Social and engineering perspectives on optimal farm management and reliable grain supply chain networks

The growth in food demand urge the need of increasing agricultural productivity and reducing food losses in a sustainable basis. New opportunities for farm management decision making have been rapidly growing with the proliferation of data and information describing agricultural systems. Farm management performance is affected by complex interactions between factors, such as crop yield, market price, culture task schedule, machinery selections, as well as local weather and environmental conditions. Appropriate farm management practices coupling with abilities to obtain real-time local agricultural information with recently vigorous developed information technologies can improve agricultural productivity, reduce losses, and improve farmers’ profits. Also, a better understanding of strength and weakness of grain supply chains provide opportunities to plan a reliable and robust food networks, thereby assisting farm management and reducing post-harvest losses. Thus, the overall objective is developing a framework to support farming decisions that enhance farm management on a sustainable and profitable basis. To bridge existed information gaps, specialized text mining tools are developed to discover real-time agricultural information by utilizing Twitter, which also provides geolocation data with finer spatial resolution. The results showed that social networks contribute more real-time regional crop planting schedules compared to official NASS reports, which can be ahead of time by five days on average at the early stage of planting. We have also identified influential agricultural stakeholders within social networks, based on social network connections of the communities observed within Twitter. The results showed that the connections of online agricultural communities are exceedingly tight and geo-location-based. This will provide new strategies for the development and deployment of targeted community learning modules for enhanced implementation of best management practices. Qualitative and quantitative analytical tools have been developed to provide decision support on farm management practices. A text mining analysis was performed to identify farming schedules and discover key influential factors behind farmers’ operational decisions from news media. The results showed strong site-specific relationships between harvest, grain price, and moisture for farm management. An optimization model, BioGrain, was developed to maximize farmers’ profits by optimizing critical farm decisions including agricultural machinery selection and harvesting schedules. The optimization modeling showed that crop moisture content is critical for optimal farm management. Farmers should balance the tradeoffs between harvestable yield and drying costs to make appropriate decisions when determining the best management strategy. Large farms outperformed small farms on profits but generated higher grain losses, due to a longer harvesting period. The change of corn price would affect optimal farm decision making when adopting on-farm drying, but not for farmers adopting elevator drying. Grain supply chains are inherently complex due to interactions between farms, grain elevators, and several kinds of grain processing facilities. We have developed an optimization model to reproduce the potential grain supply chain flows within the network based on local crop yields and agricultural infrastructure. Given potential grain transportation flows, we then study the network structure and characteristics of the Illinois grain supply chains from global and local topological perspectives. The result shows that the network has scale-free properties and good network features for supply chains. Using modularity and centrality analyses, important subgroups and facilities were identified. The results revealed two primary subgroups located in western and central Illinois. The most important facilities are identified within those regions and should be well maintained to avoid propagation of system failures