Search for: All records

The United Nations Sustainable Development Goals provide a road map for countries to achieve peace and prosperity. In this study, we address two of these sustainable development goals: achieving food security and reducing inequalities. Food banks are nonprofit organizations that collect and distribute food donations to food‐insecure populations in their service regions. Food banks consider three criteria while distributing the donated food: equity, effectiveness, and efficiency. The equity criterion aims to distribute food in proportion to the food‐insecure households in a food bank's service area. The effectiveness criterion aims to minimize undistributed food, whereas the efficiency criterion minimizes the total cost of transportation. Models that assume predetermined weights on these criteria may produce inaccurate results as the preference of food banks over these criteria may vary over time, and as a function of supply and demand. In collaboration with our food bank partner in North Carolina, we develop a single‐period, weighted multi‐criteria optimization model that provides the decision‐maker the flexibility to capture their preferences over the three criteria of equity, effectiveness, and efficiency, and explore the resulting trade‐offs. We then introduce a novel algorithm that elicits the inherent preference of a food bank by analyzing its actions within a single‐period. The algorithm does not require direct interaction with the decision‐maker. The non‐interactive nature of this algorithm is especially significant for humanitarian organizations such as food banks which lack the resources to interact with modelers on a regular basis. We perform extensive numerical experiments to validate the efficiency of our algorithm. We illustrate results using historical data from our food bank partner and discuss managerial insights. We explore the implications of different decision‐maker preferences for the criteria on distribution policies.

 

Our goal in this work is to build effective yet robust models to predict unreliable and inconsistent in-kind donations at both weekly and monthly levels for two food banks across coasts: the Food Bank of Central Eastern North Carolina in North Carolina and Los Angeles Regional Food Bank in California. We explore three factors: model, data length, and window type. For the model, we evaluate a series of classic time-series forecasting models against the state-of-the-art approaches such as Bayesian Structural Time Series modeling (BSTS) and deep learning models; for the data length, we vary training data from 2 weeks to 13 years; for the window type, we compare sliding vs. expanding. Our results show the effectiveness of different models heavily depends on the data length and the window type as well as characteristics of the food bank. Motivated by these findings, we investigate the effectiveness of employing an average of all predictions formed by considering all three factors at both monthly and weekly levels for both food banks. Our results show that this average of predictions significantly and consistently outperforms all classical models, deep learning, and BSTS for the donation prediction at both monthly and weekly levels for both food banks. 

Food banks are nonprofit hunger relief organizations that collect donations from donors and distribute food to local agencies that serve people in need. Donors consist of local supermarkets, manufacturers, and community organizations. The frequency, quantity, and type of food donated by each donor can vary each month. In this research, we propose a technique to identify the supply behavior of donors and cluster them based on these attributes. We then develop a predictive ensemble model to forecast the contribution of different donor clusters. Our study shows the necessary behavioral attributes to classify donors and the best way to cluster donor data to improve the prediction model. 

Food banks operate with an objective to serve as many of food-insecure people as possible with the limited supply available to them. This paper presents a mixed-integer programming model to identify the efficient assignment of demand zones (counties) to distribution centers (branches) and equitable allocation of donated food from the food bank branches to the demand zones. The model objective function minimizes the total cost of branch operation, the cost of receiving and distributing food, the cost of undistributed food while maintaining the maximum allowed deviation from perfect equity. Data from the Food Bank of Central and Eastern North Carolina (FBCENC) are used to characterize the major attributes controlling the food distribution system of a food bank. Results from the optimization model using FBCENC data show that the optimal allocation under perfect equity follows a particular structure depending on the shipping cost and the cost of undistributed supply. Sensitivity analyses exploring the trade-offs between efficiency and effectiveness as a function of the cost of shipping, truck capacity, and a user-specified maximum inequity cap show that marginal sacrifice in equity can significantly improve effectiveness. The corresponding improvement in effectiveness is greater when comparatively larger trucks are used and the cost of shipping is relatively higher. The analyses also suggest that while efficiency is less sensitive to the allowable limit on the deviation from perfect equity, it is sensitive to truck size. A comparison of direct shipping to branches to operating a local hub suggests the former option to be more cost efficient. 

