Next Article in Journal
Electromagnetic Interference Effect of Portable Electronic Device with Satellite Communication to GPS Antenna
Previous Article in Journal
Lightweight Dual-Attention Network for Concrete Crack Segmentation
Previous Article in Special Issue
Applying a Virtual Art Therapy System Based on the Michelangelo Effect in Patients with Spinal Cord Injury
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

A Comprehensive Review on Sensor-Based Electronic Nose for Food Quality and Safety

1
Automation Department, Technical University of Cluj-Napoca, 400114 Cluj-Napoca, Romania
2
College of Communication and Information, University of Kentucky, Lexington, KY 40506-0224, USA
3
Vice Presidency for Scientific Research and Innovation, Imam Abdulrahman bin Faisal University, Dammam 31441, Saudi Arabia
*
Author to whom correspondence should be addressed.
Sensors 2025, 25(14), 4437; https://doi.org/10.3390/s25144437
Submission received: 18 June 2025 / Revised: 10 July 2025 / Accepted: 14 July 2025 / Published: 16 July 2025
(This article belongs to the Special Issue Sensors in 2025)

Abstract

Food quality and safety are essential for ensuring public health, preventing foodborne illness, reducing food waste, maintaining consumer confidence, and supporting regulatory compliance and international trade. This has led to the emergence of many research works that focus on automating and streamlining the assessment of food quality. Electronic noses have become of paramount importance in this context. We analyze the current state of research in the development of electronic noses for food quality and safety. We examined research papers published in three different scientific databases in the last decade, leading to a comprehensive review of the field. Our review found that most of the efforts use portable, low-cost electronic noses, coupled with pattern recognition algorithms, for evaluating the quality levels in certain well-defined food classes, reaching accuracies exceeding 90% in most cases. Despite these encouraging results, key challenges remain, particularly in diversifying the sensor response across complex substances, improving odor differentiation, compensating for sensor drift, and ensuring real-world reliability. These limitations indicate that a complete device mimicking the flexibility and selectivity of the human olfactory system is not yet available. To address these gaps, our review recommends solutions such as the adoption of adaptive machine learning models to reduce calibration needs and enhance drift resilience and the implementation of standardized protocols for data acquisition and model validation. We introduce benchmark comparisons and a future roadmap for electronic noses that demonstrate their potential to evolve from controlled studies to scalable industrial applications. In doing so, this review aims not only to assess the state of the field but also to support its transition toward more robust, interpretable, and field-ready electronic nose technologies.

1. Introduction

The development of the electronic nose started in the 1960s when researchers began exploring artificial olfaction systems capable of mimicking the human sense of smell. In 1964, Wilkens and Hartman introduced the idea of simulating olfactory processes electronically, contributing to the theoretical groundwork for artificial odor recognition systems [1]. In 1965, Buck et al. explored how chemical compounds could be detected through surface effects on metals and semiconductors, laying the principles that underpin modern gas sensor technology [2]. Around the same time, Dravnieks and Trotter developed a polar vapor detector based on contact potential modulation, an early precursor to volatile organic compounds’ (VOCs) sensing mechanisms [3]. In the early 1980s, Persaud and Dodd proposed a system designed to replicate the mechanisms of human smell perception using an array of gas sensors combined with a pattern recognition algorithm [4]. A few years later, Ikegami and Kaneyasu expanded upon this idea by applying an array of semiconductor gas sensors to distinguish the freshness of food products [5]. The concept of an electronic nose (e-nose) as we know it today is considered to be an intelligent system that integrates an array of chemical or gas sensors with signal processing and pattern recognition mechanisms capable of identifying both simple and complex odors [6,7]. Early e-nose models relied on metal oxide semiconductors and conducting polymer sensors, which, despite their limitations, demonstrated the potential for detecting and discriminating complex odor patterns associated with VOCs. Over time, advancements in sensor technology, nanomaterials, and machine learning have significantly improved the performance of e-nose devices, enabling applications in various industries. Today, modern e-noses incorporate biosensors, artificial intelligence, and miniaturized designs, making them more sensitive, selective, and appropriate for real-world applications.
E-noses play an important role in improving quality control, safety, and efficiency in various sectors, making them an indispensable technological advancement. Studies [8,9] emphasize that e-noses are crucial for the following:
  • Ensuring food quality and safety: E-noses help detect spoilage, contamination, and adulteration in food products, ensuring freshness and quality control.
  • Advancing medical diagnostics: E-noses are used in disease detection by analyzing breath, sweat, or urine to identify biomarkers associated with conditions like diabetes, cancer, and infections.
  • Enhancing environmental monitoring: E-noses detect pollutants, hazardous gases, and air quality changes, aiding in environmental protection and public health.
  • Improving industrial process control: E-noses help monitor manufacturing processes, detect leaks, and ensure consistent product quality in industries such as pharmaceuticals, perfumes, and beverages.
  • Strengthening security and defense: E-noses are used in explosive and drug detection, helping in law enforcement and military and border security operations.
  • Boosting agriculture and farming: E-noses assist in monitoring soil conditions, plant health, and pest infestations by detecting VOCs released by plants to improve crop yields and reduce reliance on harmful pesticides.
  • Ensuring workplace safety: E-noses help prevent occupational hazards by detecting toxic or flammable gases in industrial and laboratory environments.

1.1. Research Questions and Contributions

Ensuring food quality and safety is vital for protecting public health, preventing foodborne illness, reducing food waste, maintaining consumer confidence, and supporting regulatory compliance and international trade. In this context, a timely review of e-noses is essential to capture recent innovations, discuss their advantages and limitations, and summarize the key research directions in order to enhance food quality and safety assurance.

1.1.1. Research Questions

The research questions we explore in this survey include the following:
  • Research Question 1 (RQ1): What are the state-of-the-art research results over the last decade in the field of e-nose systems aimed at food quality and safety?
  • Research Question 2 (RQ2): What lessons have been learned from the design and deployment of e-nose systems in laboratory and industrial settings?
  • Research Question 3 (RQ3): What research gaps exist in the application of e-noses for food quality and safety, and what are the future research directions that we must explore to address these gaps?

1.1.2. Research Contributions

We summarize the main research contributions of this work that address the research questions above as follows:
  • We present an in-depth analysis of research results over the past decade in the e-nose field designed for food quality and safety. We concluded the analysis based on a proposed taxonomy, which we developed through a comprehensive examination of peer-reviewed research papers from three scientific databases. We highlight key technological advances, practical implementations, and performance results obtained across various food sectors.
  • We identify critical lessons learned, such as the importance of e-nose components selection (sensors, signal processing unit, data pattern recognition model) according to the type of food and the need to develop suitable data pattern recognition models, as well as new sensors tailored to food quality and safety assessment.
  • We identify current research gaps, such as the lack of real-world validation and limited sensor sensitivity, and we discuss future research opportunities that will improve the reliability, scalability, and industrial applicability of e-nose technologies in food systems.

1.1.3. Organization of This Paper

We organize the rest of the paper as follows: Section 2 introduces the main components of an electronic nose. Section 3 outlines our review methodology, introduces the final dataset of documents selected for analysis, and presents relevant statistical information. Section 4 analyzes the selected research works based on our proposed taxonomy and discusses lessons learned. Section 5 addresses research gaps and recommends future research directions. Finally, in Section 6, we make some concluding remarks.

