Assessment of beer quality based on foamability and chemical composition using computer vision algorithms, near infrared spectroscopy and machine learning algorithms

Enhancing beer authentication, quality, and control assessment using non-invasive spectroscopy through bottle and machine learning modeling

Fraud in alcoholic beverages through counterfeiting and adulteration is rising, significantly impacting companies economically. This study aimed to develop a method using near-infrared (NIR) spectroscopy (1596-2396 nm) through the bottle, along with machine learning (ML) modeling for beer authentication, quality traits, and control assessment. For this study, 25 commercial beers from different brands, styles, and three types of fermentation were used. To obtain the ground-truth data, a quantitative descriptive analysis was conducted with 11 trained panelists to evaluate the intensity of 16 sensory descriptors, and volatile aromatic compounds were analyzed using gas chromatography-mass spectroscopy (GC-MS). The ML models were developed using artificial neural networks with NIR absorbance values as inputs to predict (i) type of fermentation (Model 1), (ii) intensity of 16 sensory descriptors (Model 2), and (iii) peak area of volatile aromatic compounds (Model 3). All models resulted in high overall accuracy (Model 1: 99%; Model 2: R = 0.92; Model 3: R = 0.94), and model deployment for new beer samples showed high performance (Model 1: 95%; Model 2: R = 0.83). This method enables brewers and retailers to analyze beers without opening bottles, preventing quality assurance issues, fraud, and provenance concerns. Further model training with new targets could assess additional quality traits like physicochemical parameters and origin. PRACTICAL APPLICATION: Near-infrared spectroscopy coupled with ML modeling is a novel method for assessing beer quality and authentication through the bottle. It serves as a rapid, accurate tool for predicting sensory and aroma profiles without opening the bottle. Additionally, it monitors quality traits during transport and storage.

Flavor Wheel Development from a Machine Learning Perspective

The intricate relationships between chemical compounds and sensory descriptors in distilled spirits have long intrigued distillers, sensory experts, and consumers alike. The importance and complexity of this relation affect the production, quality, and appreciation of spirits, and the success of a product. Because of that, profoundly investigating the different flavor and aroma combinations that the chemical compounds can give to a desired beverage takes an essential place in the industry. This study aims to study these relationships by employing machine learning techniques to analyze a comprehensive dataset with 3051 chemical compounds and their associated aroma descriptors for seven distilled spirit categories: Baijiu, cachaça, gin, mezcal, rum, tequila, and whisk(e)y. The study uses principal component analysis (PCA) to reduce the dimensionality of the dataset and a clustering machine learning model to identify distinct clusters of aroma descriptors associated with each beverage category. Based on these results, an aroma wheel that encapsulates the diverse olfactory landscapes of various distilled spirits was developed. This flavor wheel is a valuable tool for distillers, sensory experts, and consumers, providing a comprehensive reference for understanding and appreciating the complexities of distilled spirits.

Analytical Chemistry Strategies in the Use of Miniaturised NIR Instruments: An Overview

Miniaturized NIR instruments have been increasingly used in the last years, and they have become useful tools for many applications on a broad variety of samples. This review focuses on miniaturized NIR instruments from an analytical point of view, to give an overview of the analytical strategies used in order to help the reader to set up their own analytical methods, from the sampling to the data analysis. It highlights the uses of these instruments, providing a critical discussion including current and future trends.

Graphormer-IR: Graph Transformers Predict Experimental IR Spectra Using Highly Specialized Attention

Infrared (IR) spectroscopy is an important analytical tool in various chemical and forensic domains and a great deal of effort has gone into developing in silico methods for predicting experimental spectra. A key challenge in this regard is generating highly accurate spectra quickly to enable real-time feedback between computation and experiment. Here, we employ Graphormer, a graph neural network (GNN) transformer, to predict IR spectra using only simplified molecular-input line-entry system (SMILES) strings. Our data set includes 53,528 high-quality spectra, measured in five different experimental media (i.e., phases), for molecules containing the elements H, C, N, O, F, Si, S, P, Cl, Br, and I. When using only atomic numbers for node encodings, Graphormer-IR achieved a mean test spectral information similarity (SISμ) value of 0.8449 ± 0.0012 (n = 5), which surpasses that the current state-of-the-art model Chemprop-IR (SISμ = 0.8409 ± 0.0014, n = 5) with only 36% of the encoded information. Augmenting node embeddings with additional node-level descriptors in learned embeddings generated through a multilayer perceptron improves scores to SISμ = 0.8523 ± 0.0006, a total improvement of 19.7σ (t = 19). These improved scores show how Graphormer-IR excels in capturing long-range interactions like hydrogen bonding, anharmonic peak positions in experimental spectra, and stretching frequencies of uncommon functional groups. Scaling our architecture to 210 attention heads demonstrates specialist-like behavior for distinct IR frequencies that improves model performance. Our model utilizes novel architectures, including a global node for phase encoding, learned node feature embeddings, and a one-dimensional (1D) smoothing convolutional neural network (CNN). Graphormer-IR's innovations underscore its value over traditional message-passing neural networks (MPNNs) due to its expressive embeddings and ability to capture long-range intramolecular relationships.

