What question did this study set out to answer?

This research aims to provide a systematic framework for selecting technologies to reduce food loss and waste while managing uncertainty in stakeholder preferences.

May 15, 2026Open Access

Selecting Food Loss and Waste Mitigation Technologies Under Preference Uncertainty: An Explainable Multi-Criteria Decision Support Framework

Key Points

This research aims to provide a systematic framework for selecting technologies to reduce food loss and waste while managing uncertainty in stakeholder preferences.
Developed an explainable decision support system (XDSS) combining Best–Worst Method (BWM) and Stochastic Multi-criteria Acceptability Analysis (SMAA-2).
Evaluated 100 FLW mitigation strategies across five contextual criteria: geographic fit, product category, food supply chain stage, stakeholder role, and technology type.
Generated probabilistic rank-acceptability profiles to indicate strategy performance under various preference conditions.
Leading alternatives achieved first-rank acceptability levels between 62% and 74%.
Framework provides transparent rankings, enhancing decision-making in uncertain contexts.
The proposed XDSS lays out a methodology base for both future real-world testing and operational use.

Abstract

Food Loss and Waste (FLW) remain major challenges for global food security, environmental sustainability, and economic stability, with nearly one-third of food produced each year being lost or wasted. Although many technologies exist to mitigate FLW, they are often assessed separately, making it difficult for decision-makers to compare options and select solutions suited to specific contexts. This research introduces an explainable decision support system (XDSS) that helps prioritise FLW mitigation strategies while accounting for uncertainty in stakeholder preferences. The proposed framework combines the Best–Worst Method (BWM) with Stochastic Multi-criteria Acceptability Analysis for Group Decision-Making (SMAA-2) to produce transparent and uncertainty-aware rankings. It evaluates one hundred FLW mitigation strategies across five contextual criteria: geographic fit, product category, food supply-chain stage, stakeholder role, and technology type. Rather than producing a single fixed ranking, the system generates probabilistic rank-acceptability profiles that indicate the likelihood of each strategy performing well under different preference conditions. Illustrative scenarios demonstrate that the framework can translate qualitative user preferences into robust prioritisation outcomes, with leading alternatives achieving first-rank-acceptability levels between 62% and 74%. These results indicate that the system can support clearer and more flexible decision-making when preferences are incomplete, inconsistent, or uncertain. Although the current results are based on simulated structured cases, the proposed XDSS provides a transparent methodological foundation for future real-world validation and operational deployment. The framework offers practical value for selecting FLW technologies and for policy planning, contributing to more sustainable food systems and supporting progress toward SDG 12.3.

Read Full Paperexternally

Demander à l'IA

Bookmark

View Full Paper