2. E-Nose Components

The definition of an electronic nose presented in [6] highlights the key components of such a system that work together to detect, analyze, and interpret odors: the sensor array, the signal processing unit, and the pattern recognition system.
The sensor array is the core component of the e-nose, comprising multiple chemical or gas sensors that respond differently to various VOCs, creating a unique pattern for different odors. To provide a clearer understanding of sensor technologies, we adopted a classification based on the way each type of sensor operates. Common categories include the following:
  • Chemiresistive sensors, such as metal oxide semiconductor (MOS) sensors [10] and carbon nanotube (CNT) sensors [11], detect gases through changes in electrical resistance upon exposure to VOCs.
  • Conductometric sensors, including conducting polymer (CP) sensors [12], alter their conductivity in the presence of gas molecules.
  • Mass-sensitive sensors, such as quartz crystal microbalance (QCM) sensors [13] and surface acoustic wave (SAW) sensors [14], detect gas adsorption by measuring shifts in resonant frequency.
  • Electrochemical sensors [15] convert chemical reactions at the electrode surface into electrical signals.
  • Optical sensors [16] monitor changes in light absorption, fluorescence, or scattering in response to gas exposure.
  • Field-effect transistor (FET)-based sensors [17,18], a growing category often involving nanomaterials or 2D materials, modulate current flow through a semiconductor channel when exposed to target VOCs.
  • Bioelectronic sensors [19] integrate biological recognition elements to selectively detect specific VOCs.
Table 1 presents the advantages and application domains of the types of sensors that can be used in the e-nose development process.
The signal processing unit transforms sensor outputs into digital electronic signals for further analysis. It includes amplifiers, analog-to-digital converters (ADC), and noise filters to refine the data.
The pattern recognition system uses machine learning, artificial intelligence, or statistical algorithms to analyze the sensor data and identify unique odor patterns by comparing against a database of known smells. The most commonly used approaches include principal component analysis (PCA) [20] (reduces the dimensionality of sensor data while retaining key information), linear discriminant analysis (LDA) [21] (classifies odors by maximizing the separation between different groups), artificial neural networks (ANNs) [22] (mimics brain-like processing to learn and recognize odors patterns), support vector machines (SVMs) [23] (separate odor data into different classes using an optimal decision boundary), and k-nearest neighbors (KNNs) [24] (classify odors based on similarity to known reference samples). Within an e-nose, the sensor type determines the data characteristics, which in turn influence the pattern recognition technique needed for effective odor analysis. Therefore, MOS and CP sensors require fast, efficient recognition methods like PCA, ANNs, and SVMs due to large sensor response variations; QCM and SAW sensors produce high-precision frequency-based data, making ANNs and SVMs ideal for classification; optical and electrochemical sensors work well with PCA and LDA for chemical discrimination; and bioelectronic sensors use advanced techniques like ANNs to handle complex biological interactions in odor detection. Table 2 presents the advantages of each pattern recognition technique that can be used in the e-nose development process, the relation between these techniques and sensor types, and also the areas where they can be used.
We found that the results of the odor classification and analysis should be presented in a more user-friendly manner (i.e., mobile and web-based interfaces), making them easily accessible to e-nose end-users. This approach enhances efficient user interaction, enables real-time monitoring, and improves data visualization.
In addition to the previously discussed components, an e-nose should include a sample delivery system to ensure consistent, controlled, and repeatable exposure of VOCs to the sensor array. This component plays a key role in enhancing the accuracy and dependability of odor detection. It can include a sampling port to collect the gas sample, pre-concentrators to enhance sensitivity by capturing and releasing VOCs, filters to remove unwanted contaminants, flow controllers to regulate the gas rate and pressure, temperature and humidity controllers to maintain optimal conditions to prevent variations in sensor responses, a sample chamber to hold the gas sample for uniform interaction with the sensor array, and pumps to facilitate the movement of the gas sample through the system, and a ventilation system to ensure safe disposal of the analyzed gas after detection [25].
Figure 1 presents an overview of the e-nose components, highlighting the key elements we have discussed above and their interconnections.

3. Review Methodology

3.1. Criteria for Selecting Relevant Research Papers Used in This Review

To select the most relevant research papers used in this review, we adopted the strategy in Figure 2.
  • In the first stage, we used the Scopus, IEEE Xplore, and Web of Science (WoS) electronic databases to search for final peer-review English-language documents published between 2014 and 2025. We considered only documents of type articles, reviews, and conference papers. We conducted the search using the following words appearing in the title or abstract of the documents: electronic nose, e-nose, artificial nose, bioelectronic nose, food quality, food safety, and food freshness.
  • In the second stage, we reviewed the titles of the documents retrieved from the search query to remove duplicates.
  • In the third stage, we thoroughly reviewed the full text of the remaining documents (after removing duplicates) before making the final selection, excluding unrelated studies, highlighting the key sections relevant to our review, and identifying the taxonomy of the research literature on the development of e-nose technology/applications in the field of food quality and safety.

3.2. Preliminary Results Obtained

The search query results from the Scopus electronic database, retrieved on 6 March 2025, yielded a total of 69 peer-reviewed documents. These include 40 journal articles (58%), 11 reviews (15.9%), and 18 conference papers (26.1%).
The search results from the IEEE Xplore database, retrieved on 6 March 2025, reveal a total of 117 peer-reviewed documents. These include 21 journal/magazine articles (18%), 2 reviews (1.7%), and 94 conference papers (80.3%).
The search results from the Web of Science database, retrieved on 6 March 2025, indicate a total of 465 peer-reviewed documents. These include 298 journal articles (64.1%), 120 reviews (25.8%), and 47 conference papers (10.1%).
After removing the duplicates, the preliminary documents dataset consists of 314 journal/magazine articles, 122 reviews, and 113 conference papers.
Table 3 presents an overview of the preliminary search results for the documents.

3.3. Final List of Selected Papers and Taxonomy of the Research Literature on E-Nose for Food Quality and Safety

After a comprehensive analysis of the publications obtained in the second stage of the selection process, we excluded those that (a) do not have food quality and safety as their main research objectives (with terms like food, food quality, food safety, or food freshness appearing only in the abstract); (b) primarily review the food quality and safety field, presenting only general information about e-nose, such as definitions and descriptions of basic components, alongside other systems, devices, or techniques for this purpose; (c) rely only on the gas chromatography technique; (d) address food quality for animals; and (e) do not show promising results in terms of performance metrics.
As a result, the final list of publications includes 397 documents, which include 241 journal/magazine articles, 80 reviews from journals and conferences, and 76 conference papers from all the three electronic databases previously mentioned. Table 4 presents a summary of the final list of publications.
Figure 3 presents an additional analysis of the data, which reveals that approximately 20% of the papers in the final dataset are review studies, and most of the original research results have been published in journals/magazines. Additionally, 72% of the papers were published in the past 5 years, which demonstrates the growing interest of the scientific community in this topic.
During this stage, we classified the documents in the final dataset into five distinct classes. The first category covers review studies that focus on the application of e-nose technology in the food industry. The second category includes papers that present the development of new electronic noses for food quality and safety. The third category comprises papers that use commercially available or researcher-developed electronic noses, either alone or in combination with other techniques, for food analysis. The fourth category includes research studies that introduce new/enhanced gas sensors or novel materials for gas sensor development with applicability in the food industry. The last category includes papers that present the outcomes of applying existing algorithms or techniques for pattern recognition or their fusion, on food-related datasets available online. Figure 4a illustrates the taxonomy of the research literature on e-nose technology for food quality and safety based on these categories. Figure 4b presents the distribution of papers in the final list of publications belonging to each identified category.
As Figure 4 shows, most of research efforts have focused on analyzing food data from commercial or laboratory-based e-nose systems, followed closely by studies dedicated to developing new electronic nose systems. There are also several reviews that have been published that cover the latest advancements in the field. Studies concerning the development of new gas sensors and new algorithms/techniques for pattern recognition using existing datasets have also attracted some interest.

4. Analysis of the Research Works from the Final List of Selected Publications

In this section, we emphasize the novel contribution of our review when compared with existing reviews that we selected in our taxonomy. Additionally, we present the most promising research works from the final selection, encompassing the remaining four categories of the taxonomy, and we highlight the key lessons learned from these past studies.

4.1. Reviews

As Figure 4 shows, we identified several review articles in our final selection of papers. These published reviews covered several aspects on the topic of electronic nose.