Evaluation of spontaneous fermentation impact on the physicochemical properties and sensory profile of green and roasted arabica coffee by digital technologies

There is a growing demand for specialty coffee with more pleasant and uniform sensory perception. Wet fermentation could modulate and confer additional aroma notes to final roasted coffee brew. This study aimed to assess differences in volatile compounds and the intensities of sensory descriptors between unfermented and spontaneously fermented coffee using digital technologies. Fermented (F) and unfermented (UF) coffee samples, harvested from two Australia local farms Mountain Top Estate (T) and Kahawa Estate (K), with four roasting levels (green, light-, medium-, and dark-) were analysed using near-infrared spectrometry (NIR), and a low-cost electronic nose (e-nose) along with some ground truth measurements such as headspace/gas chromatography-mass spectrometry (HS-SPME-GC-MS), and quantitative descriptive analysis (QDA ®). Regression machine learning (ML) modelling based on artificial neural networks (ANN) was conducted to predict volatile aromatic compounds and intensity of sensory descriptors using NIR and e-nose data as inputs. Green fermented coffee had significant perception of hay aroma and flavor. Roasted fermented coffee had higher intensities of coffee liquid color, crema height and color, aftertaste, aroma and flavor of dark chocolate and roasted, and butter flavor (p < 0.05). According to GC-MS detection, volatile aromatic compounds, including methylpyrazine, 2-ethyl-5-methylpyrazine, and 2-ethyl-6-methylpyrazine, were observed to discriminate fermented and unfermented roasted coffee. The four ML models developed using the NIR absorbance values and e-nose measurements as inputs were highly accurate in predicting (i) the peak area of volatile aromatic compounds (Model 1, R = 0.98; Model 3, R = 0.87) and (ii) intensities of sensory descriptors (Model 2 and Model 4; R = 0.91), respectively. The proposed efficient, reliable, and affordable method may potentially be used in the coffee industry and smallholders in the differentiation and development of specialty coffee, as well as in process monitoring and sensory quality assurance.

The Impact of Wet Fermentation on Coffee Quality Traits and Volatile Compounds Using Digital Technologies

Fermentation is critical for developing coffee’s physicochemical properties. This study aimed to assess the differences in quality traits between fermented and unfermented coffee with four grinding sizes of coffee powder using multiple digital technologies. A total of N = 2 coffee treatments—(i) dry processing and (ii) wet fermentation—with grinding levels (250, 350, 550, and 750 µm) were analysed using near-infrared spectrometry (NIR), electronic nose (e-nose), and headspace/gas chromatography–mass spectrometry (HS-SPME-GC-MS) coupled with machine learning (ML) modelling. Most overtones detected by NIR were within the ranges of 1700–2000 nm and 2200–2396 nm, while the enhanced peak responses of fermented coffee were lower. The overall voltage of nine e-nose sensors obtained from fermented coffee (250 µm) was significantly higher. There were two ML classification models to classify processing and brewing methods using NIR (Model 1) and e-nose (Model 2) values as inputs that were highly accurate (93.9% and 91.2%, respectively). Highly precise ML regression Model 3 and Model 4 based on the same inputs for NIR (R = 0.96) and e-nose (R = 0.99) were developed, respectively, to assess 14 volatile aromatic compounds obtained by GC-MS. Fermented coffee showed higher 2-methylpyrazine (2.20 ng/mL) and furfuryl acetate (2.36 ng/mL) content, which induces a stronger fruity aroma. This proposed rapid, reliable, and low-cost method was shown to be effective in distinguishing coffee postharvest processing methods and evaluating their volatile compounds, which has the potential to be applied for coffee differentiation and quality assurance and control.

The Study of Mineral Elements Participation in Brewing Products Foam Structure Formation

The article is devoted to the study of the metal ion’s role in the formation of the foam colloidal structure. The nitrogenous compounds prominent role as structuring compounds of the colloidal film on the carbon dioxide bubbles surface is noted. Based on the calculation of the correlation strength between beer samples organic compounds obtained based on light and dark malts, as well as barley and corn as unmalted raw materials, the effect of Ca, Mg, Mn, Co and Na mineral ions on the foam structure was evaluated. It was shown that Ca, Mg, Mn, Na, Co ions take part in the foam structure formation in conjunction with the grain raw materials thiol nitrogen, and Ca, Mg, Mn ions in conjunction with catechins of both grain and hop raw materials. High correlation coefficients between foam resistance and all ions (r-0.991÷0.998), foam resistance, catechins and Co ions (r 0.991), as well as foam resistance and Ca and Mg ions (r -0.987) ensure the elasticity of the foam structure colloidal film.

Developing colloidal structure of beer by grain organic compounds

The present article introduces the problem of determining the general structure of beer as a complex system of related biomolecules. The objective was to establish the correlation of various quantities of organic compounds in beer formulation. The research featured samples of filtered pasteurized beer obtained from a retail chain shop in Moscow (Russia). The experiment relied on standard research methods, including instrumental methods of analysis, e.g., high-performance liquid chromatography (HPLC). The obtained experimental data underwent a statistical analysis using the Statistica software (StatSoft, 2016). The research established the correlation between the type of grain (barley or wheat malt) and the content of organic compounds, e.g., β-glucan, polyphenols, soluble nitrogen, etc. The research also revealed some patterns in the distribution of proteins, which served as a framework for the system of organic compounds. The distribution of thiol proteins proved to depend on the dissolution degree of the grain and was different in barley light, barley dark, and wheat malt samples. The fraction distribution of β-glucan depended on the color of the malt. In light beer samples, it concentrated in high- and medium-molecular fractions of nitrogenous substances, in dark beer – in low-molecular fractions (≤ 63%). Initial wort density and alcohol content affected the amount of catechins and total polyphenols. Nitrogenous compounds depended on the color, initial extract, and alcohol content. The nitrogenous structure and other organic compounds of beer proved to depend on protein substances. The research also revealed a number of factors that affected the fraction distribution of biomolecules in different beer sorts.

