Food Production Line Research Articles

A Candy Defect Detection Method Based on StyleGAN2 and Improved YOLOv7 for Imbalanced Data.

Quality management in the candy industry is a vital part of food quality management. Defective candies significantly affect subsequent packaging and consumption, impacting the efficiency of candy manufacturers and the consumer experience. However, challenges exist in candy defect detection on food production lines due to the small size of the targets and defects, as well as the difficulty of batch sampling defects from automated production lines. A high-precision candy defect detection method based on deep learning is proposed in this paper. Initially, pseudo-defective candy images are generated based on Style Generative Adversarial Network-v2 (StyleGAN2), thereby enhancing the authenticity of these synthetic defect images. Following the separation of the background based on the color characteristics of the defective candies on the conveyor belt, a GAN is utilized for negative sample data enhancement. This effectively reduces the impact of data imbalance between complete and defective candies on the model's detection performance. Secondly, considering the challenges brought by the small size and random shape of candy defects to target detection, the efficient target detection method YOLOv7 is improved. The Spatial Pyramid Pooling Fast Cross Stage Partial Connection (SPPFCSPC) module, the C3C2 module, and the global attention mechanism are introduced to enhance feature extraction precision. The improved model achieves a 3.0% increase in recognition accuracy and a 3.7% increase in recall rate while supporting real-time recognition scenery. This method not only enhances the efficiency of food quality management but also promotes the application of computer vision and deep learning in industrial production.

Hardware Design Strategies Oriented to Edge Computing and AgriFood Electronics for Image Processing using Cellular Automata

There has been a growing interest from academia and industry in developing circuits and systems for edge computing and quality control tasks in food production lines, where image-processing is frequently required. This paper outlines the required considerations for designing a fruit classification system based on image-processing using Cellular Automata (CA) models and integrating it into reconfigurable hardware (HW) such as Field Programmable Gate Arrays (FPGAs). Parallel processing in CA requires numerous processing elements to be implemented and mapping CA models to HW generally comes with limitations. Homogeneous CA arrays are easier to design and implement in HW but can be resource-demanding. To fill this gap, this study explores different alternatives for the HW implementation of CA models, particularly trading computational-parallelism for a more optimized use of the available HW resources. We conducted experimental tests of the designed HW system using the Digilent Nexys development board, and the operation was validated against software-based benchmarks for image-processing, particularly concerning edge-detection. The presented study provides a broader range of design solutions for the HW implementation of two-dimensional CA models and a better understanding of their advantages and disadvantages. The results show that solutions focusing on instruction-parallelism add some complexity to the conception and require more design effort, compared to homogeneous CA models composed of identical cells. However, the instruction-parallel design solutions can significantly improve the HW resource utilization, especially when implementing computationally intensive CA rules in FPGAs.

Monitoring and risk assessment for some quaternary ammonium compounds as antimicrobial agents in some processed and packaged food

ABSTRACT Quaternary ammonium compounds (QACs) might be applied as fungicides, herbicides, and algaecide in agriculture or as antimicrobial agents in food preservation. There are risks either from the application of QACs as food additives or the contamination during the disinfection and sterilisation of the food production lines by QACs in food production factories. Exceeding the maximum residue limits (MRLs) wasn’t accepted by European legislation and Codex guidelines, leading to several risks in human health. Inspecting and monitoring of QACs in 143 packaged food products, such as liquid milk, juices, baby food products, herbal drinks, and frozen vegetables and fruits were carried out in some Egyptian local markets during 2022–2023. The In-house validated QuEChERS extraction method was operated and followed by LC-MS/MS analysis. Only one violated baby food sample that was contaminated with 0.064 mg/kg DDAC and 0.335 mg/kg BACs. According to the EU MRLs, there were no higher residue levels of QACs in juices, milk products, and frozen vegetables with acceptable concentrations. All analysed samples have acceptable results with concentrations less than 0.1 mg/kg, which means that they were all safe for human consumption and health. Our study supported the analytical procedure for the routine inspection of QACs in food.

