A Demand-Driven Software Approach with Dynamic Micro-Expiry and Just-in-Time Processing to Reduce Platelet Wastage in Blood Banks
Published in: IEEE | Official IEEE Link: IEEE Xplore Document #11584332
Contribution Breakdown
Sahil Tomar (dev-S-t) contributed to ideation, problem formulation, SARIMA and XGBoost predictive model development, and simulation software execution.
Key Findings & Numerical Benchmark Results
- Platelet Wastage Reduction: Reduced simulated platelet wastage from 11.2% to 2.5%, representing a ~78% relative reduction.
- Supply Fulfillment: Preserved 99.1% demand fulfillment across blood bank inventory nodes.
- Forecasting Accuracy: SARIMA model recorded the lowest Mean Absolute Error (MAE: 5.85) among evaluated time-series approaches.
- Statistical Validation: Results verified across 30 independent simulation iterations with paired t-tests (p < 1.22×10⁻¹²).
Access Note: The IEEE abstract is accessible publicly via the IEEE Xplore link above. Supplemental institutional repository and preprint links may be made available alongside the main IEEE citation.