Evaluation of Improvements in the Separation of Monolayer and Multilayer Films via Measurements in Transflection and Application of Machine Learning Approaches

Small plastic packaging films make up a quarter of all packaging waste generated annually in Austria. As many plastic packaging films are multilayered to give barrier properties and strength, this fraction is considered hardly recyclable and recovered thermally. Besides, they can not be separated from recyclable monolayer films using near-infrared spectroscopy in material recovery facilities. In this paper, an experimental sensor-based sorting setup is used to demonstrate the effect of adapting a near-infrared sorting rig to enable measurement in transflection. This adaptation effectively circumvents problems caused by low material thickness and improves the sorting success when separating monolayer and multilayer film materials. Additionally, machine learning approaches are discussed to separate monolayer and multilayer materials without requiring the near-infrared sorter to explicitly learn the material fingerprint of each possible combination of layered materials. Last, a fast Fourier transform is shown to reduce destructive interference overlaying the spectral information. Through this, it is possible to automatically find the Fourier component at which to place the filter to regain the most spectral information possible.

Rapid Identification of Wild Gentiana Genus in Different Geographical Locations Based on FT-IR and an Improved Neural Network Structure Double-Net

Gentiana Genus, a herb mainly distributed in Asia and Europe, has been used to treat the damp heat disease of the liver for over 2000 years in China. Previous studies have shown significant differences in the compositional contents of wild Gentiana Genus samples from different geographical origins. Therefore, the traceable geographic locations of the wild Gentiana Genus samples are essential to ensure practical medicinal value. Over the last few years, the developments in chemometrics have facilitated the analysis of the composition of medicinal herbs via spectroscopy. Notably, FT-IR spectroscopy is widely used because of its benefit of allowing rapid, nondestructive measurements. In this paper, we collected wild Gentiana Genus samples from seven different provinces (222 samples in total). Twenty-one different FT-IR spectral pre-processing methods that were used in our experiments. Meanwhile, we also designed a neural network, Double-Net, to predict the geographical locations of wild Gentiana Genus plants via FT-IR spectroscopy. The experiments showed that the accuracy of the neural network structure Double-Net we designed can reach 100%, and the F1_score can reach 1.0.

Digital Detection of Olive Oil Rancidity Levels and Aroma Profiles Using Near-Infrared Spectroscopy, a Low-Cost Electronic Nose and Machine Learning Modelling

The success of the olive oil industry depends on provenance and quality-trait consistency affecting the consumers' acceptability/preference and purchase intention. Companies rely on laboratories to analyze samples to assess consistency within the production chain, which may be time-consuming, cost-restrictive, and untimely obtaining results, making the process more reactive than predictive. This study proposed implementing digital technologies using near-infrared spectroscopy (NIR) and a novel low-cost e-nose to assess the level of rancidity and aromas in commercial extra-virgin olive oil. Four different olive oils were spiked with three rancidity levels (N = 17). These samples were evaluated using gas-chromatography-mass-spectroscopy, NIR, and an e-nose. Four machine learning models were developed to classify olive oil types and rancidity (Model 1: NIR inputs; Model 2: e-nose inputs) and predict the peak area of 16 aromas (Model 3: NIR; Model 4: e-nose inputs). The results showed high accuracies (Models 1–2: 97% and 87%; Models 3–4: R = 0.96 and 0.93). These digital technologies may change companies from a reactive to a more predictive production of food/beverages to secure product quality and acceptability.

Detecting the content of the bright blue pigment in cream based on deep learning and near-infrared spectroscopy

The excessive content of additives in food is a radical problem that affects human health. However, traditional chemical methods are limited by a long cycle, low accuracy, and strong destructiveness, so a fast and accurate alternative is urgently needed. This paper proposes a prediction model introducing near-infrared spectroscopy and deep learning to perform fast and accurate non-destructive detection of artificial bright blue pigment in cream. The model results show that R² is 0.9638, RMSEP is 0.0157, and RPD is 4.4022. In the preprocessing part, this paper compares the traditional preprocessing methods (SNV, MSC, SG) horizontally and innovatively proposes the use of autoencoders to mitigate the dimensionality of data, which has immensely improved the follow-up prediction effect. In addition, it tries to perform regression prediction on spectral data and establish a fully connected convolutional neural network model through deep learning, whose result indicators prove better than those of traditional methods such as PLSR and MLR. When constructing the deep learning model, this paper applies knowledge evolution to compress the model to achieve a lower calculation cost and higher accuracy. Compared with the traditional methods, the model proposed in this paper has greater accuracy and higher speed with samples undamaged, which is worth popularizing.

Digital Assessment and Classification of Wine Faults Using a Low-Cost Electronic Nose, Near-Infrared Spectroscopy and Machine Learning Modelling

The winemaking industry can benefit greatly by implementing digital technologies to avoid guesswork and the development of off-flavors and aromas in the final wines. This research presents results on the implementation of near-infrared spectroscopy (NIR) and a low-cost electronic nose (e-nose) coupled with machine learning to detect and assess wine faults. For this purpose, red and white base wines were used, and treatments consisted of spiked samples with 12 faults that are traditionally formed in wines. Results showed high accuracy in the classification models using NIR and e-nose for red wines (94–96%; 92–97%, respectively) and white wines (96–97%; 90–97%, respectively). Implementing new and emerging digital technologies could be a turning point for the winemaking industry to become more predictive in terms of decision-making and maintaining and increasing wine quality traits in a changing and challenging climate.