Physical Contamination Detection in Food Industry Using Microwave and Machine Learning

The detection of contaminants in food products after packaging by a non-invasive technique is a serious need for companies operating in the food industry. In recent years, many technologies have been investigated and developed to overcome the intrinsic drawbacks of the currently employed techniques, such as X-rays and metal detector, and to offer more appropriate solutions with respect to techniques developed in the academic domain in terms of acquisition speed, cost, and the penetration depth (infrared, hyperspectral imaging). A new method based on MW sensing is proposed to increase the degree of production quality. In this paper, we are going to present a novel approach from measurements setup to a binary classification of food products as contaminated or uncontaminated. The work focuses on combining MW sensing technology and ML tools such as MLP and SVM in a complete workflow that can operate in real time in a food production line. A very good performance accuracy that reached 99.8% is achieved using the non-linear SVM algorithm, while the accuracy of the performance of the MLP classifier reached 99.3%.

Descattering for transmissive inspection in production line using slanted linear image sensors.

We propose a descattering method that can be easily applied to food production lines. The system consists of several sets of linear image sensors and linear light sources slanted at different angles. The images captured by these sensors are partially clear along the direction perpendicular to the sensors. We computationally integrate these images on the frequency domain into a single clear image. The effectiveness of the proposed method is assessed by simulation and real-world experiments. The results show that our method recovers clear images. We demonstrate the applicability of the proposed method to a real production line by a prototype system.

A novel low-cost capacitance-based approach for non-destructive evaluation of powder quality for food processing industries

In a high-volume food production line, several stainless-steel hoppers store, mix and dispense powdered ingredients to form food powders of the desired composition. Improper mixing at high speeds and caking (formation of lumps) during storage hamper powder homogeneity and quality. Currently, low throughput offline techniques such as spectroscopic analysis and powder flow tests are used to screen for powder mixing uniformity and caking. Herein, a flexible electrode system is proposed to detect capacitance changes based on the dielectric differences between different powders (beverage powders and their mixtures) and their subsequent variation from poor mixing/caking. The detection scheme successfully differentiated different powder mixtures and detected caking in powders non-invasively with good detection speed (50 ms) and low detection limits (15% for powder mixing and 1–43% for powder caking). Furthermore, the capacitive system integrated with an industrial food hopper successfully showcased an in-situ automated detection of powder caking.

An Effective Optimization Approach to Minimize Waste in a Complex Industrial System

Industrial waste is a major environmental concern. This paper proposes an optimization approach to better manage the operations of a food production line, with the aim of minimizing material waste. The configuration considered is inspired by a real situation. It is a two-stage flow shop with dedicated machines at stage 1 and identical parallel machines at stage 2. A mathematical model is first proposed that considers the minimization of the number of idle times exceeding an acceptable threshold as an objective. A positive time lag for the product and machine setup is considered as a constraint. To solve large problem instances, a hybrid genetic algorithm (HGA) is proposed. In addition, we developed a two-phase procedure based on a decision tree to select the parameters of the HGA. The results show that we could achieve the minimization of material waste without altering too much, another industrial goal, which is the total production time.

Reduce and refine: Plasma treated water vs conventional disinfectants for conveyor-belt cleaning in sustainable food-production lines

Sustainable and microbiologically secure foodstuff production lines are of increasing scientific interest and are in the focus of recent research programs. Additionally, they are of great importance for the production industry due to the prevention of food-borne illnesses caused by pathogens such as Salmonella sp., Listeria monocytogenes, or Escherichia coli. These pathogens are responsible for production losses, loss of customer acceptance, and severe food-borne illnesses. A pathogenic threat is frequently combated with sanitizing steps of the production lines. For conveyor band cleaning, this study compares the cleaning abilities of nitric acid (HNO3) and plasma treated water (PTW), which have been sprayed via a commercially available nozzle on two different polymeric surfaces (polysiloxane and polyurethane). Additionally, the cleaning agents HNO3 and PTW have been characterized through their pH and their conductivity. These findings have been underpinned by experiments that focus on a possible influence of nozzle abrasion, such as brass and stainless-steel nanoparticles, on the antimicrobial potential of PTW and HNO3. Adversely acting effects like an enhanced abrasion of conveyer band materials due to PTW or HNO3 treatment have been checked by using light microscopic micrographs and topographic scans in high-resolution mode. Based on the presented results of the experiments, the suitability of an in-place sanitation step in foodstuff production lines has been demonstrated on a laboratory scale.