4.1.1. Reviews on Advancements in Electronic Nose Systems for Food Industry Applications

One category of review papers explored recent advancements in electronic nose systems and their applications in the food industry. The studies [26,27,28,29,30,31,32,33,34,35,36,37] describe the technologies used to develop each of the e-nose components and discuss some proposed solutions of this type of systems. Some reviews focus on the advancement of electronic nose technology in relation to a specific food category: meat [38,39,40,41], berries [42], oils [43,44], fruits and vegetables juice [45], milk and dairy [46,47], tea [48], and wine [49,50].

4.1.2. Reviews on Sensor Development and Pattern Recognition Techniques in Electronic Nose Systems

Reviews such as [51,52,53,54,55,56,57,58] investigate the progress made in the development of sensors with high sensitivity and selectivity for detecting VOCs emitted from food, and highlight the challenges related to sensor stability, cross-sensitivity, environmental interference, and also their integration into e-nose-based applications. The authors of [59,60,61,62] reviewed pattern recognition techniques that can be applied to electronic nose systems. In addition, they also discussed current challenges and potential future directions of these methods.

4.1.3. Reviews on Recent Sensing Technologies for Food Quality Assessment

Another category of reviews focuses on recent sensing technologies for food quality assessment. For example, refs. [63,64,65,66,67,68,69,70,71,72,73] present comprehensive reviews of electronic sensing technologies (e-nose, e-tongue, e-eye) and their applications. Similarly, refs. [74,75] highlight electronic noses as intelligent detection tools in the food industry, along with technologies such as computer vision, intelligent tracing systems, intelligent colorimetric films, and near-infrared spectroscopy.

4.1.4. Our Review on Sensor-Based Electronic Nose for Food Quality and Safety

Our review offers a comprehensive and structured examination of the past decade’s research on sensor-based e-nose devices for ensuring food quality and safety. In contrast with earlier efforts and past publications mentioned in Section 4.1.2, which focus on the technologies used to develop e-nose components and explore proposed solutions, the novelty of our contributions includes the following:
  • Development and application of a unique taxonomy ensuring broad coverage and reduced selection bias: We conducted an extensive analysis of peer-reviewed studies across three major scientific databases, the taxonomy. The taxonomy enables a systematic evaluation of technological advancements (both sensors and pattern recognition techniques), practical implementations, and performance outcomes across the following food and beverage sectors: meat, seafood, vegetables and fruits, spices, oils, coffee, tea, diary, and alcoholic beverages.
  • Decade-long coverage of research results: By capturing trends over an extended period, our review offers an up-to-date perspective on technological evolution and trends.
  • Lessons-learned synthesis: Our review identifies critical lessons learned from the existing literature in each category from the taxonomy we developed. The lessons learned will help guide both future academic research and practical development of e-nose systems for food quality and safety.
  • Identification of unresolved research gaps: Our review reveals notable gaps that must be addressed in the future. These gaps include the lack of e-nose real-world validation, limitations in sensor sensitivity and stability, challenges in achieving miniaturize and portable e-noses, lack of standardized testing protocols, limited real-time processing capabilities, and insufficient support for user-friendly visualization of odor classification and identification outcomes.
Our review serves as a valuable resource for researchers, especially those new to the field of electronic noses for food quality and safety, because it provides a comprehensive foundation and state-of-the-art, in-depth information on current technologies, applications, and research directions in this area.

4.2. Electronic Nose Systems for Food Quality and Safety

Thirty-one percent of the journal/magazine articles and conference papers in the final list of selected publications present electronic nose systems for odor authentication and recognition, for quality assessment, and for quality monitoring of various food and beverage products, including meat (i.e., chicken, beef, pork), seafood (i.e., fish, prawn), vegetables and fruits (i.e., tomato, broccoli, banana, avocado), spices, oils (i.e., palm, olive, sunflower, essential oils), coffee, tea, diary (i.e., milk), and alcoholic beverages (i.e., rice wine, beer, scotch, whiskey, liquor). These papers describe the physical components of the e-noses developed, along with the data analysis approaches used for odor classification and identification. Table 5 briefly presents the analysis of these e-noses from the following perspectives: design architecture (sensor array and signal processing unit), data analysis techniques, evaluation and performance metrics, and application area.

Lessons Learned

Based on the detailed analysis of electronic nose-type systems for food quality and safety that we performed above in Table 5, we found that most of these systems are homemade, low-cost, using commercial MOS sensors and simple microcontroller boards in the Arduino and ESP32 range (Arduino Uno R3, Arduino Mega 2560, Arduino Uno, ATmega8, Arduino Due, PIC, PIC18F45K22, Arduino Nano, Gizduino, ESP32, and the Node MCU IoT platform) or single-board computers such as Raspberry Pi (Raspberry Pi 3 and 4, S3C6410-based Linux platform), or data acquisition cards like NI DAQ card—USB-6009 and PCI6035E, available on the market. Table 6 provides a complete list of all the gas sensors used in e-nose solutions from Table 5 along with their key characteristics. The characteristics presented are important because sensor range defines e-nose sensitivity, and power and response time are critical for portable or multi-sensor systems. For example, heater-based sensors need long startup or preheat times, which limits portability. Table 5 and Table 6 offer researchers a comprehensive overview of the most commonly used sensors in the development of the e-nose system, their corresponding application areas, and the algorithms that achieve good performance levels. These tables provide practical insights and recommendations that will significantly reduce the efforts of designers and implementors required for sensor selection. Creating such selection tables is a crucial preliminary step in the sensor selection process.
For the classification of the acquired data, Table 5 reveals that the main techniques presented in Table 2 (PCA, LDA, ANNs, SVM, KNN) and variations of those (i.e., CNN, PNN, LSTM, MLP, SVR) are the most used. Current trends focus on implementing machine learning technologies, such as CNN and LSTM, and general AI solutions, because processing platforms have become more accessible, affordable, and powerful. Although these solutions involving AI methods and algorithms are not novel, researchers can now truly benefit from their use, thanks to the emergence of cost-effective and high-performance hardware platforms. The use of other models, such as Linear Regression, k-Means, Random Forest, Stochastic Gradient Descent, Naïve Bayes, Fuzzy logic, Discrete Fourier Transformation, Quadratic Discriminant Analysis, and Extreme Gradient Boosting, also yielded results with an accuracy of over 80%, as reported by the authors. Some research works combined PCA with other techniques (i.e., SVN, LDA, PNN, KNN, k-Means) for the following reasons: (a) it reduces the number of features of the acquired data while retaining the most relevant information; (b) it helps filter out sensor noise and irrelevant variations in the data; (c) it pre-processes the data, which results in faster training times, better generalization, higher accuracy, especially when the raw sensor data are noisy or redundant; and (d) it reduces data to two or three dimensions, guiding researchers in the selection of the appropriate pattern recognition technique. In these cases, the performance metrics yield highly encouraging results. We can conclude that PCA acts as a smart pre-processing step that makes the data more manageable and informative for learning tasks.
Most proposed e-nose solutions lack support for the user-friendly visualization of the odor classification and the identification of the results. It is important to ensure that the analysis results are effectively communicated to end users in real time through software applications and easy-to-use graphical user interfaces.
Our in-depth analysis of the research studies also reveals that the developed e-nose systems do not target the achievement of high-performance measurement prototypes or products; rather, the authors focused on concept validation with their prototype systems. This is supported by the following conclusions:
  • Cost: The sensors used in the experiments belong to the cheap components’ class, usually included in gas measurement systems, where their main feature is the detection of the presence of a certain gas component. Another characteristic of such systems is the low manufacturing cost. The documentation that comes with the sensors used is brief, containing little relevant information, omitting aspects like the manufacturer-recommended schematics, calibration and compensation methods depending on temperature and relative humidity values, or formulas for converting the voltage or resistance measured by the microcontroller back into the actual physical quantity measured by the sensor. In many cases, the datasheets do not include important characteristics such as precision, accuracy, repeatability, stability over time, or startup periods. Most sensors used are analog, and they do not integrate calibration circuits, drift, compensation or control mechanisms, or an ADC within the same package. As a result, their overall measurement performance is typically poor, and they are further affected by the required external electronics. The BME688 [125] sensor used in [124,152] stands out in a positive way because it includes important circuitry besides the sensing element, which supports advanced functions such as filtering, signal conditioning, the ADC, the compensation table and algorithm, and digital communication with the processing unit, ESP32. The BME688 development kit uses eight sensors instead of one to form a sensor array, which enhances detection performance, especially for low-cost setups. Though calibrated, sensors differ slightly, and tracking signal trends over time across multiple sensors improve reliability. Additionally, free gas flow causes variations in individual sensor responses before steady state, making arrays beneficial.
  • Power usage: The energy consumption required by the sensors used is high, and they are suitable for integration with systems powered permanently from the main power outlet. Sensors with heaters that are common in most studies have long response times, between 10 and 300 s, and operate at 200–400 °C. This leads to high power consumption unsuitable for portable devices and faster aging that requires frequent recalibrations. The recommended preheating time, or sensor warm-up time, until the first correct measurements can be extensively long, up to 2 to 7 days in some cases. The power consumed during measurement ranges from 0.3 to 1.0 W, and the continuous operation of the heating element in some sensors makes them unsuitable for use in portable electronic nose systems.
  • Data collection: The data acquisition platforms are not designed for instrumentation systems. Most of the proposed solutions use low-resolution ADCs (10- or 12-bit), typically with a 0 to 5 V input range. This leads to an effective resolution per bit of 5 to 10 mV. Temperature and relative humidity compensation are generally based on low-accuracy T and RH sensors (±1 °C for temperature and ±4% for relative humidity), with a few exceptions. In platforms based on ESP32, measurement performance in terms of used digits is further limited by the built-in ADC, which typically offers an effective resolution of only 8 bits. Some proposed electronic nose systems utilize industrial-grade measurement platforms (i.e., PCI6035E, AD7606) and compensation sensors (sensors helping in adjusting the measurement depending on ambient parameters such as temperature, humidity, or pressure) for temperature and relative humidity (i.e., SHT15), which outperform those commonly used in standard gas detectors.
  • Portability: Very few papers focus on low-power or portable systems. However, the question of whether a measurement system can be powered from the main outlet is a valid one. In this case, what are the time and logistical efforts required to make the system operational at a different location? For example, when considering the BME688 sensor, a system using it requires 30 min to reach maximum accuracy after power cycling. Other manufacturers do not specify this time requirement, but in some cases, it could require days.
  • Data processing: Generally, the accuracy of measurement data is verified through thousands of hours of operation and repeated measurements, ideally conducted on multiple similar devices operating in parallel. Some past research solutions [103,115,119,138,141,163] used professional techniques (i.e., Gas Chromatography–Mass Spectrometry (GC–MS) [176]) to compare their experimental results with reference ones. In all the papers that we reviewed, the number of samples collected by the sensor arrays and used by the machine learning algorithms is rather small (<300 samples). In these conditions, expecting authors to validate their work with equipment that has been running for a full year is not feasible. This raises questions regarding the performance metrics obtained. To validate the results obtained, standard test/evaluation scenarios should be run, not just particular test sets created by the authors of the papers.