Measurement Strategies for the Classification of Edible Oils Using Low-Cost Miniaturised Portable NIR Instruments

Miniaturised near-infrared (NIR) instruments have been increasingly used in the last few years, and they have become useful tools for many applications on different types of samples. The market already offers a wide variety of these instruments, each one having specific requirements for the correct acquisition of the instrumental signal. This paper presents the development and optimisation of different measuring strategies for two miniaturised NIR instruments in order to find the best measuring conditions for the rapid and low-cost analysis of olive oils. The developed strategies have been applied to the classification of different samples of olive oils, obtaining good results in all cases.

Application of Artificial Intelligence in Food Industry—a Guideline

Artificial intelligence (AI) has embodied the recent technology in the food industry over the past few decades due to the rising of food demands in line with the increasing of the world population. The capability of the said intelligent systems in various tasks such as food quality determination, control tools, classification of food, and prediction purposes has intensified their demand in the food industry. Therefore, this paper reviews those diverse applications in comparing their advantages, limitations, and formulations as a guideline for selecting the most appropriate methods in enhancing future AI- and food industry–related developments. Furthermore, the integration of this system with other devices such as electronic nose, electronic tongue, computer vision system, and near infrared spectroscopy (NIR) is also emphasized, all of which will benefit both the industry players and consumers.

Smart Detection of Faults in Beers Using Near-Infrared Spectroscopy, a Low-Cost Electronic Nose and Artificial Intelligence

Early detection of beer faults is an important assessment in the brewing process to secure a high-quality product and consumer acceptability. This study proposed an integrated AI system for smart detection of beer faults based on the comparison of near-infrared spectroscopy (NIR) and a newly developed electronic nose (e-nose) using machine learning modelling. For these purposes, a commercial larger beer was used as a base prototype, which was spiked with 18 common beer faults plus the control aroma. The 19 aroma profiles were used as targets for classification ma-chine learning (ML) modelling. Six different ML models were developed; Model 1 (M1) and M2 were developed using the NIR absorbance values (100 inputs from 1596–2396 nm) and e-nose (nine sensor readings) as inputs, respectively, to classify the samples into control, low and high concentration of faults. Model 3 (M3) and M4 were based on NIR and M5 and M6 based on the e-nose readings as inputs with 19 aroma profiles as targets for all models. A customized code tested 17 artificial neural network (ANN) algorithms automatically testing performance and neu-ron trimming. Results showed that the Bayesian regularization algorithm was the most adequate for classification rendering precisions of M1 = 95.6%, M2 = 95.3%, M3 = 98.9%, M4 = 98.3%, M5 = 96.8%, and M6 = 96.2% without statistical signs of under- or overfitting. The proposed system can be added to robotic pourers and the brewing process at low cost, which can benefit craft and larger brewing companies.

Interactive Advertising on HbbTV: An Experimental Analysis of Emotions

Interactivity in television (and sustainability, thanks to virtualization) is a growing phenomenon, driven especially by the implementation of the HbbTV (hybrid broadcast broadband television) standard. Networks, media agencies and advertisers are trying to adapt to and profit from the options generated by the interaction with the viewer. Through an experimental methodology including the viewing of an advertising block which included conventional and interactive advertisements, the emotions of the viewer were collected. It was concluded that the order of the announcements is not decisive, that the emotions of anger and sadness predominate over those of joy due to a negative predisposition towards viewing advertising proposals and the content of the advertisement itself, and that less intrusive formats are more accepted.

Integrating a Low-Cost Electronic Nose and Machine Learning Modelling to Assess Coffee Aroma Profile and Intensity

Aroma is one of the main attributes that consumers consider when appreciating and selecting a coffee; hence it is considered an important quality trait. However, the most common methods to assess aroma are based on expensive equipment or human senses through sensory evaluation, which is time-consuming and requires highly trained assessors to avoid subjectivity. Therefore, this study aimed to estimate the coffee intensity and aromas using a low-cost and portable electronic nose (e-nose) and machine learning modeling. For this purpose, triplicates of nine commercial coffee samples with different intensity levels were used for this study. Two machine learning models were developed based on artificial neural networks using the data from the e-nose as inputs to (i) classify the samples into low, medium, and high-intensity (Model 1) and (ii) to predict the relative abundance of 45 different aromas (Model 2). Results showed that it is possible to estimate the intensity of coffees with high accuracy (98%; Model 1), as well as to predict the specific aromas obtaining a high correlation coefficient (R = 0.99), and no under- or over-fitting of the models were detected. The proposed contactless, nondestructive, rapid, reliable, and low-cost method showed to be effective in evaluating volatile compounds in coffee, which is a potential technique to be applied within all stages of the production process to detect any undesirable characteristics on–time and ensure high-quality products.