COMBING IMAGING TO DIFFERENTIATE CONNECTIVE TISSUE DISEASE-ASSOCIATED INTERSTITIAL LUNG DISEASE FROM IDIOPATHIC PULMONARY FIBROSIS

SESSION TITLE: Medical Student/Resident Pulmonary Manifestations of Systemic Disease Posters SESSION TYPE: Med Student/Res Case Rep Postr PRESENTED ON: October 18-21, 2020 Introduction: Parenchymal lung involvement in Sjogren's syndrome has generally been associated with a nonspecific interstitial pneumonia pattern (NSIP)1,2. We present a patient with chest CT findings compatible with usual interstitial pneumonia (UIP) pattern including exuberant honeycombing, straight edge sign, and the upper lobe sign. After thorough evaluation, the patient was diagnosed with Sjogren's syndrome with exuberant honeycombing, a rare finding in this specific connective tissue disease. Case Presentation: A 36-year-old non-Spanish speaking Guatemalan female presented with progressive chest pain, subacute on chronic dyspnea, and a dry cough. On admission, she had hypoxemic respiratory failure which responded to supplemental oxygen. CT angiogram was negative for pulmonary emboli but showed extensive bibasilar “exuberant” honeycombing. She immigrated to the US 2 years before and had no significant past medical conditions or medication use - prescription or herbal. Exposure history was notable for exposure to indoor smoke from firewood burning stove 10 years ago and she is currently employed on a food production line. Otherwise, infectious and social histories were negative. On examination, she had poor dentition with several capped teeth, cracked lips, and dry Velcro crackles on lung auscultation. Autoimmune serology was significant for positive SSA/SSB and ANA. Her QuantiFERON gold test was likewise positive. Echocardiogram and right heart catheterization showed severe pulmonary hypertension with normal PCWP with a positive vasodilator response. In the context of the UIP pattern on CT chest, examination, and autoimmune serology, she was diagnosed with Sjogren's related interstitial lung disease (ILD). She gradually improved in the hospital with conservative management including supplemental oxygen. A twelve-week course of isoniazid and rifapentine was initiated for LTBI. Sjogren's related ILD therapy is planned after completion of LTBI treatment. She is undergoing evaluation in the Pulmonary Hypertension Clinic and the Transplant Clinic. Discussion: Although mild basilar and reticular honeycombing is a well-described pattern in Sjogren's related ILD, extensive honeycombing is extremely rare1,2. Connective Tissue Disease-related-UIP (CTD-UIP) is differentiated from Idiopathic Pulmonary Fibrosis-associated-UIP (IPF-UIP) in this patient by the presence of the anterior upper lobe sign, straight-edge sign, and exuberant honeycombing sign. (See figure 1) Individually, these signs have low sensitivity (<25%) albeit high specificity (87-94%) for CTD-IPF3. When present together, the sensitivity increases to 95.5% with a likelihood ratio LR(+) of 5.28 for CTD-UIP3. Conclusions: We report an unusual case with extensive “exuberant” honeycombing due to Sjogren's disease. When present, these radiographic signs are valuable aids in differentiating CTD-related-UIP from IPF UIP. Reference #1: Kim, E. J., Collard, H. R. & King, T. E. Rheumatoid arthritis-associated interstitial lung disease: the relevance of histopathologic and radiographic pattern. Chest 136, 1397 -1405 (2009). Reference #2: Mueller-Mang, C., Grosse, C., Schmid, K., Stiebellehner, L. & Bankier, A. A. What Every Radiologist Should Know about Idiopathic Interstitial Pneumonias. RadioGraphics 27, 595 -615 (2007) Reference #3: Chung, J. H. et al. CT Features of the Usual Interstitial Pneumonia Pattern: Differentiating Connective Tissue Disease -Associated Interstitial Lung Disease From Idiopathic Pulmonary Fibrosis. Am. J. Roentgenol. 210, 307 -313 (2017) Disclosures: No relevant relationships by Timothy Dempsey, source=Web Response No relevant relationships by Samuel Falde, source=Web Response No relevant relationships by Kaiser Lim, source=Web Response

