Original papersGrain bin monitoring via electromagnetic imaging
Introduction
The ever-growing world population requires reliable and high-quality resources for food. Grains such as wheat and rice constitute the majority of globally consumed daily food: the total wheat and rice-paddy equivalent produced globally in 2011 were 704.08 and 722.76 million tons, respectively (FAO, 2013). Grains are harvested in large quantities and require safe places for storage, typically large metal containers referred to as bins, or silos, are used. Moreover, current challenges in food production due to scarcity of resources, such as limited land and water, as well as increased weather variability caused by climate change, require enhanced post-harvest management. Grains are usually stored dry with the hope that their characteristics will remain unchanged during storage. Harvested grain may be stored for periods of a few weeks to a few years before being used (Maier et al., 2010, Muir, 1998), and a major concern during this period is the possibility of unacceptable deterioration (Maier et al., 2010, Muir, 1998, Metz et al., 2007). Changes in temperature, humidity, moisture content, and external agents such as insects, birds and rodents, can cause spoilage and mold. These factors can cause losses of 3–10% (developed countries) and up to (developing countries) of harvested grain (Muir and White, 2000). To prevent grain contamination, and to extend viability, grain bins should be monitored periodically. This can be performed either in situ or ex situ. Moisture content and temperature are the most important factors contributing to grain quality. Immunity to different insects and pathogens is also a factor. For instance, insect activities begin at temperatures above 15 °C and mold growth increases at relative humidities over and temperatures between 25 °C and 35 °C (Millls, 1989). Insects and mold in a grain bin can cause small high temperature pockets (known as hot-spots) to form, accelerating deterioration. For example, insect reproduction can result in the creation of pockets of 42 °C, while mold can lead to the production of mycotoxins that can heat up localized regions (Maier et al., 2010, Muir, 1998, Metz et al., 2007). Maintaining uniform temperature and moisture content during storage, as well as early detection of hot-spots so that appropriate action can be taken before major losses occur, are important. Different methods have been developed and applied in grain bin management to monitor grain quality during storage and detect possible contaminations (Maier et al., 2010). The most basic of these techniques is ‘human sensory exposure’ where a person enters a grain bin and inspects the grain visually by taking samples and/or smelling the ventilated air to determine if mold has formed (Maier et al., 2010, White, 2000). This method can be quite time consuming, ineffective for large bins and erroneous as it lacks exhaustiveness and is prone to human error. Also, due to a number of reported fatalities related to farm workers getting trapped upon entry in filled bins, there has been a call from labor unions to bring in regulation to completely seal off bins from human access.
The most common automatic techniques that are widely used for detecting grain spoilage are temperature and moisture measurement. Depending on the size of a silo, one or more temperature and/or moisture sensor carrying cables are vertically hung in each bin (Fig. 1 (Intellifarms, 2009)). The sensors on each cable are typically separated by 1.2 m or more1 (White, 2000, Intellifarms, 2009), and the number of cables suspended in a grain bin depends on several factors including: the size of the silo (especially its diameter), the climatic conditions and the stored commodity (Shelton, 1998). One of the main advantages of this system is its real-time temperature and moisture monitoring capability. However, the typical sensor spacing of 1.2 m on a given cable results in low spatial resolution. Temperature and humidity sensors are locally sensitive, providing detection limited to a radius of 30–60 cm (White, 2000, Intellifarms, 2009, Mills et al., 1989). Grain is a good thermal insulator and a hot-spot far from the sensor can only be detected when it becomes large enough to be within a sensor’s range (Muir, 1998, White, 2000). These cables are also expensive and require reinforcing the ceiling of the grain bin to account for forces exerted on the cables during unloading.
Other automated techniques, such as measuring the CO2 produced by the living organisms inside the bin, or using acoustic sensors to detect sound generated by insects, have also been reported in literature (Muir, 1998, White, 2000, Mills et al., 1989, Maier et al., 2010, Yan et al., 2006). Such techniques, however, have practical limitations to be used in farm-scale bins.
An alternative, automated technique to detect spoilage and mold within grain bins, is to use electromagnetic imaging (EMI). EMI can be adopted for non-invasive estimation of the electrical properties of an object-of-interest (OI). EMI has been extensively used in medical imaging (Meaney et al., 2007, Abubakar et al., 2002, Semenov and Corfield, 2008), security scanners (Sheen et al., 2001), geophysical surveying (Abubakar and Van den Berg, 2000, Abubakar et al., 2008) and industrial non-destructive testing (Zoughi, 2000). Its objective is to obtain quantitative information (such as physical properties) and qualitative information (such as shape and location) regarding an OI and its internal features (Franchois and Pichot, 1997, Joachimowicz et al., 1991). Applied to grain bins, EMI would produce a global map of the electrical properties of grain throughout the bin. The grain industry already infers moisture content, temperature and bulk density of grain through correlation with grains’ electrical properties (Nelson et al., 2000, Yan et al., 2006). EMI offers several advantages over existing methods that include: global sensitivity, the use of low-cost electromagnetic radiation2 and the ability to provide images without disturbing or interacting with the grain. Another feature that makes EMI a better alternative to current grain bin monitoring techniques is its relatively high spatial resolution compared to existing techniques.3 Most grain bin monitoring technologies are affected by grain fines and dust. While these particles, over time, can accumulate and coat sensors in the storage bin, the limited thickness of this coating will generally render it invisible to EMI.
