Artificial Intelligence replication of District Hospital Oxygen Concentrator Fleet Scheduling Under Unmeasured Sieve-Bed Degradation Sensing Gaps: A Healthcare Robustness-Constrained Evaluation

District hospitals that depend on pressure-swing-adsorption oxygen concentrators face a scheduling problem in which the true remaining life of each sieve bed is unobservable, because the sensing channels that would reveal zeolite degradation are either absent or too coarse to inform dispatch decisions. the present analysis treats that sensing gap as a formal estimation problem and evaluates, through a replication-based replication, how a linear estimator of concentrator output recovers the parameters of a declared data-generating process when the degradation state is latent. Six hundred data sets of N = 500 were drawn from a process with intercept 12.0, exposure 0.42, group -3.1, severity 1.35 and error SD 4.2, and the ordinary least squares estimator was characterised across replications. The estimator recovered the exposure coefficient with negligible bias (-0.00015) and a root mean squared error of 0.01509, while the intercept carried a small positive bias (0.02128) and a root mean squared error of 0.83137; empirical coverage of the nominal 95% confidence intervals was 1.0 for both parameters. The results establish that a linear scheduling model can remain statistically well behaved even when the degradation signal is unmeasured, provided the latent state enters the outcome through the declared severity term. The contribution is a robustness-constrained evaluation protocol that separates estimator behaviour from the sensing architecture, and the paper argues that fleet scheduling under sensing gaps should be assessed by coverage and bias of the scheduling coefficients rather than by point predictions of remaining bed life.