Measuring the production performance indicators for food processing industry

Production performance measurement is the most important activity to achieve the higher productivity in food industries. The Overall Equipment Effectiveness (OEE) is a standard for measuring production productivity. However, the uncertainty of OEE components could mask the true production performance estimation in food processing lines. In this study, the production speed and stoppage duration measurements as major uncertainties to estimate the OEE and its components were investigated. In order to handle the uncertainties, two methods based on fuzzy arithmetic and interval arithmetic were conducted to supplement the classical-OEE. To survey the classical-OEE, the operational data field of edible oil purification process over a period of two years were examined. The comparison results of filed data revealed that the new OEE index has been increased around 4% by maintenance improvements. In such index, the total availability, performance efficiency and quality rate were improved by 3%, 2% and 5%, respectively. The research results showed that the spread values of Fuzzy-OEE are more explicit to show the worst and best values, while the interval arithmetic of OEE reflects the results by tighter limits. The proposed methods are more convenient to quantify uncertainty with limits to understand the OEE variations helping to make better decisions on uncertainties in food production lines.

Multi-source domain adaptation for quality control in retail food packaging

Retail food packaging contains information which informs choice and can be vital to consumer health, including product name, ingredients list, nutritional information, allergens, preparation guidelines, pack weight, storage and shelf life information (use-by/best before dates). The presence and accuracy of such information is critical to ensure a detailed understanding of the product and to reduce the potential for health risks. Consequently, erroneous or illegible labeling has the potential to be highly detrimental to consumers and many other stakeholders in the supply chain. In practice, due to the high volume of food packages that go through the supply chain, mistakes do occur therefore good quality of images are needed to verify the correctness of the information. In this paper, a multi-source deep learning-based domain adaptation system is proposed and tested to identify and verify the presence and legibility of use-by date information from food packaging photos taken as part of the validation process as the products pass along the food production line. This was achieved by improving the generalization of the techniques via incorporating new loss functions and making use of multi-source datasets in order to extract domain-invariant representations for all domains and aligning distribution of all pairs of source and target domains in a common feature space, along with the class boundaries. The proposed system performed very well in the conducted experiments, for automating the verification process and reducing labeling errors that could otherwise threaten public health and contravene legal requirements for food packaging information and accuracy. Comprehensive experiments on our food packaging datasets demonstrate that the proposed multi-source deep domain adaptation method significantly improves the classification accuracy and therefore has great potential for application and beneficial impact in food manufacturing control systems.

Контроль качества розлива и маркировки пищевых продуктов с использованием интеллектуальных технологий

Статья посвящена повышению эффективности работы пищевого предприятия за счёт разработки и внедрения автоматизированной системы управления технологическим манипулятором на линии розлива и маркировки пищевых напитков с использованием интеллектуальных технологий. Показана необходимость в решении кинематических задач, особенно задач обратной кинематики манипулятора-робота. Проведенный в работе анализ показал, что существующие универсальные алгоритмы расчёта кинематики просты для записи в общем виде, но из-за ряда неустранимых недостатков ресурсоёмки и обладают в некоторых случаях большой вычислительной погрешностью. Если технологические манипуляторы вынуждены работать именно в таких зонах, в которых этот метод приводит к существенным вычислительным ошибкам, то это может привести к замедлению движения манипулятора, что также приводит к снижению производительности, или потребовать вынужденно изменять режим работы с целью удаления рабочего органа от центра рабочей зоны, что займет больше места для каждого манипулятора. Поэтому для расчета кинематики предложено вместо универсального алгоритма использовать набор простых тригонометрических выражений по условиям конкретных движений, что позволяет сохранить требуемую точность во всём объёме рабочей зоны. Для управления манипулятором, поиска оптимальной траектории и взаимодействия с программой-моделью реальности (в среде SolidWorks) использовались программы, созданные в среде LabVIEW с инструментарием NI-SoftMotion в качестве контроллера для управления положением модели. Для автоматизации контроля качества розлива и маркировки пищевых продуктов и модернизации системы управления технологическим манипулятором разработаны алгоритмы решения обратной кинетической задачи для роботов типа PUMA, SCARA и KUKA, а также программное обеспечение, реализующее разработанные алгоритмы. Показаны результаты анализа использования системы оптического распознавания изделий, учитывающего совместное движение манипулятора и изделий. В результате проведенных исследований разработана система виртуального прототипирования мехатронной системы, позволяющая выбрать контроллер управлением движением манипулятора и требуемые двигатели и параметры оптимизации линии розлива и маркировки пищевых продуктов с использованием интеллектуальных технологии. По полученным результатам разработана улучшенная модель робота PUMA 560, к которой добавлены средства для передвижения манипулятора с использованием интеллектуальных технологий, что позволит повысить эффективность работы линии розлива и маркировки пищевых продуктов, улучшить их показатели качества.

