Conference Agenda

Tribo-electrostatic separation (TES) as a dry fractionation technique for the separation of protein was performed in this work. The experiments were carried out using a custom-built solid particle dispenser, coupled with a triboelectrostatic separator, with the latter comprising of a cubic chamber with two electrodes on which the separated particles were deposited. Both cereal and protein flours of known particle sizes were tested with the electrostatic separator via the particle dispenser with air as the dispersing medium. Using a tribocharging tube, Faraday cup, and an electrometer, the type of charge (positive or negative) acquired by the components of the flour sample and their chargeability in coulombs per unit mass were established prior to separation. A high voltage generator was utilized to supply the electric field required to ensure proper separation within the separator. The effects of electric field strength (100 – 200 kV/m) and air flow rate, varied between 5 – 15 l/min on the yield, protein enrichment and separation efficiency were explored. Scanning Electron Microscopy (SEM) analysis of the structure was carried out on the pre- and post-TES materials which further confirmed high separation efficiency of the process. The positive results from these experiments prove that the TES technique is a viable means for the commercial fractionation of both cereal and pulse flours without the drawbacks associated with existing wet fractionation techniques. Further experiments comparing the functional properties of the post-TES with the pre-TES materials are currently underway to explore the effect of this technique on these and other feed materials.

Peanut is the one of the top 10 allergens that is causing product recalls in Canada. Current strategies to identify such allergens are mostly reactive and often compounded by the delays in sampling, laboratory testing, decision making and traceability challenges. The current study focuses on using NIR spectrometry to identify the peanut allergen contamination by automating data collection, apply machine learning algorithm to predict the allergen contamination and to provide real-time feedback. The current study leverages the Texas Instruments DLP® NIRscan™ Nano Evaluation Module(900 – 1700 nm) that is integrated with Raspberry Pi Module 3 to collect the spectral signature of the wheat flour . The samples contained two categories: pure and contaminated wheat flour. First, the spectral signature of pure wheat sample are obtained. The wheat flour is contaminated with Peanut contaminated samples are acquired. Applying a machine learning pipeline that is being split into the following phases - training-test split, feature preprocessing, model selection, hyperparameter tuning, feature selection and model evaluation on the test set. The Light Gradient Boosting Machines (LGBM) and eXtreme Gradient Boosting machines (XGBoost) achieved the best average balanced accuracy. The final trained models were then evaluated on the test set, which was also used to calculate their final SHAP values. The random forest and XGBoost models were found to be the most effective classifiers, outperforming SVCs, decision trees, and the more computationally efficient Light Gradient Boosting Machine (LGBM) algorithm. The model predictions achieve above 90% balanced accuracy that predicts the presence or absence of allergen in question

The present antinutrients and unpleasant flavours constrain consumers’ acceptance of kidney bean flour. Extrusion can remove antinutrients by heat, pressure, and shear. However, its ability to modify beany flavour has not been well-investigated. While researchers have investigated the effects of extrusion temperature and feed moisture, the effects of acid or alkali injection into the extruder are yet to be systematically elucidated. This study investigated the effects of injecting acetic acid or sodium carbonate solutions, at three levels of concentration (0.05, 0.10, 0.15 mol/L), combined with three levels of temperature profile (die temperature set at 90, 110, 130°C) and two levels of feed moisture (25, 30%) during extrusion on the removals of antinutrients (condensed tannins, trypsin inhibitor activity, raffinose family oligosaccharides) and beany flavour (volatile compounds) in whole kidney bean flour. Results showed that all concentration levels of acetic acid and sodium carbonate solution increased the reduction of condensed tannin compared with water, especially at 130°C extrusion temperature. The addition of acetic acid and sodium carbonate at 0.15 mol/L concentration resulted in 72 and 90% reduction of total raffinose oligosaccharide content, as compared with 17% when water alone was added. The addition of sodium carbonate under three investigated concentrations reduced the total volatile compound in the bean flours by 45-58% as compared with water (23-33%) and acetic acid (11-27%), primarily due to aldehydes, alcohols, and aromatic hydrocarbons. These results suggest that the injection of sodium carbonate solution during extrusion could effectively reduce antinutrients and beany flavour compounds in kidney bean flour.

