Generative AI Based Demand Forecasting for Third-Party Logistics

The article discusses the growing role of advanced technologies, such as artificial intelligence (AI) and machine learning (ML), in-demand forecasting within the contract logistics sector, particularly among 3PL operators. In the face of the complexity of global supply chains and dynamically changing market conditions, traditional forecasting methods prove insufficient. The article highlights challenges related to demand forecasting, such as volatility and irregularity in supply, as well as the need to integrate modern technologies that enable better data analysis and identification of hidden patterns. Specifically, the traditional ARIMA model supported by ML is compared to modern solutions based on the generative AI (gen-AI) model - ChatGPT, showcasing their impact on forecast accuracy. Case studies of 3PL warehouses demonstrated that the use of AI and ML technology significantly improves operational efficiency, though certain limitations were also noted, particularly in areas with high data variability. The compensation for model illusion was based on ex-post verification of forecast accuracy during studied periods, linked to forecast updates driven by refreshing data sources with new input data. The procedure for creating forecasts is outlined in the overall process map shown in the text. The article emphasizes the need for further development of hybrid forecasting approaches that integrate traditional statistical models with advanced machine learning techniques.