Self-Reported Emotions and Facial Expressions on Consumer Acceptability: A Study Using Energy Drinks

Emotional responses elicited by foods are of great interest for new product developers and marketing professionals, as consumer acceptance proved to be linked to the emotions generated by the product in the consumers. An emotional measurement is generally considered an appropriate tool to differentiate between the products of similar nutritional value, flavour, liking and packaging. Novel methods used to measure emotions include self-reporting verbal and visual measurements, and facial expression techniques. This study aimed to evaluate the explicit and implicit emotional response elicited during the tasting of two different brands (A and B) of energy drinks. The explicit response of consumers was assessed using liking (nine-point hedonic scale), and emotions (EsSense Profile^®—Check-All-That-Apply questionnaire), and implicit emotional responses were evaluated by studying facial expressions using the Affectiva Affdex^® software. The familiarity of the product and purchase intent were also assessed during the study. The hedonic rating shows a significant difference in liking between the two brands of energy drink during the tasting session. For the explicit emotional responses, participants elicited more positive emotions than the negative emotions for both energy drinks. However, participants expressed “happy”, “active” and “eager” emotions more frequently for energy drink A. On the other hand, the implicit emotional responses through facial expressions indicated a high level of involvement of the participants with energy drink B as compared to energy drink A. The study showed that overall liking and the explicit and implicit emotional measurements are weakly to moderately correlated.

Advances in infrared spectroscopy combined with artificial neural network for the authentication and traceability of food

The authentication and traceability of food attract more attention due to the increasing consumer awareness regarding nutrition and health, being a new hotspot of food science. Infrared spectroscopy (IRS) combined with shallow neural network has been widely proven to be an effective food analysis technology. As an advanced deep learning technology, deep neural network has also been explored to analyze and solve food-related IRS problems in recent years. The present review begins with brief introductions to IRS and artificial neural network (ANN), including shallow neural network and deep neural network. More notably, it emphasizes the comprehensive overview of the advances of the technology combined IRS with ANN for the authentication and traceability of food, based on relevant literature from 2014 to early 2020. In detail, the types of IRS and ANN, modeling processes, experimental results, and model comparisons in related studies are described to set forth the usage and performance of the combined technology for food analysis. The combined technology shows excellent ability to authenticate food quality and safety, involving chemical components, freshness, microorganisms, damages, toxic substances, and adulteration. As well, it shows excellent performance in the traceability of food variety and origin. The advantages, current limitations, and future trends of the combined technology are further discussed to provide a thoughtful viewpoint on the challenges and expectations of online applications for the authentication and traceability of food.

Low-Cost Methods to Assess Beer Quality Using Artificial Intelligence Involving Robotics, an Electronic Nose, and Machine Learning

Beer quality is a difficult concept to describe and assess by physicochemical and sensory analysis due to the complexity of beer appreciation and acceptability by consumers, which can be dynamic and related to changes in climate affecting raw materials, consumer preference, and rising quality requirements. Artificial intelligence (AI) may offer unique capabilities based on the integration of sensor technology, robotics, and data analysis using machine learning (ML) to identify specific quality traits and process modifications to produce quality beers. This research presented the integration and implementation of AI technology based on low-cost sensor networks in the form of an electronic nose (e-nose), robotics, and ML. Results of ML showed high accuracy (97%) in the identification of fermentation type (Model 1) based on e-nose data; prediction of consumer acceptability from near-infrared (Model 2; R = 0.90) and e-nose data (Model 3; R = 0.95), and physicochemical and colorimetry of beers from e-nose data. The use of the RoboBEER coupled with the e-nose and AI could be used by brewers to assess the fermentation process, quality of beers, detection of faults, traceability, and authentication purposes in an affordable, user-friendly, and accurate manner.

A Digital Approach to Model Quality and Sensory Traits of Beers Fermented under Sonication Based on Chemical Fingerprinting

The development of digital tools based on artificial intelligence can produce affordable and accurate methodologies to assess quality traits and sensory analysis of beers. These new and emerging technologies can also assess new products in a near real-time fashion through virtual simulations before the brewing process. This research was based on the development of specific digital tools (four models) to assess quality traits and sensory profiles of beers produced using sonication and traditional brewing techniques. Results showed that models developed using supervised machine learning (ML) regression algorithms based on near-infrared spectroscopy (NIR) were highly accurate in the estimation of physicochemical parameters (Model 1; R = 0.94; b = 0.91). Outputs from Model 1 were then used as inputs to obtain estimations of the intensity of sensory descriptors (Model 2; R = 0.99; b = 0.98), liking of sensory attributes (Model 3; R = 0.97; b = 0.99), and the classification of fermentation treatments using supervised classification ML algorithms (Model 4; 96% accuracy). These new digital tools can aid craft brewing companies for product development at lower costs and maintain specific quality traits and sensory profiles, creating original styles of beers to get positioned in the market.

Machine Learning Modeling of Wine Sensory Profiles and Color of Vertical Vintages of Pinot Noir Based on Chemical Fingerprinting, Weather and Management Data

Important wine quality traits such as sensory profile and color are the product of complex interactions between the soil, grapevine, the environment, management, and winemaking practices. Artificial intelligence (AI) and specifically machine learning (ML) could offer powerful tools to assess these complex interactions and their patterns through seasons to predict quality traits to winegrowers close to harvest and before winemaking. This study considered nine vintages (2008–2016) using near-infrared spectroscopy (NIR) of wines and corresponding weather and management information as inputs for artificial neural network (ANN) modeling of sensory profiles (Models 1 and 2 respectively). Furthermore, weather and management data were used as inputs to predict the color of wines (Model 3). Results showed high accuracy in the prediction of sensory profiles of vertical wine vintages using NIR (Model 1; R = 0.92; slope = 0.85), while better models were obtained using weather/management data for the prediction of sensory profiles (Model 2; R = 0.98; slope = 0.93) and wine color (Model 3; R = 0.99; slope = 0.98). For all models, there was no indication of overfitting as per ANN specific tests. These models may be used as powerful tools to winegrowers and winemakers close to harvest and before the winemaking process to maintain a determined wine style with high quality and acceptability by consumers.