4.3. Food Analysis Based on Previously Developed/Commercial Electronic Nose Systems

Thirty-nine percent of the publications in the final set of selected publications present food and beverage quality assessment and quality monitoring solutions that integrate commercial e-noses or electronic nose systems developed by other research labs. The research works in this taxonomy category perform some sort data acquisition based on these devices, followed by, in the majority of cases, different data processing and prediction algorithms.

4.3.1. Commercial Electronic Nose Systems

Our analysis found that the most used commercial e-noses are PEN3 [177] from Airsense Analytics Inc., Fox 3000 [178] from Alpha MOS, Fox 4000 [179] from Alpha MOS, FOODsniffer [180], NeOse Pro [181] from Aryballe Technologies, and Cyranose@ 320 [182] produced by Sensigent. Table 7 summarizes their main characteristics, along with a few promising papers that employed them, considering aspects such as data analysis techniques, evaluation and performance metrics, and application domains.

4.3.2. Electronic Noses Developed by Academic Research Groups

Our analysis reveals that several academic research groups developed electronic noses that other researchers later employed them in their own studies. For instance, research groups from the University of Rome Tor Vergata developed LibraNose [203] and also other prototypes [204,205], very similar to LibraNose. Researchers from the Industrial Engineering School of the University of Extremadura, Spain, designed a low-cost, high-accurate electronic nose [206], and those from the Institute of Agrophysics PAS in Lublin, Poland, created the Agrinose [207] system. In the Department of Biosystems Engineering, Faculty of Agriculture, Bu-Ali Sina University, Hamedan, Iran, researchers designed an e-nose [208] that has been successfully used in several studies regarding garlic quality assessment. Table 8 summarizes the main characteristics of these e-noses, along with a few promising papers that employed them, considering aspects such as data analysis techniques, evaluation and performance metrics, and application domains.

Lessons Learned

Numerous research studies use commercial electronic nose instruments, like those mentioned in Table 7, because of their user-friendly operation and dependable performance. These instruments allow researchers to save time, ensure consistent reliability, and focus on more advanced research questions. Additionally, commercial e-noses facilitate the generation of reproducible and comparable results, all without the need to navigate the challenges of developing custom hardware.
Our analysis highlights several ways in which the research community adopted commercial electronic noses, as follows:
  • To apply well-known pattern recognition techniques to assess the quality and safety of different types of food: Studies such as [178,188,194,195,196] used PEN3, Fox 3000, and Fox 4000 commercial electronic noses to apply algorithms such as PCA, Random Forest, and ANNs to evaluate the quality of various food categories (meat, fruits, jams, milk). In [186,189,197], the researchers applied well-known pattern recognition techniques (PCA, PLS-DA, Partial Least Squares regression) to both e-nose data (PEN3, Fox 4000) and chromatography analysis results (GC–MS), leveraging the complementary strengths of these methods in chemical analysis and pattern recognition.
  • To develop novel models or algorithms for odor identification and classification: Studies such as [183,191,193,202] used data collected from commercial electronic nose devices (PEN3, Cyranose@ 320) to develop novel data models for the identification and classification of odors, demonstrating strong performance.
  • To confirm the ability of commercial e-noses to recognize and classify aromas: Studies such as [199,200] proved that the FOODsniffer e-nose can accurately classify meat based on its data analysis, with results validated against GC–MS analysis and physicochemical measurements. The authors of [201] validated NeOse Pro to evaluate the quality of the plant-based beverage by applying PCA and LDA to the e-nose data and comparing the results with those obtained using the same algorithms on GC–MS data.
The previous classification also applies to electronic noses developed by research groups. Table 8 shows that the e-nose designed within the Department of Biosystems Engineering, Bu-Ali Sina University, can successfully assess the quality of garlic by applying well-known pattern recognition techniques (PCA, LDA, SMV, Backpropagation Neural Network) [224], and the e-noses developed by the University of Rome Tor Vergata work well with PLS-DA, PCA, and LDA to identify bacteria in food and beverage [204,205,213]. In [206,209,218], researchers applied techniques such as PCA, Random Forest regression, PLS-DA, and Partial Least Squares to data collected from LibraNose and the e-nose developed by the Industrial Engineering School of the University of Extremadura to assess the quality of meat, roasted coffee, and almonds. The authors of these studies validated the obtained results using GC–MS and High Performance Liquid Chromatography. Studies [210,211,212] used LibraNose data in the development of novel models or algorithms for meat quality assessment, while other studies [221,222] used Agrinose to implement models for the assessment of the suitability of bread for consumption after storage and to identify rapeseed spoilage. These last two studies validated their results by comparing them with those obtained from GC–MS and Fourier Transform Infrared Spectroscopy analysis.
During our analysis, we identified new sensors integrated into e-noses developed by research groups. Table 9 completes the list of sensors presented in Table 6 by including the new ones mentioned in Section 4.3.2 of this review.
An in-depth analysis of the research studies reveals that these instruments can accurately detect certain types of food, but this does not necessarily imply they are suitable for detecting all food classes. The selection of such instruments should not be arbitrary; the selected sensors must be carefully evaluated based on the primary VOCs present in the target sample. Comparing the results obtained from using e-nose data and various algorithms with those from chromatography analyses, which are highly accurate in the identification of VOCs in food, serves to validate the findings. Chromatography provides precise, specific chemical data, but is time-consuming, expensive, and requires skilled operators. E-noses may be more advantageous over GC–MS in distinguishing the integral aroma profile, although they cannot identify the explicit VOCs of different samples. Combining e-noses data analysis with complementary technologies analysis, such as human sensory evaluation, GC–MS, or e-tongue to assess food quality can achieve high detection accuracy. However, this approach often involves significant time due to data fusion processes and incurs substantial costs.
Similar to the approaches analyzed in the first category of our taxonomy, those in the second category also focus on concept validation conducted at the laboratory level.

