Monitoring Grain Bed Conditions during Truck Transport of Soybeans in Brazil

Abstract

Abstract. Reducing losses in the grain production system is of great interest for society and the scientific community given the anticipated increases in global food demand in the near future. Losses in grain production can be identified in several stages of the supply chain, including pre-harvesting, harvesting, transportation and postharvest handling and processing. Loss estimates at each stage are often fraught with assumptions and the data are often collected by non-representative means. Truck transport is commonly used worldwide for the distribution of goods for trade. In Brazil, truck transport is usually the most economical way to provide distribution of goods in places where inexpensive (e.g., railways) or natural (e.g., ports, rivers) transport alternatives are not available. Truck transport plays a significant role in moving raw materials and processed products from the agricultural-rich north central regions of Brazil to the port cities in the south. Current estimates of grain loss during truck transport are reported to be between 1-5% of a typical load of grain (27,000 kg) and losses stem from inadequately drying the grain prior to transport and the use of non-grain trucks to haul grain for distances upwards of 2,500 to 3,000 km. For soybeans, these long hauls may take up to 2-3 days during the rainy season and the grain can go “out of condition” once it reaches the port. To assess the condition of the grain during truck transport, a wireless instrumentation system was designed, constructed, and tested to monitor the temperature, relative humidity, and carbon dioxide levels in the grain bed. Laboratory testing of the design was conducted, including individual sensor evaluation in both static and dynamic environments, a demonstration of entire system functionality, and pre-implementation testing of a power regulation system to be applied to future field work. Results showed that the monitoring probe construction did not hinder CO 2 measurement and demonstrated the function of the probe and power regulation system. The system was shown to have a CO 2 sensor accuracy of 8.5 % of reading and a diffusion gradient from inside probe chambers to grain bed of 6 % with a response time lag caused by probe housing of about 5 min. After assessment, the probe and its wireless data acquisition system proved ready for the field.