FSMM: Food System Monitoring Model for Re-optimization

Wenjun Wu, Dong Liu, Lai Jiang, Kai GuoJournal of Industrial Engineering and Applied Science 2024 · DOI

Abstract

The instability of the global food system has aroused growing attention worldwide. Although there is sufficient food produced to feed every person, many people suffer from hunger, and current systems are often environmentally harmful. The Food System Monitoring Model (FSMM) provides a score-based evaluation using entropy and subjective weighting to assess the state of food systems across 178 countries. By measuring efficiency, profitability, sustainability, and equity through nine individual features, the model identifies key drivers of system health. Using Gray Forecast and Boosting Regression models, the research predicts transition timelines to sustainable states and identifies energy use as a primary lever for re-optimization.

Introduction: The Paradox of Global Food Production

The global food system is currently caught in a paradox: we produce enough calories to feed the world's population, yet hunger persists, and the methods used to achieve this abundance are often environmentally destructive. Most existing models prioritize efficiency and profitability, leading to a "race to the bottom" regarding soil health, emissions, and social equity. The FSMM addresses this by shifting the focus toward equity and sustainability, identifying which external factors—such as energy use or government policy—drive improvements across diverse economies.

Methodology and Data Integration

The FSMM employs a sophisticated stack of mathematical models to analyze longitudinal data from 2000 to 2019 for 186 countries, drawing from the FAO, World Bank, and IDMC.

  1. Dual-Score Evaluation: The researchers developed two primary scores: $Y_1$ (Efficiency/Profitability) and $Y_2$ (Equity/Sustainability). Entropy weights were used to objectively determine the importance of nine features, including Food CPI, emissions ($CO_2$, $CH_4$), and displacement numbers.
Score ranges for Efficiency/Profitability and Sustainability/Equity metrics
  1. Indicator Correlation: To understand the relationships between external factors and food system health, the model examines correlations between indicators such as Food CPI, PPI, and net export values.
Correlation heatmap of food system indicators including Food CPI and PPI
  1. Predictive and Causal Modeling: A Gray Forecast Model (GM(1,1)) was applied to estimate the "time to optimality" for each country. Furthermore, an XGBoost-style Boosting Regression model identified the importance of 10 external indicators, achieving a 90% fit for the sustainability-oriented scores.

Results: The Energy Pivot and Transition Timelines

The most striking finding is that shifting the goal from profit to sustainability makes Total Energy Use the dominant factor. In the re-optimized model, energy importance jumped to 26%, suggesting that food system reform is fundamentally an energy transition problem.

The model also highlights a significant divide between developed and developing nations. Sustainability in developed countries is heavily influenced by renewable energy consumption, whereas developing countries' scores are more closely tied to social factors like infant mortality and hunger levels.

Regarding timelines, the Gray Forecast model identified "model citizens" like Iceland and Indonesia, which could reach sustainability in under 8 years. However, for the majority of nations, the transition is much slower, with estimates suggesting a 100-year climb under current standards.

Conclusion and Future Directions

While the FSMM provides a robust framework for national-level analysis, it faces limitations in granularity. Currently, it treats each nation as a single unit, potentially masking regional disparities. Future work will focus on applying the FSMM to smaller, local food systems and integrating more nuanced socio-cultural data to refine the "Equity" score. Despite these constraints, the model offers a clear roadmap: optimizing plant-based yields (vegetables and wheat) and transitioning energy use are the most efficient paths toward global food system sustainability.

Summary updated

Abstract

The unstability of global food system have aroused growing attention worldwide. Althrough there is sufficient food produced to feed every person, many people in the world are suffering from hunger. Moreover, the current food system is harmful to the environment. As our global population continues to rise, the ability to produce more food while sustaining, and even improving, the health of our environment counts. The FSMM(Food System Monitoring Model) provides score rank using entropy weight and subjective weight to evaluate the current state of the food system for most countries in the world. It contains 9 individual features to measure multiple aspects, including efficiency, profitability, sustainability and equity. Considering the future trend of the food system, we apply Gray Forecast Model to forecast the next 5 years of the development of the food system to figure out the time for each country to change into equity and sustainability. To find out the relation between external factors and food system, The FSMM uses Boosting Regression Model to find the correlation between the scores and 10 indicators which represent the influences of energy use, government, agriculture, social factors and population ratio. The FSMM fits the data from 178 countries from 2000 to 2019. After simulating and adjusting the parameters to decrease the loss of the model, we gets the importance rate of each indicator. We find the indicators whose importance rate change significantly when the food system is optimized for equity and sustainability. And we use these indicatos as dependent variable and apply Linear Regression to fit by using the yields of the wheat, meat and vegetables as independent variable. By doing this, We find out the direct impact of re-optimization on the food system. We finally get the importance rate and the coefficient of two regression model and analyze the specific influence of various factors on the food system and compare the differences between the influence on developing and developed countries. After the overall analysis of the influence on developed and developing countries, we pick some specific developing and developed countries to analyze and discuss the scalability and adaptability of the model.
BibTeX
@article{wu2024fsmm,
  author = {Wenjun Wu and Dong Liu and Lai Jiang and Kai Guo},
  title = {{FSMM: Food System Monitoring Model for Re-optimization}},
  journal = {Journal of Industrial Engineering and Applied Science},
  year = {2024},
  doi = {10.70393/6a69656173.323433},
}
Figure