We have developed a prototype EMI system capable of monitoring grain bins for early detection of spoilage conditions. The system uses side-mounted antennas to both illuminate the interior of the bin and to collect electromagnetic field measurements. To these measurements we apply an optimization algorithm known as the finite-element contrast source inversion (FEM-CSI) to produce images. In this manuscript, we present synthetic validation of the proposed system and experimental validation from a small-scale prototype.
The paper is organized as follows: a brief description of electromagnetic imaging is given in Section 2; a synthetic validation of the system is outlined in Section 3; a detailed description of the experimental setup, including data calibration and reconstruction results, is presented in Section 4; and finally, the paper concludes, including a brief discussion of our ongoing work, in Section 5.
Section snippets
A brief introduction to electromagnetic imaging
EMI is an imaging modality that uses active transmitters and receivers of electromagnetic radiation to obtain quantitative and qualitative images of the complex dielectric profile of an OI. Most EMI systems use antennas as transmitters and/or receivers. Due to the propagation and scattering properties of electromagnetic signals at the lower end of the frequency spectrum (including radio and microwave frequencies), multiple antennas are usually placed such that they surround the OI in order to
Synthetic analysis
The software used for synthetic analysis consists of algorithms that solve two problems: an electromagnetic forward problem and an inverse problem. In this part of the study, an assessment was performed to monitor the variation in electromagnetic signals due to the change in grains’ electrical properties. This assessment involved running a number of simulations representing scenarios where the electrical properties change due to the presence of moisture, insect infestation, or mold. These
Laboratory-scale experimental system
For the second phase of this feasibility study a laboratory-scale metallic cylindrical chamber, with r = 15 cm and h = 30 cm, was constructed. This scaled experimental system consists of 24 monopole antennas each 5 cm long, distributed in three layers inside the chamber. The antennas were polarized along the radius of the enclosure () and were connected to a 24-to-2 RF multiplexer/switch that was followed by a vector network analyzer (VNA) analogous to that shown in Fig. 2. The RF switch enabled
Conclusion
Successful detection of grain spoilage using EMI has been presented for both synthetic and experimental data sets. The installation and analysis of an EMI system inside a hopper style grain bin at the CWBCSGR at the University of Manitoba is ongoing. Future work will evaluate EMI grain bin monitoring at scale during grain storage. Studies show that the dielectric properties of materials are also dependent on their temperature level (Ryynänen, 1995). Use of the proposed grain monitoring
Acknowledgments
The authors would like to thank Dr. Amer Zakaria for the access to his FEM-CSI algorithm, Dr. Majid Ostadrahimi for his technical advise and Dr. Fuji Jian for his help at the CWBCGSR facility. The authors would also like to thank Mitacs-Canadafor their financial support.
References (32)
- et al.
Non-linear three-dimensional inversion of cross-well electrical measurements
Geophys. Prospect.
(2000) - et al.
Imaging of biomedical data using a multiplicative regularized contrast source inversion method
IEEE Trans. Microwave Theory Tech.
(2002) - et al.
2.5D forward and inverse modeling for interpreting low-frequency electromagnetic measurements
Geophysics
(2008) - et al.
A 3-d dual-polarized near-field microwave imaging system
IEEE Trans. Microw. Theory Tech.
(2014) - et al.
A contrast source inversion method
Inverse Prob.
(1997) - FAO, 2013. Production. <http://faostat3.fao.org/home/index.html#HOME> July 19,...
- et al.
Microwave imaging-complex permittivity reconstruction with a Levenberg–Marquardt method
IEEE Trans. Antenn. Propag.
(1997) - et al.
A wideband microwave tomography system with a novel frequency selection procedure
IEEE Trans. Biomed. Eng.
(2010) - et al.
On super-resolution with an experimental microwave tomography system
IEEE Antenn. Wireless Propag. Lett.
(2010) - Intellifarms, 2009. Moisture cable. <https://intellifarms.com/solutions/binmanager/Moisture_Cable.html.aug> (23,...
Inverse scattering: an iterative numerical method for electromagnetic imaging
IEEE Trans. Antenn. Propag.
Monitoring carbon dioxide concentration for early detection of spoilage in stored grain
Julius-Kühn-Archiv
Cited by (45)
Digital Twins: A novel traceability concept for post-harvest handling
2023, Smart Agricultural TechnologyIntegrated approach on stored grain quality management with CO<inf>2</inf> monitoring-A review
2022, Journal of Stored Products ResearchCitation Excerpt :The most common methods of grain spoilage monitoring tried by many grain management companies and researchers involve using grain temperature and moisture monitoring, systematic periodic inspection for visible mold, detection of off odors, etc. (Fleurat-lessard, 2015; Gonzales et al., 2009; Kebeli et al., 2000; Wawrzyniak et al., 2013). Attempts have been made to identify the most reliable monitoring technique for in-bin or in-storage grain quality using either continuous monitoring of moisture gradients, temperature variation, air pressure, and gas concentrations, etc. (Asefi et al., 2015, 2017; Gilmore et al., 2017, 2019; Lewis et al., 2019; Maier et al., 2010; Steidle Neto and Lopes, 2015). These monitoring techniques and integrated grain management strategies have been verified through experimental and empirical studies on the safe storage guidelines for each grain type at different storage conditions (Fleurat-Lessard, 2017).
Applications of new technologies for monitoring and predicting grains quality stored: Sensors, Internet of Things, and Artificial Intelligence
2022, Measurement: Journal of the International Measurement ConfederationFlow monitoring by microwave imaging inside a cuboid cavity: Theory and numerical feasibility analysis
2021, Flow Measurement and Instrumentation