Hunger relief organizations are mostly non-profits that collect food from various sources and redirect them to the people in need. This is to combat the prevalent food insecurity affecting children, the unemployed, students, seniors and so on. Previous research has focused on the demand/donation side of food rescue operations, but the distribution or supply side - especially in reducing the uncertainty associated with food demand - has received significantly lower attention. In this study, we obtained data from a local hunger relief organization, specifically a food pantry to develop estimates of the demand they expect to receive in the future. To do this, we fit the growth of the food pantry client population to a logistic growth model to obtain a good fit. We then obtained data for frequency of visits to develop estimates of the number of visits expected in the future, using time series models. This will be combined with the allocation policy for food distribution to develop estimates of true demand. This study has merit for hunger relief organizations. It will aid decision making relative to food distribution, while also providing data for planning purposes. 

Hunger and food insecurity are present in each American county. Government and non-government organizations are working to address food insecurity in the United States. Food banks are nonprofit hunger relief organizations that collect food and monetary donations from donors and distribute food to local agencies which serve people in need. Contributions come from retail donors, communities, and food manufacturers. The uncertainty of donation amounts and frequency is a challenge for food banks in the fight against hunger. In this research, we analyze local food bank donation data and propose a predictive model to forecast the contribution of different donors. Our study shows the necessary behavioral attributes to classify donors and the best way to cluster donor data to improve the prediction model. We also compare the accuracy of prediction for different conventional forecasting techniques with the proposed Support Vector Regression (SVR) model. 

Non-profit hunger relief organizations rely on the goodwill of donors for their in-kind cash, food donations and other supplies to alleviate hunger, reduce human suffering and save lives. However, these organizations struggle with changing demand and supply patterns, disruptions caused by very low donations even though they must make strategic distribution decisions. Food distribution forecasts based on times series models can be useful for these decisions. Yet, it is plausible that food distribution by hunger relief organizations (and demand by the people in need) are driven by certain underlying factors. In this research, we used Visual Analytics (VA) to study the effect of certain underlying factors on the forecast generated for food distribution to the aid recipients. Specifically, we used already tested forecasting techniques to predict the expected quantity of distributed food for the underlying factors identified. 

In the fight against hunger, Food Banks must routinely make strategic distribution decisions under uncertain supply (donations) and demand. One of the challenges facing the decision makers is that they tend to rely heavily on their prior experiences to make decisions, a phenomenon called cognitive bias. This preliminary study seeks to address cognitive bias through a visual analytics approach in the decision-making process. Using certain food bank data, interactive dashboards were prepared as an alternative to the customary spreadsheet format. A preliminary study was conducted to evaluate the effectiveness of the dashboard and results indicated dashboards reduced the amount of confirmation bias. 

Food insecurity affects more than 41 million people annually in the United States. Within the Feeding America network, approximately 200 food banks are working throughout the US to serve people in need with donated food. Satisfying hunger need of food insecure people with limited supply is a challenge for these food banks. A numerical study is performed on data from Food Bank of Central and Eastern North Carolina (FBCENC) to capture the major attributes controlling its food distribution system. FBCENC seeks to distribute donated food equitably so that each service area (county) receives food proportional to its demand while minimizing the undistributed food donations. In addition to seeking equitable and effective food distribution policies, FBCENC wants to identify distribution branches to maximize the accessibility of the counties to donated food. An assignment and distribution model is developed to minimize the cost of maintaining a user-specified cap on the maximum inequity in food distribution. A sensitivity analysis between the user-specified maximum inequity cap and effectiveness shows the effectiveness of donated food distribution can be improved significantly by sacrificing equitable distribution slightly.