Beer and Consumer Response Using Biometrics: Associations Assessment of Beer Compounds and Elicited Emotions

Some chemical compounds, especially alcohol, sugars, and alkaloids such as hordenine, have been reported as elicitors of different emotional responses. This preliminary study was based on six commercial beers selected according to their fermentation type, with two beers of each type (spontaneous, bottom, and top). Chemometry and sensory analysis were performed for all samples to determine relationships and patterns between chemical composition and emotional responses from consumers. The results showed that sweeter samples were associated with higher perceived liking by consumers and positive emotions, which corresponded to spontaneous fermentation beers. There was high correlation (R = 0.91; R² = 0.83) between hordenine and alcohol content. Beers presenting higher concentrations of both, and higher bitterness, were related to negative emotions. Further studies should be conducted, giving more time for emotional response analysis between beer samples, and comparing alcoholic and non-alcoholic beers with similar styles, to separate the effects of alcohol and hordenine. This preliminary study was a first attempt to associate beer compounds with the emotional responses of consumers using non-invasive biometrics.

A Digital Approach to Evaluate the Effect of Berry Cell Death on Pinot Noir Wines’ Quality Traits and Sensory Profiles Using Non-Destructive Near-Infrared Spectroscopy

Berry cell death (BCD) is linked to the development of flavors and aromas in berries and wine. The BCD pattern and rate within a growing season start at around 90–100 days after anthesis (DAA), and the rate until harvest depends on environmental factors. This study assessed the BCD effects on berry and wine composition from a boutique commercial vineyard in Victoria, Australia, using fluorescent imaging. Results showed differences in wine sensory profiles from the two blocks studied, mainly related to variations in BCD, due to differences in altitude between blocks. Furthermore, two machine learning (ML) models were constructed using near-infrared spectroscopy (NIR) measurements from full berries as inputs and living tissue (LT) and dead tissue (DT) from berries as targets (Model 1). Model 2 was developed using Brix, LT, DT from the east and west sides of canopies as inputs and using 19 sensory descriptors from wines as targets. High correlation and performances were achieved for both models without signs of overfitting (R = 0.94 and R = 0.80, respectively). These models could be used for decision-making purposes as an objective and comprehensive berry maturity assessment obtained in a non-destructive, accurate, and in a real-time fashion close to harvest, to secure specific wine styles.

Beer Aroma and Quality Traits Assessment Using Artificial Intelligence

Increasing beer quality demands from consumers have put pressure on brewers to target specific steps within the beer-making process to modify beer styles and quality traits. However, this demands more robust methodologies to assess the final aroma profiles and physicochemical characteristics of beers. This research shows the construction of artificial intelligence (AI) models based on aroma profiles, chemometrics, and chemical fingerprinting using near-infrared spectroscopy (NIR) obtained from 20 commercial beers used as targets. Results showed that machine learning models obtained using NIR from beers as inputs were accurate and robust in the prediction of six important aromas for beer (Model 1; R = 0.91; b = 0.87) and chemometrics (Model 2; R = 0.93; b = 0.90). Additionally, two more accurate models were obtained from robotics (RoboBEER) to obtain the same aroma profiles (Model 3; R = 0.99; b = 1.00) and chemometrics (Model 4; R = 0.98; b = 1.00). Low-cost robotics and sensors coupled with computer vision and machine learning modeling could help brewers in the decision-making process to target specific consumer preferences and to secure higher consumer demands.

Development of a Rapid Method to Assess Beer Foamability Based on Relative Protein Content Using RoboBEER and Machine Learning Modeling

Foam-related parameters are associated with beer quality and dependent, among others, on the protein content. This study aimed to develop a machine learning (ML) model to predict the pattern and presence of 54 proteins. Triplicates of 24 beer samples were analyzed through proteomics. Furthermore, samples were analyzed using the RoboBEER to evaluate 15 physical parameters (color, foam, and bubbles), and a portable near-infrared (NIR) device. Proteins were grouped according to their molecular weight (MW), and a matrix was developed to assess only the significant correlations (p < 0.05) with the physical parameters. Two ML models were developed using the NIR (Model 1), and RoboBEER (Model 2) data as inputs to predict the relative quantification of 54 proteins. Proteins in the 0–20 kDa group were negatively correlated with the maximum volume of foam (MaxVol; r = −0.57) and total lifetime of foam (TLTF; r = −0.58), while those within 20–40 kDa had a positive correlation with MaxVol (r = 0.47) and TLTF (r = 0.47). Model 1 was not as accurate (testing r = 0.68; overall r = 0.89) as Model 2 (testing r = 0.90; overall r = 0.93), which may serve as a reliable and affordable method to incorporate the relative quantification of important proteins to explain beer quality.

