Tackling pharmaceutical waste with new AI demand forecasting model

Takeda Pharmaceutical has launched a new artificial intelligence (AI) model to forecast demand for its medicines in Japan, in efforts to optimise the company’s supply chain, strengthen stable supply, and reduce the disposal of expired pharmaceuticals.

The pharmaceutical supply chain faces unique challenges. Sudden shifts in demand, caused by factors such as infectious disease outbreaks or changes in treatment protocols, must be managed alongside strict quality control and product expiry dates. An overestimation of demand leads to the waste of expired products, while an underestimation creates the risk of shortages, preventing patients from receiving necessary treatment.

For Takeda Pharmaceutical demand forecasting previously relied on a combination of statistical analysis of past sales and the knowledge of staff. The new AI model can analyse vast amounts of data to identify non-linear trends and complex correlations, enabling more accurate and comprehensive predictions. The system incorporates human experience and knowledge into the final decision-making process. This approach increases the frequency of forecast updates and allows for greater flexibility in production planning, creating a more agile response to sudden market changes.

The company will first apply the model to products with high shipping frequency and sufficient historical data for the AI to learn from. It plans to expand the range of applicable products as more data is accumulated. Optimising inventory is expected to yield both environmental and financial benefits by reducing waste and improving cash flow.