Improving supply chain planning for perishable food: data-driven implications for waste prevention

Abstract

Waste in the perishable food supply chain is a challenge that data-driven forecasting methods can tackle. However, integrating such methods in supply chain planning requires development efforts. In this regard, understanding user expectations is the first development step. This study scrutinizes the expectations of a data-driven forecasting method for perishable food. The intended development is a joint initiative of a consortium containing three perishable grocery handling firms. Besides planning expectations, the study identifies and ranks demand-sensing factors that can enable data-driven forecasting for food waste prevention. As the participating firms compete in the same region, horizontal collaboration implications are additionally explored in this context. Accordingly, the study extracts relevant performance measures parallelized to food waste. A two-round Delphi study is used to collect the expectations from a data-driven forecasting method. Individual semi-structured interviews with experts from the initiative firms are conducted in the first Delphi round. Based on the extracted propositions in each interview, industrial experts jointly readdressed and ranked the extracted propositions in the second Delphi round, i.e., focus group workshop. The results reveal that the perishability characteristic emerges as a common expectation in linking supply chain planning with data-driven forecasting. This empirical study contributes to the research on supply chain forecasting and addresses the pertinent aspects of developing data-driven approaches to prevent food waste.