Modeling Pinot Noir Aroma Profiles Based on Weather and Water Management Information Using Machine Learning Algorithms: A Vertical Vintage Analysis Using Artificial Intelligence

: Wine aroma profiles are determinant for the specific style and quality characteristics of final wines. These are dependent on the seasonality, mainly weather conditions, such as solar exposure and temperatures and water management strategies from veraison to harvest. This paper presents machine learning modeling strategies using weather and water management information from a Pinot noir vineyard from 2008 to 2016 vintages as inputs and aroma profiles from wines from the same vintages assessed using gas chromatography and chemometric analyses of wines as targets. The results showed that artificial neural network (ANN) models rendered the high accuracy in the prediction of aroma profiles (Model 1; R = 0.99) and chemometric wine parameters (Model 2; R = 0.94) with no indication of overfitting. These models could offer powerful tools to winemakers to assess the aroma profiles of wines before winemaking, which could help adjust some techniques to maintain/increase the quality of wines or wine styles that are characteristic of specific vineyards or regions. These models can be modified for different cultivars and regions by including more data from vertical vintages to implement artificial intelligence in winemaking.

Bubbles, Foam Formation, Stability and Consumer Perception of Carbonated Drinks: A Review of Current, New and Emerging Technologies for Rapid Assessment and Control

Quality control, mainly focused on the assessment of bubble and foam-related parameters, is critical in carbonated beverages, due to their relationship with the chemical components as well as their influence on sensory characteristics such as aroma release, mouthfeel, and perception of tastes and aromas. Consumer assessment and acceptability of carbonated beverages are mainly based on carbonation, foam, and bubbles, as a flat carbonated beverage is usually perceived as low quality. This review focuses on three beverages: beer, sparkling water, and sparkling wine. It explains the characteristics of foam and bubble formation, and the traditional methods, as well as emerging technologies based on robotics and computer vision, to assess bubble and foam-related parameters. Furthermore, it explores the most common methods and the use of advanced techniques using an artificial intelligence approach to assess sensory descriptors both for descriptive analysis and consumers' acceptability. Emerging technologies, based on the combination of robotics, computer vision, and machine learning as an approach to artificial intelligence, have been developed and applied for the assessment of beer and, to a lesser extent, sparkling wine. This, has the objective of assessing the final products quality using more reliable, accurate, affordable, and less time-consuming methods. However, despite carbonated water being an important product, due to its increasing consumption, more research needs to focus on exploring more efficient, repeatable, and accurate methods to assess carbonation and bubble size, distribution and dynamics.

Emerging Technologies Based on Artificial Intelligence to Assess the Quality and Consumer Preference of Beverages

Beverages is a broad and important category within the food industry, which is comprised of a wide range of sub-categories and types of drinks with different levels of complexity for their manufacturing and quality assessment. Traditional methods to evaluate the quality traits of beverages consist of tedious, time-consuming, and costly techniques, which do not allow researchers to procure results in real-time. Therefore, there is a need to test and implement emerging technologies in order to automate and facilitate those analyses within this industry. This paper aimed to present the most recent publications and trends regarding the use of low-cost, reliable, and accurate, remote or non-contact techniques using robotics, machine learning, computer vision, biometrics and the application of artificial intelligence, as well as to identify the research gaps within the beverage industry. It was found that there is a wide opportunity in the development and use of robotics and biometrics for all types of beverages, but especially for hot and non-alcoholic drinks. Furthermore, there is a lack of knowledge and clarity within the industry, and research about the concepts of artificial intelligence and machine learning, as well as that concerning the correct design and interpretation of modeling related to the lack of inclusion of relevant data, additional to presenting over- or under-fitted models.

Chocolate Quality Assessment Based on Chemical Fingerprinting Using Near Infra-red and Machine Learning Modeling

Chocolates are the most common confectionery and most popular dessert and snack across the globe. The quality of chocolate plays a major role in sensory evaluation. In this study, a rapid and non-destructive method was developed to predict the quality of chocolate based on physicochemical data, and sensory properties, using the five basic tastes. Data for physicochemical analysis (pH, Brix, viscosity, and color), and sensory properties (basic taste intensities) of chocolate were recorded. These data and results obtained from near-infrared spectroscopy were used to develop two machine learning models to predict the physicochemical parameters (Model 1) and sensory descriptors (Model 2) of chocolate. The results show that the models developed had high accuracy, with R = 0.99 for Model 1 and R = 0.93 for Model 2. The thus-developed models can be used as an alternative to consumer panels to determine the sensory properties of chocolate more accurately with lower cost using the chemical parameters.

Development of Artificial Neural Network Models to Assess Beer Acceptability Based on Sensory Properties Using a Robotic Pourer: A Comparative Model Approach to Achieve an Artificial Intelligence System

Artificial neural networks (ANN) have become popular for optimization and prediction of parameters in foods, beverages, agriculture and medicine. For brewing, they have been explored to develop rapid methods to assess product quality and acceptability. Different beers (N = 17) were analyzed in triplicates using a robotic pourer, RoboBEER (University of Melbourne, Melbourne, Australia), to assess 15 color and foam-related parameters using computer-vision. Those samples were tested using sensory analysis for acceptability of carbonation mouthfeel, bitterness, flavor and overall liking with 30 consumers using a 9-point hedonic scale. ANN models were developed using 17 different training algorithms with 15 color and foam-related parameters as inputs and liking of four descriptors obtained from consumers as targets. Each algorithm was tested using five, seven and ten neurons and compared to select the best model based on correlation coefficients, slope and performance (mean squared error (MSE). Bayesian Regularization algorithm with seven neurons presented the best correlation (R = 0.98) and highest performance (MSE = 0.03) with no overfitting. These models may be used as a cost-effective method for fast-screening of beers during processing to assess acceptability more efficiently. The use of RoboBEER, computer-vision algorithms and ANN will allow the implementation of an artificial intelligence system for the brewing industry to assess its effectiveness.