Ensuring food safety of eggs is crucial to preventing the risk of foodborne illnesses for consumers. However, the traditional egg surface decontamination method, washing with hot water containing disinfecting chemicals can inadvertently damage the eggshell cuticle, acting as a natural barrier, preventing bacterial access inside egg, and maintaining egg quality during storage. To address this challenge, it is essential to explore innovative technologies for egg surface decontamination. Non-thermal emerging technologies, such as engineering water nanostructures (EWNS) and cold plasma, offer promising alternatives, in which their effectiveness in deactivating bacteria on egg surface without any adverse impact on egg quality have been proved in our previous studies. However, understanding the impact of these technologies on the eggshell cuticle structure is crucial for informed decision-making regarding their practical application. In this study, the effects of non-thermal emerging technologies on eggshell cuticle chemical composition and coverage were evaluated, utilizing ATR Infrared spectroscopy and synchrotron-based X-ray micro-CT tomography. Our findings indicate that unwashed eggs (control) with intact cuticles exhibit a highly intense polysaccharide band, signifying strong glycosylation of the cuticle. In contrast, washed eggs show a weak polysaccharide band and strong carbonate absorption peaks due to cuticle thickness reduction during the washing process, and significantly lower cuticle coverages. On the other side, no significant differences in the eggshell chemical compositions and coverage of unwashed eggs and those treated with EWNS, and cold plasma were observed, suggesting that these methods do not have any adverse impacts on the eggshell cuticle chemical composition and coverage.

Leafy vegetables, particularly lettuce (Lactuca sativa), are prone to postharvest deterioration due to high water content and rapid loss of green color following harvest. This research aimed to investigate the effects of different storage light spectra on shelf life and visual characteristics of lettuce while minimizing postharvest quality losses. Since green light does not induce stomatal opening and as the consequence minimal impact on wilting of the product, in the present study the impact of two wavelengths of green LEDs with peaks at 500 nm and 530 nm, compared to the white LEDs (400–700 nm), and dark storage (control) over a 14-day storage period were assessed. Lettuce leaves were exposed to constant light intensity (10 µmol m-2 s-1), photoperiod (12 h-d), and air temperature (5°C). Lettuce stored under 500 nm and 530 nm green LEDs exhibited a significantly extended shelf life compared to dark storage. Noteworthy improvements in stiffness, and maintaining color were obvious by postharvest exposure to green LEDs. Exposure to 530 nm green LEDs led to notably higher chlorophyll index compared to other treatments. Photosynthesis was run over an extended duration, resulting in higher levels of total soluble sugar (TSS), while lower transpiration compared to samples under dark storage and white LEDs. Under green light, relative water content (RWC) was higher than its value under dark storage and white LEDs. These findings highlight the positive impact of green LEDs on quality retention and visual attributes of lettuce, offering valuable insights for postharvest practices.

The bin piler is a mechanical arm used to pile tubers within a bulk storage facility. This study evaluates a localization system designed to track the position of the bin-piler, enabling the creation of a 3D quality map of a tuber storage facility. Conducted at Dalhousie University, the research involved two experiments to determine localization accuracy in a 32 m^3 grid using stationary anchors at varied and uniform heights. The varied setup had a total anchor height range of 2.09 m, while the uniform setup held all anchors at an equal height of 0.52 m. Initial tests yielded accuracies of 1.19 m and 75 m, respectively, revealing geometric limitations in the localization algorithm.

Implementing a multivariate Support Vector Machine regressor model improved localization accuracy to 0.23 m and 0.49 m for varied and uniform anchor heights. Further optimization of hyperparameters was conducted, training models using a series of values for both the regularization parameter and epsilon, ranging from 0.1 – 100 and 0.1 – 1, respectively. With optimized regularization and epsilon, the accuracy further improved to 0.18 m for the varied anchor height setup, and 0.32 m for uniform heights. The synchronization of the localization information with other sensors that detect quality information of tuber, such as size, would create storage quality maps for better traceability.

Key Words:

Food traceability

UWB sensors

Wireless sensing

Yield quality