Comparison of the reliability of programmable logic controller and electromagnetic relay control in industrial production line

This paper is aimed at determining the reliability of Industrial Production Line using Programmable Logic Controller. Production lines in the manufacturing sector are automated using Programmable Logic Controller module, which is programmed using programming software. The programming software used is the Siemens S7-200 while the programming language is the Ladder Programming language. Most manufacturing industries in Nigeria today still use relays for control purposes. The wirings from the control panel are so massive that any fault from the machine usually takes hours to trace and replace using the traditional electromagnetic relay. Alternative and better control methods use Programmable Logic Controller. Research articles on the reliability of PLC in a food production line are scarce. This paper is written to address this gap. The study uses the statistical analytical method, which provides useful and higher accuracy of the result as it allows high tolerance to uncertainties. The study determines the failure rate, the mean time between failures (MTBF), mean time to failure (MTTF), and availability as a means of determining the reliability of the network. Accordingly, the study showed that the mean availability of the Production line using PLC for three consecutive years is 84% while the reliability is 86%. Consequently, the use of PLC should be encouraged in industrial automation as the fault can be detected easily, thereby reducing downtime as compared to the use of electromagnetic relay in the control circuitry. Keywords: Programmable logic controller, availability, reliability, ladder programming language, relay logic

An ADOxx Based Environment for Problem Based Learning in Manufacturing Systems Design

The Problem Based Learning (PBL) as student centred approach and learning-by-doing method is suited for the modern higher education. However, the first contact with the method can be overwhelming for the students, in the absence of prior domain knowledge. The preparation of the learning material can be time and resource consuming for the teacher. The goal of the research was the implementation of an environment that should enhance the learning experience for the student and reduce the implementation burden for the teacher. The environment is based on the ADOxx platform and allows the collaboration of the learner teams and the teacher-learner interaction on three levels. The Metamodeling level supports the development of the domain-specific language used in the modelling of the manufacturing system; this activity stimulates and directs the gathering and consolidation of domain-specific knowledge. The modelling level allows the development of alternative design solution using models of the factory components. The Simulation level allows the analysis of these variants. The environment supports the teacher in developing instructional scaffolding and uses cases to ease the learners the first time contact with PBL. The functionality of the environment is presented using the case of designing a flexible food production line.

TDR application for moisture content estimation in agri-food materials

Cereals and legumes are invaluable resources, as they represent the raw materials of many foods and beverages. Cereals are also largely used as livestock feeds, thus indirectly influencing the quality of dairy products and meat. Because of their key role in human nutrition and well-being, the safety and quality of these agri-food materials are extremely important topics in food science. In this regard, moisture content is one of the ultimate factors influencing the quality, safety and price of the final food product; hence, it is crucial to monitor water content of materials in the food production line. Starting from these considerations, this paper describes the use of time domain reflectometry (TDR) for both in-line and off-line moisture content sensing of agrifood materials. In particular, after a brief description of the basic principles of TDR, two representative application cases for moisture content measurements of agri-food materials are reported and commented on.

Business Analytics in Performance Assessment of Food Manufacturing System

Transparency is a jargon in the food industry particularly driven by regulators and consumers. Thoughits a common term, understanding and applying it is still unclear to the food industry specifically what transparencymeans, why to pursue it, what it involves, or how to improve it (http://www.streetdirectory.com, 2017). The realitiesof todays competitive food manufacturing industry, along with pressure from stakeholders, require business leaders tobe more aggressive and creative with cost reduction initiatives (corepractice.com, 2017). This led to the developmentof a system that will enable food production line to automate inventory and product production monitoring. This willenable transparencies in terms of viewing available raw materials vis-´a-vis the production order. This paper will providearchitecture of a food manufacturing system that could lead to transparencies.

