problem, the ask, the pilot metrics, and the cost posture — on one page for a busy desk.
A single-facility, low-cost demonstration with success criteria set in advance and results published either way.
Three governed-AI behaviors — a safe first move, a refusal to guess, a contradiction caught — each paired with the metric a pilot would measure.
What this is
The United States is losing operational control of its own critical-infrastructure knowledge. Federal building repair backlogs more than doubled to $370 billion in seven years, and the skilled-trades workforce that holds the operational knowledge is retiring faster than it can be replaced. Every retirement and every contract turnover walks years of undocumented troubleshooting expertise out the door.
AI can now capture that expertise in the field — but only if it can be trusted near safety-critical systems. That trust is the whole point of the Lindsey Provenance Discipline: AI whose every claim is traceable to its evidence. This package makes the case and proposes a low-cost federal pilot to demonstrate it.
Written as an independent practitioner — no employer or specific facility is named or implicated. Every macro figure is public federal or industry data, cited to a primary source; pilot outcomes are framed as targets to measure, not results achieved.
Cite the package: Zenodo DOI 10.5281/zenodo.20519245 (all versions) · 10.5281/zenodo.20519246 (this version) · CC BY 4.0.