4.4. Gas Sensors for Electronic Nose Systems

In the final set of selected publications, we also found several research efforts that focus on the development of new gas sensors for the food industry. Table 10 summarizes the key results from the selected articles.

Lessons Learned

Recent advancements in e-nose technologies demonstrate that sensor performance can be significantly improved through material innovation and system integration.
Surface and chemical modifications, such as those applied to Silicon NanoWires or graphene, enhance sensitivity and selectivity, which is critical for detecting specific analyses in complex environments. The use of advanced materials like graphene and graphene with Metal Phthalocyanines enables precise gas discrimination, expanding the applicability of e-noses in food quality and safety.
Bio-inspired approaches, particularly those mimicking the diversity of biological olfactory systems, show strong potential when combined with AI and machine learning for analyzing multi-dimensional VOCs data. Moreover, the integration of e-nose systems with compact, low-power computing architectures, such as memristor-based accelerators, addresses challenges in energy efficiency and real-time data processing, essential for portable devices.
Efforts toward miniaturization, including the use of Film Bulk Acoustic Resonator sensors and reference drift compensation, have made portable e-noses more viable for field deployment. Colorimetric and bioelectronic sensor innovations, leveraging either chemical complexes or olfactory receptors, offer energy-independent or ultra-sensitive detection, pushing the boundaries of low-resource and high-precision sensing.
These developments show that future e-nose systems will increasingly rely on the synergy of novel materials, bio-inspired sensing strategies, smart signal processing, and system-level integration to meet the demands of next-generation applications in food quality and safety assessment.

4.5. Food-Related Datasets and Algorithms/Techniques for Pattern Recognition Used on Them

Several publications in the final collection of articles emphasize the use of available food-related databases, which various algorithms/techniques for pattern recognition used. Table 11 summarizes the e-nose datasets and the related studies that utilized them, as well as the methods and performance metrics reported.

Lesson Learned

Our analysis reveals the following results:
  • Public datasets provide a valuable foundation for developing and testing new models or algorithms for odor identification and classification.
  • Public datasets accelerate comparative research. The availability of well-structured datasets has enabled researchers to benchmark different models, promoting transparency and repeatability. As Table 11 shows, deep learning models outperform traditional classifiers in some cases. Additionally, approaches that combine multiple classifiers tend to boost accuracy and model stability. Such comparisons are possible because the researchers employed the same dataset.
  • Diverse model strategies provide complementary insights. The use of a wide range of algorithms across datasets shows that no single approach performs best across all datasets and applications. Different algorithms excel under specific data characteristics and task requirements.
  • Model performance is dataset dependent. Even if the authors of the cited research efforts reported high accuracies, these are heavily influenced by the specific dataset, number of classes, sensor types, and experimental conditions.

5. Research Gaps and Future Research Opportunities

Despite promising progress, a significant gap remains between lab scale e-nose prototypes and practical, market-ready solutions for food quality and safety monitoring. Current limitations include insufficient sensor sensitivity and stability, challenges in device miniaturization and portability, lack of standardized testing protocols, limited real-time processing capabilities, and lack of support for the user-friendly visualization of the odor classification and identification results. Furthermore, there is a lack of robust data fusion strategies and comprehensive odor reference datasets to support reliable decision making in diverse real-world scenarios.
To address these gaps, future research should focus on the following areas:
  • Sensor technology: We must develop novel gas-sensitive materials with enhanced selectivity and sensitivity for food VOCs, new gas sensors with fast response time, adaptive calibration methods, and sensor baseline correction techniques to improve the stability of the gas sensors and integrate bio-inspired or biomimetic sensors.
  • Data processing: We must implement deep learning algorithms for pattern recognition and VOC classification. We must develop and implement efficient multi-sensor (e-nose, e-tongue, e-eye) data fusion algorithms for a more holistic food profiling approach. Additionally, we must also develop standardized odor databases and reference libraries and real-time data analysis platforms for on-site decision making.
  • Miniaturization and portability: We must integrate micro-electro-mechanical systems/ nano-electro-mechanical systems technology for compact and low-power devices. We must also develop reliable wireless and IoT-enabled e-noses for remote monitoring.
  • Standardization: We must develop standardized testing protocols across different food types and storage conditions.
To consolidate the practical contributions of this review, Table 12 summarizes the key research gaps identified in current electronic nose applications and the corresponding solutions proposed in this work.

6. Conclusions

The application of electronic noses in the food industry has witnessed significant growth over the past decade, demonstrating considerable potential in monitoring food quality and safety, detecting spoilage, assessing freshness, and verifying food authenticity across a wide range of products, including meat, seafood, fruits and vegetables, spices, oils, dairy, and beverages. We arrive at this conclusion based on our analysis of more than 350 peer-reviewed documents retrieved from three scientific databases (Scopus, IEEE Xplore, WoS) using a targeted keyword search. This analysis followed a proposed taxonomy that categorized the publications into five distinct classes: (1) publications that present the development of new electronic noses; (2) publications that use commercially available or researcher-developed electronic noses, either alone or in combination with other techniques; (3) publications that introduce new/enhanced gas sensors or novel materials for the development of gas sensors; (4) publications that present the outcomes of applying existing algorithms or techniques for pattern recognition, or their fusion, on food-related datasets available online; and (5) review studies. Moreover, the analysis revealed that, despite notable progress, several challenges remain. E-nose systems still rely on general-purpose sensor arrays, limiting their adaptability to diverse types of food. In addition, issues related to sensor noise, drift, calibration, temperature, or modularization continue to hinder widespread industrial adoption. Moreover, while many studies report high classification accuracy, there is often a lack of standardization in methodologies/protocols for data pre-processing, feature selection, model deployment, and testing, along with limited application in real-world scenarios. Future research directions must focus on gas sensor technology, data processing, miniaturization and portability, and standardization.

Author Contributions

Conceptualization, T.S. and G.D.M.; methodology, T.S. and G.D.M.; investigation, T.S., G.D.M., T.C.R. and E.A.A.-S.; writing—original draft preparation, T.S., G.D.M., S.F. and T.C.R.; writing—review and editing, S.Z. and S.F.; supervision, S.Z. and S.F. All authors have read and agreed to the published version of the manuscript.

Funding

This research was achieved through the FoodNose project, ctr. no. 4/01.07.2024, funded by the Technical University of Cluj-Napoca through the GNaC ARUT 2023 Program.