Odor Fingerprint Analysis Using Feature Mining Method Based on Olfactory Sensory Evaluation

In this paper, we aim to use odor fingerprint analysis to identify and detect various odors. We obtained the olfactory sensory evaluation of eight different brands of Chinese liquor by a lab-developed intelligent nose. From the respective combination of the time domain and frequency domain, we extract features to reflect the samples comprehensively. However, the extracted feature combined time domain and frequency domain will bring redundant information that affects performance. Therefore, we proposed data by Principal Component Analysis (PCA) and Variable Importance Projection (VIP) to delete redundant information to construct a more precise odor fingerprint. Then, Random Forest (RF) and Probabilistic Neural Network (PNN) were built based on the above. Results showed that the VIP-based models achieved better classification performance than PCA-based models. In addition, the peak performance (92.5%) of the VIP-RF model had a higher classification rate than the VIP-PNN model (90%). In conclusion, odor fingerprint analysis using a feature mining method based on the olfactory sensory evaluation can be applied to monitor product quality in the actual process of industrialization.

Multiscale prediction of functional self-assembled materials using machine learning: high-performance surfactant molecules

Various physical properties of functional materials can be induced by controlling their chemical molecular structures. Therefore, molecular design is crucial in the fields of engineering and materials science. With its remarkable development in various fields, machine learning combined with molecular simulation has recently been found to be effective at predicting the electronic structure of materials (Nat. Commun., 2017, 8, 872 and Nat. Commun., 2017, 8, 13890). However, previous studies have used similar microscale information as input and output data for machine learning, i.e., molecular structures and electronic structures. In this study, we determined whether multiscale data can be predicted using machine learning via a self-assembly functional material system. In particular, we investigated whether machine learning can be used to predict dispersion and viscosity, as the representative physical properties of a self-assembled surfactant solution, from the chemical molecular structures of a surfactant. The results showed that relatively accurate information on these physical properties can be predicted from the molecular structure, suggesting that machine learning can be used to predict multiscale systems, such as surfactant molecules, self-assembled micelle structures, and physical properties of solutions. The results of this study will aid in further development of the application of machine learning to materials science and molecular design.

Assessment of Beer Quality Based on a Robotic Pourer, Computer Vision, and Machine Learning Algorithms Using Commercial Beers

Sensory attributes of beer are directly linked to perceived foam-related parameters and beer color. The aim of this study was to develop an objective predictive model using machine learning modeling to assess the intensity levels of sensory descriptors in beer using the physical measurements of color and foam-related parameters. A robotic pourer (RoboBEER), was used to obtain 15 color and foam-related parameters from 22 different commercial beer samples. A sensory session using quantitative descriptive analysis (QDA^® ) with trained panelists was conducted to assess the intensity of 10 beer descriptors. Results showed that the principal component analysis explained 64% of data variability with correlations found between foam-related descriptors from sensory and RoboBEER such as the positive and significant correlation between carbon dioxide and carbonation mouthfeel (R = 0.62), correlation of viscosity to sensory, and maximum volume of foam and total lifetime of foam (R = 0.75, R = 0.77, respectively). Using the RoboBEER parameters as inputs, an artificial neural network (ANN) regression model showed high correlation (R = 0.91) to predict the intensity levels of 10 related sensory descriptors such as yeast, grains and hops aromas, hops flavor, bitter, sour and sweet tastes, viscosity, carbonation, and astringency.

PRACTICAL APPLICATIONS

This paper is a novel approach for food science using machine modeling techniques that could contribute significantly to rapid screenings of food and brewage products for the food industry and the implementation of Artificial Intelligence (AI). The use of RoboBEER to assess beer quality showed to be a reliable, objective, accurate, and less time-consuming method to predict sensory descriptors compared to trained sensory panels. Hence, this method could be useful as a rapid screening procedure to evaluate beer quality at the end of the production line for industry applications.

Artificial Neural Network to Predict Vine Water Status Spatial Variability Using Multispectral Information Obtained from an Unmanned Aerial Vehicle (UAV)

Water stress, which affects yield and wine quality, is often evaluated using the midday stem water potential (Ψ_stem). However, this measurement is acquired on a per plant basis and does not account for the assessment of vine water status spatial variability. The use of multispectral cameras mounted on unmanned aerial vehicle (UAV) is capable to capture the variability of vine water stress in a whole field scenario. It has been reported that conventional multispectral indices (CMI) that use information between 500–800 nm, do not accurately predict plant water status since they are not sensitive to water content. The objective of this study was to develop artificial neural network (ANN) models derived from multispectral images to predict the Ψ_stem spatial variability of a drip-irrigated Carménère vineyard in Talca, Maule Region, Chile. The coefficient of determination (R²) obtained between ANN outputs and ground-truth measurements of Ψ_stem were between 0.56–0.87, with the best performance observed for the model that included the bands 550, 570, 670, 700 and 800 nm. Validation analysis indicated that the ANN model could estimate Ψ_stem with a mean absolute error (MAE) of 0.1 MPa, root mean square error (RMSE) of 0.12 MPa, and relative error (RE) of −9.1%. For the validation of the CMI, the MAE, RMSE and RE values were between 0.26–0.27 MPa, 0.32–0.34 MPa and −24.2–25.6%, respectively.