A novel library-independent approach based on high-throughput cultivation in Bioscreen and fingerprinting by FTIR spectroscopy for microbial source tracking in food industry.

The source tracking of fungal contamination in the food industry is an important aspect of food safety. Currently, all available methods are time consuming and require the use of a reference library that may limit the accuracy of the identification. In this study, we report for the first time, a library-independent FTIR spectroscopic approach for MST of fungal contamination along the food production line. It combines high-throughput microcultivation and FTIR spectroscopy and is specific on the genus and species level. Therefore, such an approach possesses great importance for food safety control in food industry.

Software Architecture and Framework for Programmable Logic Controllers: A Case Study and Suggestions for Research

Programmable Logic Controllers (PLCs) are widely used for control and automation of machines and processes, so the quality of PLC software has a direct impact on production efficiency. This paper describes a PLC program for a food production line that was completely rewritten using a different software architecture and framework. The PLC hardware and the production line equipment were not changed, so this project provides an opportunity to quantify the impact of different PLC software architecture on production efficiency. The average number of cases of products produced per production hour during the first ten months with the new program was 6.1% higher than the average during the previous ten months with the old program. PLC software, unlike most other software, is often used by the end-user for troubleshooting. In this case, a relatively simple architecture and framework that favors the end-user significantly improved production efficiency over a more sophisticated architecture and framework that favors the software developer. Suggestions for further research on software architecture and framework are given.

Investigation of the effect of UV-LED exposure conditions on the production of vitamin D in pig skin

The dietary intake of vitamin D is currently below the recommended intake of 10–20μg vitamin D/day. Foods with increased content of vitamin D or new products with enhanced vitamin D are warranted. Light-emitting diodes (LEDs) are a potential new resource in food production lines. In the present study the exposure conditions with ultraviolet (UV) LEDs were systematically investigated in the wavelength range 280–340nm for achieving optimal vitamin D bio-fortification in pig skin. A wavelength of 296nm was found to be optimal for vitamin D3 production. The maximum dose of 20kJ/m2 produced 3.5–4μg vitamin D3/cm2 pig skin. Vitamin D3 produced was independent on the combination of time and intensity of the LED source. The increased UV exposure by UV-LEDs may be readily implemented in existing food production facilities, without major modifications to the process or processing equipment, for bio-fortifying food products containing pork skin.

Methicillin Resistant &lt;em&gt;Staphylococcus aureus&lt;/em&gt; (MRSA) and Methicillin Resistant Coagulase Negative &lt;em&gt;Staphylococci&lt;/em&gt; (MRCoNS) Isolated from Fish and Fish Handlers in Maiduguri, Nigeria

Methicillin-Resistant <em>Staphylococcus aureus</em> (MRSA) and Methicillin-Resistant Coagulase Negative <em>Staphylococci</em> (MRCoNS) are of increasing importance to animal and public health where it has been isolated in food production line. The presence of <em>Staphylococci</em> on fish is an indication of contamination or disease in fish. This study was conducted to isolate and phenotypically characterize MRSA and MRCoNS from fresh fish, fish handlers, utensils used for fish processing and the scavenging animals within two major markets supplying fresh fish in Maiduguri, Nigeria. A total isolation rate of 121 (21.1%), 38 (6.6%) and 39 (6.8%) were detected as <em>S. aureus</em>, MRSA and MRCoNS, respectively. MRSA was isolated from fresh fish (2.6%), fish handlers (15%), utensils (4.5%) and scavenging animals (15.9%). MRCoNS was isolated from fresh fish 9 (3.4%), fish handlers 8 (13.3%), utensils 9 (5.8%) and scavenging animals 13 (13.8%). MRSA isolates showed resistance to cefoxitin (100%), gentamicin (89.5%), ciprofloxacin (94.7%), oxacillin (76.3%) and tetracycline (68.4%). Similarly, MRCoNS isolates were resistant to cefoxitin (100%), oxacillin (79.4%), cephazolin (66.6%) and tetracycline (64.1%). MRSA and MRCoNS were detected in most parts of fresh fish and the marketing environments supplying fish to consumers in Maiduguri. Further studies are needed to elucidate transmission routes of MRSA in relation to fresh fish and to provide tools for preventing the spread of MRSA.