Acknowledgments

We thank the anonymous reviewers for their valuable comments, which helped us improve the organization, content, and presentation of this paper.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
ADCanalog-to-digital converters
ANNsartificial neural networks
CNNConvolutional Neural Network
CNTcarbon nanotube
CPconducting polymer
GC–IMSGas Chromatography with Ion Mobility Spectrometry
GC–MSGas Chromatography–Mass Spectrometry
KNNsk-nearest neighbors
LDAlinear discriminant analysis
LSTMLong Short-Term Memory
MLPMultilayer Perceptron
MOSmetal oxide semiconductor
MSEMean Squared Error (MSE)
PCAprincipal component analysis
PICProgrammable Interface Controller
PLS-DAPartial Least Squares-Discriminant Analysis
PNNProbabilistic Neural Network
PPMParts per Million
QCMquartz crystal microbalance
R2R-squared
RMSERoot Mean Square Error
SAWsurface acoustic wave
SVMssupport vector machines
SVRSupport Vector Machine Regression Technique
VOCsvolatile organic compounds
WoSWeb of Science

References

  1. Wilkens, W.F.; Hartman, J.D. An electronic analog for the olfactory processes. Ann. N. Y. Acad. Sci. 1964, 116, 608–612. [Google Scholar] [CrossRef] [PubMed]
  2. Buck, T.; Allen, F.; Dalton, J. Detection of Chemical Species by Surface Effects on Metals and Semiconductors; Bell Telephone System Technical Publications; Bell Telephone Laboratories: Murray Hill, NJ, USA, 1965; Available online: https://books.google.ro/books?id=08s_mwEACAAJ (accessed on 7 July 2025).
  3. Dravnieks, A.; Trotter, P. Polar vapour detector based on thermal modulation of contact potential. J. Sci. Instruments 1965, 42, 624. [Google Scholar] [CrossRef]
  4. Persaud, K.; Dodd, G. Analysis of discrimination mechanisms in the mammalian olfactory system using a model nose. Nature 1982, 299, 352–355. [Google Scholar] [CrossRef] [PubMed]
  5. Ikegami, A.; Kaneyasu, M. Olfactory detection using integrated sensors. Dig. Tech. Pap. Transducers 1985, 85, 136–139. [Google Scholar]
  6. Gardner, J.W.; Bartlett, P.N. A brief history of electronic noses. Sens. Actuators B Chem. 1994, 18, 210–211. [Google Scholar] [CrossRef]
  7. Yu, X.; Fu, L.; Wang, T.; Liu, Z.; Niu, N.; Chen, L. Multivariate chemical analysis: From sensors to sensor arrays. Chin. Chem. Lett. 2024, 35, 109167. [Google Scholar] [CrossRef]
  8. Freddi, S.; Sangaletti, L. Trends in the Development of Electronic Noses Based on Carbon Nanotubes Chemiresistors for Breathomics. Nanomaterials 2022, 12, 2992. [Google Scholar] [CrossRef] [PubMed]
  9. Furizal, F.; Ma’arif, A.; Firdaus, A.; Rahmaniar, W. Future Potential of E-Nose Technology: A Review. Int. J. Robot. Control Syst. 2023, 3, 449–469. [Google Scholar] [CrossRef]
  10. Goel, N.; Kunal, K.; Kushwaha, A.; Kumar, M. Metal oxide semiconductors for gas sensing. Eng. Rep. 2023, 5, e12604. [Google Scholar] [CrossRef]
  11. Luo, K.; Peng, H.; Zhang, B.; Chen, L.; Zhang, P.; Peng, Z.; Fu, X. Advances in carbon nanotube-based gas sensors: Exploring the path to the future. Coord. Chem. Rev. 2024, 518, 216049. [Google Scholar] [CrossRef]
  12. Abdel-Hakim, M.; Sayed, M.M. Conducting Polymers Sensor. In Handbook of Nanosensors: Materials and Technological Applications; Ali, G.A.M., Chong, K.F., Makhlouf, A.S.H., Eds.; Springer Nature: Cham, Switzerland, 2024; pp. 229–262. [Google Scholar] [CrossRef]
  13. Johannsmann, D.; Langhoff, A.; Leppin, C. Studying Soft Interfaces with Shear Waves: Principles and Applications of the Quartz Crystal Microbalance (QCM). Sensors 2021, 21, 3490. [Google Scholar] [CrossRef] [PubMed]
  14. Mandal, D.; Banerjee, S. Surface Acoustic Wave (SAW) Sensors: Physics, Materials, and Applications. Sensors 2022, 22, 820. [Google Scholar] [CrossRef] [PubMed]
  15. Williams, D.E. Electrochemical sensors for environmental gas analysis. Curr. Opin. Electrochem. 2020, 22, 145–153. [Google Scholar] [CrossRef]
  16. Mishra, V.; Rashmi, R.; Khera, S. Optical Gas Sensors; IntechOpen: London, UK, 2022. [Google Scholar] [CrossRef]
  17. Salaün, A.C.; Le Bihan, F.; Pichon, L. Field Effect Transistor Technologies for Biological and Chemical Sensors. In Smart Sensors for Environmental and Medical Applications; John Wiley & Sons, Ltd.: Hoboken, NJ, USA, 2020; Chapter 2; pp. 11–41. [Google Scholar] [CrossRef]
  18. Wang, C.; Chen, Z.; Chan, C.L.J.; Wan, Z.A.; Ye, W.; Tang, W.; Ma, Z.; Ren, B.; Zhang, D.; Song, Z.; et al. Biomimetic olfactory chips based on large-scale monolithically integrated nanotube sensor arrays. Nat. Electron. 2024, 7, 157–167. [Google Scholar] [CrossRef]
  19. Balasurya, S.; Kokilavani, S.; Janani, B.; Sudheer Khan, S. 5–Recent advances in bioelectronic noses based on olfactory receptors. In Nanotechnology-Based E-Noses; Gupta, R.K., Nguyen, T.A., Bilal, M., Ahmadi, M., Eds.; Woodhead Publishing Series in Electronic and Optical Materials; Woodhead Publishing: Sawston, UK, 2023; pp. 125–141. [Google Scholar] [CrossRef]
  20. Jolliffe, I.T.; Cadima, J. Principal component analysis: A review and recent developments. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 2016, 374, 20150202. [Google Scholar] [CrossRef] [PubMed]
  21. Xanthopoulos, P.; Pardalos, P.M.; Trafalis, T.B. Linear Discriminant Analysis. In Robust Data Mining; Springer: New York, NY, USA, 2013; pp. 27–33. [Google Scholar] [CrossRef]
  22. Walczak, S.; Cerpa, N. Artificial Neural Networks. In Encyclopedia of Physical Science and Technology, 3rd ed.; Meyers, R.A., Ed.; Academic Press: New York, NY, USA, 2003; pp. 631–645. [Google Scholar] [CrossRef]
  23. Wang, L. Support Vector Machines: Theory and Applications; Springer: Berlin/Heidelberg, Germany, 2005. [Google Scholar] [CrossRef]
  24. Zhang, Z. Introduction to machine learning: K-nearest neighbors. Ann. Transl. Med. 2016, 4, 218. [Google Scholar] [CrossRef] [PubMed]
  25. Borowik, P.; Grzywacz, T.; Tarakowski, R.; Tkaczyk, M.; Ślusarski, S.; Dyshko, V.; Oszako, T. Development of a Low-Cost Electronic Nose with an Open Sensor Chamber: Application to Detection of Ciboria batschiana. Sensors 2023, 23, 627. [Google Scholar] [CrossRef] [PubMed]
  26. Zhai, Z.; Liu, Y.; Li, C.; Wang, D.; Wu, H. Electronic Noses: From Gas-Sensitive Components and Practical Applications to Data Processing. Sensors 2024, 24, 4806. [Google Scholar] [CrossRef] [PubMed]
  27. Li, Y.; Wang, Z.; Zhao, T.; Li, H.; Jiang, J.; Ye, J. Electronic nose for the detection and discrimination of volatile organic compounds: Application, challenges, and perspectives. TrAC Trends Anal. Chem. 2024, 180, 117958. [Google Scholar] [CrossRef]
