Finite expected preparation cost:
ProvedOWMR.fifo_prep_boundapproximation-algorithmsinventoryinventory-theoryoperations-researchrenewal-theorystochastic-control
Let be i.i.d. demand vectors with law on , each coordinate with positive finite mean; the coordinates may be dependent. Fix and reals . Call period a preparation period unless and for every . Then there is a finite constant such that, for all , with set to ,
With the period by which all prescribed pipelines have arrived and the amount by which retailer 's position must fall to reach its target, this bounds the expected preparation length and the cost of the backlog accumulated during preparation (Appendix B, Step 1), which grows at most linearly in cumulative demand.
Preamble
import Definitions.Def_owmr_network_model open MeasureTheory open scoped ENNReal NNReal
Formal statement
namespace OWMR
theorem fifo_prep_bound {n : ℕ} (μ : Measure (Fin n → ℝ≥0)) [IsProbabilityMeasure μ]
(hpos : ∀ i, 0 < ∫⁻ a, (a i : ℝ≥0∞) ∂μ) (hfin : ∀ i, ∫⁻ a, (a i : ℝ≥0∞) ∂μ < ∞)
(T0 : ℕ) (b : Fin n → ℝ) :
∃ C : ℝ≥0∞, C ≠ ∞ ∧ ∀ T N : ℕ, T ≤ N →
∫⁻ z, ∑ t ∈ Finset.range T,
(if T0 ≤ t + 1 ∧ ∀ i, b i ≤ ∑ s ∈ Finset.range t, ((padHist N z s i : ℝ≥0) : ℝ) then 0
else ENNReal.ofReal (1 + ∑ i, ∑ s ∈ Finset.range t, ((padHist N z s i : ℝ≥0) : ℝ)))
∂(Measure.pi fun _ : Fin N => μ) ≤ C := by sorry
end OWMRSource
M. Hu, "A 5/2 Guarantee for Echelon Stock (s,S) Policies in One-Warehouse Multi-Retailer Systems", working paper (manuscript of November 2026), Appendix B (proof of Proposition 2), Step 1, p. E.4: the preparation duration and the backlog accumulated during preparation have finite expected cost under finite first demand moments.