Unplanned downtime in U.S. oil and gas operations compels firms to hold expensive critical spares, yet operators still experience stockouts of consequential items while accumulating slow-moving maintenance, repair, and operations (MRO) inventory. This tension between high service-level requirements and working-capital pressure persists because demand forecasting, criticality classification, and working-capital management are typically treated as separate problems, and evidence specific to U.S. oil and gas contexts is fragmented. This paper proposes an integrated predictive inventory and demand-planning framework that segments items by demand pattern, economic value, and operational criticality; assigns forecasting methods to each segment; converts forecast distributions into criticality-differentiated replenishment policies; and evaluates outcomes jointly on service and financial measures. The study adopts a conceptual, integrative design, supported by an illustrative analytical demonstration based on a hypothetical portfolio of 1,000 stock-keeping units. The demonstration indicates that segment-specific forecasting could improve accuracy relative to a uniform baseline, that criticality-differentiated policies could reduce stockout frequency without requiring higher inventory investment, and that selective rationalisation of slow-moving stock could release working capital. These outputs rest on stated assumptions and do not constitute empirical validation. The paper concludes that operators should adopt the framework in sequence: first establishing data integrity and criticality-informed classification, then piloting segment-level forecasting, and finally embedding forecasts in replenishment policy and working-capital governance.
Keywords: MRO inventory, critical spare parts, demand forecasting, intermittent demand, predictive inventory management