  28. Wang, M.; Chen, Y. Electronic nose and its application in the food industry: A review. Eur. Food Res. Technol. 2024, 250, 21–67. [Google Scholar] [CrossRef]
  29. Wang, Y.; Feng, Y.; Zhang, B.; Upadhyay, A.; Xiao, Z.; Luo, Y. Machine learning-supported sensor array for multiplexed foodborne pathogenic bacteria detection and identification. Trends Food Sci. Technol. 2024, 154, 104787. [Google Scholar] [CrossRef]
  30. Ali, A.; Mansol, A.S.; Khan, A.A.; Muthoosamy, K.; Siddiqui, Y. Electronic nose as a tool for early detection of diseases and quality monitoring in fresh postharvest produce: A comprehensive review. Compr. Rev. Food Sci. Food Saf. 2023, 22, 2408–2432. [Google Scholar] [CrossRef] [PubMed]
  31. Roy, M.; Yadav, B. Electronic nose for detection of food adulteration: A review. J. Food Sci. Technol. 2022, 59, 846–858. [Google Scholar] [CrossRef] [PubMed]
  32. Al-Dayyeni, W.S.; Al-Yousif, S.; Taher, M.M.; Al-Faouri, A.W.; Tahir, N.M.; Jaber, M.M.; Ghabban, F.; Najm, I.A.; Alfadli, I.M.; Ameerbakhsh, O.Z.; et al. A Review on Electronic Nose: Coherent Taxonomy, Classification, Motivations, Challenges, Recommendations and Datasets. IEEE Access 2021, 9, 88535–88551. [Google Scholar] [CrossRef]
  33. Mohd Ali, M.; Hashim, N.; Abd Aziz, S.; Lasekan, O. Principles and recent advances in electronic nose for quality inspection of agricultural and food products. Trends Food Sci. Technol. 2020, 99, 1–10. [Google Scholar] [CrossRef]
  34. Bonah, E.; Huang, X.; Aheto, J.H.; Osae, R. Application of electronic nose as a non-invasive technique for odor fingerprinting and detection of bacterial foodborne pathogens: A review. J. Food Sci. Technol. 2020, 57, 1977–1990. [Google Scholar] [CrossRef] [PubMed]
  35. Sanaeifar, A.; ZakiDizaji, H.; Jafari, A.; de la Guardia, M. Early detection of contamination and defect in foodstuffs by electronic nose: A review. TrAC Trends Anal. Chem. 2017, 97, 257–271. [Google Scholar] [CrossRef]
  36. Gliszczyńska-Świgło, A.; Chmielewski, J. Electronic Nose as a Tool for Monitoring the Authenticity of Food. A Review. Food Anal. Methods 2017, 10, 1800–1816. [Google Scholar] [CrossRef]
  37. Loutfi, A.; Coradeschi, S.; Mani, G.K.; Shankar, P.; Rayappan, J.B.B. Electronic noses for food quality: A review. J. Food Eng. 2015, 144, 103–111. [Google Scholar] [CrossRef]
  38. Sun, X.; Wang, S.; Jia, W. Research Progress of Electronic Nose and Near-Infrared Spectroscopy in Meat Adulteration Detection. Chemosensors 2024, 12, 35. [Google Scholar] [CrossRef]
  39. Khaled, A.Y.; Parrish, C.A.; Adedeji, A. Emerging Nondestructive Approaches for Meat Quality and Safety Evaluation—A Review. Compr. Rev. Food Sci. Food Saf. 2021, 20, 3438–3463. [Google Scholar] [CrossRef] [PubMed]
  40. Zaukuu, J.L.Z.; Bazar, G.; Gillay, Z.; Kovacs, Z. Emerging trends of advanced sensor based instruments for meat, poultry and fish quality—A review. Crit. Rev. Food Sci. Nutr. 2020, 60, 3443–3460. [Google Scholar] [CrossRef] [PubMed]
  41. Jia, W.; Liang, G.; Wang, Y.; Wang, J. Electronic Noses as a Powerful Tool for Assessing Meat Quality: A Mini Review. Food Anal. Methods 2018, 11, 2916–2924. [Google Scholar] [CrossRef]
  42. Wang, D.; Zhang, M.; Mujumdar, A.S.; Yu, D. Advanced Detection Techniques Using Artificial Intelligence in Processing of Berries. Food Eng. Rev. 2022, 14, 176–199. [Google Scholar] [CrossRef]
  43. Modesti, M.; Taglieri, I.; Bianchi, A.; Tonacci, A.; Sansone, F.; Bellincontro, A.; Venturi, F.; Sanmartin, C. E-Nose and Olfactory Assessment: Teamwork or a Challenge to the Last Data? The Case of Virgin Olive Oil Stability and Shelf Life. Appl. Sci. 2021, 11, 8453. [Google Scholar] [CrossRef]
  44. Majchrzak, T.; Wojnowski, W.; Dymerski, T.; Gębicki, J.; Namieśnik, J. Electronic noses in classification and quality control of edible oils: A review. Food Chem. 2018, 246, 192–201. [Google Scholar] [CrossRef] [PubMed]
  45. Osmólska, E.; Stoma, M.; Starek-Wójcicka, A. Juice Quality Evaluation with Multisensor Systems—A Review. Sensors 2023, 23, 4824. [Google Scholar] [CrossRef] [PubMed]
  46. Nagraik, R.; Sharma, A.; Kumar, D.; Chawla, P.; Kumar, A.P. Milk adulterant detection: Conventional and biosensor based approaches: A review. Sens. Bio-Sens. Res. 2021, 33, 100433. [Google Scholar] [CrossRef]
  47. Yakubu, H.G.; Kovacs, Z.; Toth, T.; and, G.B. Trends in artificial aroma sensing by means of electronic nose technologies to advance dairy production—A review. Crit. Rev. Food Sci. Nutr. 2023, 63, 234–248. [Google Scholar] [CrossRef] [PubMed]
  48. Kaushal, S.; Nayi, P.; Rahadian, D.; Chen, H.H. Applications of Electronic Nose Coupled with Statistical and Intelligent Pattern Recognition Techniques for Monitoring Tea Quality: A Review. Agriculture 2022, 12, 1359. [Google Scholar] [CrossRef]
  49. Alfieri, G.; Modesti, M.; Riggi, R.; Bellincontro, A. Recent Advances and Future Perspectives in the E-Nose Technologies Addressed to the Wine Industry. Sensors 2024, 24, 2293. [Google Scholar] [CrossRef] [PubMed]
  50. Rodríguez-Méndez, M.L.; De Saja, J.A.; González-Antón, R.; García-Hernández, C.; Medina-Plaza, C.; García-Cabezón, C.; Martín-Pedrosa, F. Electronic Noses and Tongues in Wine Industry. Front. Bioeng. Biotechnol. 2016, 4, 81. [Google Scholar] [CrossRef] [PubMed]
  51. Shuba, A.; Umarkhanov, R.; Bogdanova, E.; Anokhina, E.; Burakova, I. Possibilities of an Electronic Nose on Piezoelectric Sensors with Polycomposite Coatings to Investigate the Microbiological Indicators of Milk. Sensors 2024, 24, 3634. [Google Scholar] [CrossRef] [PubMed]
  52. Nemufulwi, M.I.; Swart, H.C.; Mhlongo, G.H. Advances of Nano-Enabled ZnFe2O4 Based-Gas Sensors for VOC Detection and Their Potential Applications: A Review. Processes 2023, 11, 3122. [Google Scholar] [CrossRef]
  53. Kumar, A.; Castro, M.; Feller, J.F. Review on Sensor Array-Based Analytical Technologies for Quality Control of Food and Beverages. Sensors 2023, 23, 4017. [Google Scholar] [CrossRef] [PubMed]
  54. Wu, K.; Debliquy, M.; Zhang, C. Metal–oxide–semiconductor resistive gas sensors for fish freshness detection. Compr. Rev. Food Sci. Food Saf. 2023, 22, 913–945. [Google Scholar] [CrossRef] [PubMed]
  55. Chia, M.R.; Phang, S.W.; Ahmad, I. Emerging Applications of Versatile Polyaniline-Based Polymers in the Food Industry. Polymers 2022, 14, 5168. [Google Scholar] [CrossRef] [PubMed]
