nx_maint_rul.nx

nx_maint_rul.nx -- remaining-useful-life (RUL) PROGNOSTICS: the learning loop's prediction. From an asset's health-metric HISTORY (the R4 ledger), estimate the wear rate (first-order system identification) and PREDICT time-to-failure, so the platform acts BEFORE failure -- proactive, not reactive. Operator (2026-06-23): "full learning loop ... close the feedback loops on physical things for wear and tear." Researcher-grounded (knowledge/fetched/ maint_c_prognostics, maint_c_condmon, maint_e_sysid, maint_e_bathtub): prognostics/RUL estimate condition + remaining life from the trend. THE EXCEED: OBD / Sense report the PRESENT (reactive -- "it's broken now"). We predict the FUTURE from the trend (proactive -- "it will hit failure in ~N"). Augury does this at enterprise price/cloud; we do it sovereign + cheap + broad. Method (integer/no-float): wear_rate = decline / span (first-order); RUL = (current - failure_threshold) / wear_rate. A flat/improving trend -> STABLE (no predicted failure). Honest: first-order linear extrapolation; nonlinear degradation (bathtub end-of-life knee) + Kalman state estimation = next. NEVER-BRICK (#26): reads a metric series, predicts; no syscalls, no writes. genealogy_id: project-maintenance-platform-sclass-2026-06-23 (RUL prognostics; researcher-grounded) license_tier: ORIGINAL