Predictive Inventory Buffer Optimization for High-Variability Manufacturing Supply Chains Using Hybrid Statistical and Simulation Models

Modern manufacturing systems frequently experience production interruptions due to supplier variability, transportation uncertainty, fluctuating demand patterns, and inaccurate inventory buffer allocation. This research proposes a hybrid quantitative framework integrating stochastic demand forecasting, Monte Carlo simulation, and multi-echelon inventory optimization to improve safety stock allocation and supply continuity in complex manufacturing environments. The study will develop a predictive inventory planning model capable of dynamically adjusting buffer inventory based on supplier lead-time variability, historical disruption frequency, production criticality, and operational demand volatility. The methodology combines statistical forecasting models with discrete-event simulation to evaluate inventory positioning across manufacturing and distribution networks. Optimization algorithms will be applied to minimize total inventory holding costs while maintaining target service levels and production continuity. The proposed framework will be validated using manufacturing and procurement datasets from high-volume industrial operations involving make-to-order and mixed-mode production systems. Performance indicators such as stockout frequency, order fulfillment rate, production downtime, and inventory carrying costs will be quantitatively evaluated. The research aims to establish a scalable methodology for resilient manufacturing supply chain planning under uncertain operational conditions.