Modeling Drug Degradation Kinetics in Smart Polymer Packaging Under Variable Transport Conditions with Real-Time Signal Feedback

Abstract

We conducted a simulation-based physics-informed packaging study in which polymer barrier design, moisture ingress, humidity-responsive signal generation, and degradation kinetics were solved jointly for a thirty-day transport–storage horizon. The simulated cohort contained 900 package histories spanning baseline polymer blisters, barrier-coated packs, and desiccant-assisted packs. Internal moisture transport was driven by time-varying temperature and humidity, while retained potency followed a moisture-coupled Arrhenius law. An embedded sensing trace converted the hidden internal moisture state into a continuously readable signal, and a random-forest estimator used the first seven days of data to infer remaining shelf life. The average shelf life increased from 7.61 d in the baseline package to 8.59 d in the barrier-coated design and 9.73 d in the desiccant-assisted design. The signal-moisture correlation was 0.833; shelf-life prediction achieved RMSE = 0.913 d, MAE = 0.731 d, and R² = 0.857. The study is intentionally narrow: it asks how early signal feedback can quantify moisture-driven stability risk for smart polymer packaging and how material choices shift that risk. Because the work was executed as a completed computational experiment, the paper reports finished trajectories, design comparisons, and prediction results rather than hypothetical outcomes.