Reduction of Food Wastage in a Hotel Staff Cafeteria through an Integrated Behavioral and Technological Intervention: A Quasi-Experimental Study in Southern Sri Lanka

Background: The hospitality industry has to deal with significant economic and environmental losses due to food wastage, which is especially aggravated in resource-constrained settings. Therefore, there is a need to design and adopt effective abatement measures. The purpose of this study was to assess an integrated intervention based on behavioral-technological interventions that would help decrease food wastage in a hotel staff cafeteria to gauge its effects on the mass, composition, and economic-environmental outcomes of food wastage.
Materials and Methods: In a Sri Lankan resort, a quasi-experiment was conducted over the course of 28 days. A multi-faceted intervention, which included immersive Virtual Reality (VR) training, real-time Internet of Things (IoT) feedback dashboards, and standardized portion-control tools, was adopted after 14 days of the baseline. The gravimetrically measured daily food wastage was analyzed using segmented regression in a quasi‑experimental pre‑post design of daily measurements.
Results: The intervention resulted in a 26.9% decrease in food wastage (p=0.020), and all meals had significant decreases. Avoidable food wastage decreased by 41.0%, which is closely associated with decreased variation of portions (r=0.71). Projections indicated that on an annualized basis, savings of 613,495 LKR and 0.75 tons of CO2e (CO2 emission) have been mitigated, as well as enhanced wastage sorting. While these findings demonstrated the potential of a multi-component approach, the results are limited to a single-site, short-duration study. Peak service gap, which was observed in high-volume service, would indicate a constraint and a goal of incorporating behavioral and operational solutions.
Conclusion: This study presents a tested framework for reducing food wastage in hospitality staff cafeterias and demonstrates that hybrid behavioural-technological approaches are effective in resource-constrained settings. However, the findings are based on a single site and a short period, limiting the generalizability.