  56. Zhang, H.; Chan-Park, M.B.; Wang, M. Functional Polymers and Polymer–Dye Composites for Food Sensing. Macromol. Rapid Commun. 2020, 41, 2000279. [Google Scholar] [CrossRef] [PubMed]
  57. He, S.; Yuan, Y.; Nag, A.; Feng, S.; Afsarimanesh, N.; Han, T.; Mukhopadhyay, S.C.; Organ, D.R. A Review on the Use of Impedimetric Sensors for the Inspection of Food Quality. Int. J. Environ. Res. Public Health 2020, 17, 5220. [Google Scholar] [CrossRef] [PubMed]
  58. Ponzoni, A.; Baratto, C.; Cattabiani, N.; Falasconi, M.; Galstyan, V.; Nunez-Carmona, E.; Rigoni, F.; Sberveglieri, V.; Zambotti, G.; Zappa, D. Metal Oxide Gas Sensors, a Survey of Selectivity Issues Addressed at the SENSOR Lab, Brescia (Italy). Sensors 2017, 17, 714. [Google Scholar] [CrossRef] [PubMed]
  59. Ji, H.; Pu, D.; Yan, W.; Zhang, Q.; Zuo, M.; Zhang, Y. Recent advances and application of machine learning in food flavor prediction and regulation. Trends Food Sci. Technol. 2023, 138, 738–751. [Google Scholar] [CrossRef]
  60. Lin, Y.; Ma, J.; Wang, Q.; Sun, D.W. Applications of Machine Learning Techniques for Enhancing Nondestructive Food Quality and Safety Detection. Crit. Rev. Food Sci. Nutr. 2023, 63, 1649–1669. [Google Scholar] [CrossRef] [PubMed]
  61. Salehi, F. Recent Advances in the Modeling and Predicting Quality Parameters of Fruits and Vegetables during Postharvest Storage: A Review. Int. J. Fruit Sci. 2020, 20, 506–520. [Google Scholar] [CrossRef]
  62. Sun, Q.; Zhang, M.; Mujumdar, A.S. Recent developments of artificial intelligence in drying of fresh food: A review. Crit. Rev. Food Sci. Nutr. 2019, 59, 2258–2275. [Google Scholar] [CrossRef] [PubMed]
  63. Gil, M.; Rudy, M.; Duma-Kocan, P.; Stanisławczyk, R. Electronic Sensing Technologies in Food Quality Assessment: A Comprehensive Literature Review. Appl. Sci. 2025, 15, 1530. [Google Scholar] [CrossRef]
  64. Tibaduiza, D.; Anaya, M.; Gómez, J.; Sarmiento, J.; Perez, M.; Lara, C.; Ruiz, J.; Osorio, N.; Rodriguez, K.; Hernandez, I.; et al. Electronic Tongues and Noses: A General Overview. Biosensors 2024, 14, 190. [Google Scholar] [CrossRef] [PubMed]
  65. Grassi, S.; Benedetti, S.; Casiraghi, E.; Buratti, S. E-sensing systems for shelf life evaluation: A review on applications to fresh food of animal origin. Food Packag. Shelf Life 2023, 40, 101221. [Google Scholar] [CrossRef]
  66. Poeta, E.; Liboà, A.; Mistrali, S.; Núñez-Carmona, E.; Sberveglieri, V. Nanotechnology and E-Sensing for Food Chain Quality and Safety. Sensors 2023, 23, 8429. [Google Scholar] [CrossRef] [PubMed]
  67. Lu, L.; Hu, Z.; Hu, X.; Li, D.; Tian, S. Electronic tongue and electronic nose for food quality and safety. Food Res. Int. 2022, 162, 112214. [Google Scholar] [CrossRef] [PubMed]
  68. Modesti, M.; Tonacci, A.; Sansone, F.; Billeci, L.; Bellincontro, A.; Cacopardo, G.; Sanmartin, C.; Taglieri, I.; Venturi, F. E-Senses, Panel Tests and Wearable Sensors: A Teamwork for Food Quality Assessment and Prediction of Consumer’s Choices. Chemosensors 2022, 10, 244. [Google Scholar] [CrossRef]
  69. Calvini, R.; Pigani, L. Toward the Development of Combined Artificial Sensing Systems for Food Quality Evaluation: A Review on the Application of Data Fusion of Electronic Noses, Electronic Tongues and Electronic Eyes. Sensors 2022, 22, 577. [Google Scholar] [CrossRef] [PubMed]
  70. Aouadi, B.; Zaukuu, J.L.Z.; Vitális, F.; Bodor, Z.; Fehér, O.; Gillay, Z.; Bazar, G.; Kovacs, Z. Historical Evolution and Food Control Achievements of Near Infrared Spectroscopy, Electronic Nose, and Electronic Tongue—Critical Overview. Sensors 2020, 20, 5479. [Google Scholar] [CrossRef] [PubMed]
  71. Tan, J.; Xu, J. Applications of electronic nose (e-nose) and electronic tongue (e-tongue) in food quality-related properties determination: A review. Artif. Intell. Agric. 2020, 4, 104–115. [Google Scholar] [CrossRef]
  72. Son, M.; Park, T.H. The bioelectronic nose and tongue using olfactory and taste receptors: Analytical tools for food quality and safety assessment. Biotechnol. Adv. 2018, 36, 371–379. [Google Scholar] [CrossRef] [PubMed]
  73. Peris, M.; Escuder-Gilabert, L. Electronic noses and tongues to assess food authenticity and adulteration. Trends Food Sci. Technol. 2016, 58, 40–54. [Google Scholar] [CrossRef]
  74. Wang, D.; Zhang, M.; Jiang, Q.; Mujumdar, A.S. Intelligent System/Equipment for Quality Deterioration Detection of Fresh Food: Recent Advances and Application. Foods 2024, 13, 1662. [Google Scholar] [CrossRef] [PubMed]
  75. Lin, H.; Jiang, H.; Adade, S.Y.S.S.; Kang, W.; Xue, Z.; Zareef, M.; and, Q.C. Overview of advanced technologies for volatile organic compounds measurement in food quality and safety. Crit. Rev. Food Sci. Nutr. 2023, 63, 8226–8248. [Google Scholar] [CrossRef] [PubMed]
  76. Amalia, A.; Rivai, M.; Purwanto, D. Indoor Beef Quality Identification Using Gas Sensor Array and Probabilistic Neural Network Method. In Proceedings of the 2024 IEEE International Conference on Artificial Intelligence and Mechatronics Systems (AIMS), Virtual, 22–23 February 2024; pp. 1–5. [Google Scholar] [CrossRef]
  77. Hanwei Electronics Co., Ltd. MQ Series Gas Sensors. Available online: https://www.hwsensor.com/ (accessed on 14 March 2025).
  78. Figaro Engineering Inc. TGS Series Gas Sensors. Available online: https://www.figaro.co.jp/en/product/sensor/ (accessed on 14 March 2025).
  79. Mohebali, B.; Tahmassebi, A.; Meyer-Baese, A.; Gandomi, A.H. Chapter 14–Probabilistic neural networks: A brief overview of theory, implementation, and application. In Handbook of Probabilistic Models; Samui, P., Tien Bui, D., Chakraborty, S., Deo, R.C., Eds.; Butterworth-Heinemann: Oxford, UK, 2020; pp. 347–367. [Google Scholar] [CrossRef]
  80. Damdam, A.N.; Ozay, L.O.; Ozcan, C.K.; Alzahrani, A.; Helabi, R.; Salama, K.N. IoT-Enabled Electronic Nose System for Beef Quality Monitoring and Spoilage Detection. Foods 2023, 12, 2227. [Google Scholar] [CrossRef] [PubMed]
  81. Aosong Electronics Co., Ltd. AM2302 (DHT22) Temperature and Humidity Sensor. Available online: https://cdn.sparkfun.com/assets/f/7/d/9/c/DHT22.pdf (accessed on 14 March 2025).
  82. Winsen Electronics Technology Co., Ltd. MH-Z19C–NDIR CO2 Sensor Module. Available online: https://www.winsen-sensor.com/sensors/co2-sensor/mh-z19c.html (accessed on 14 March 2025).
  83. Winsen Electronics Technology Co., Ltd. Electrochemical Gas Detection Module. User’s Manual V2.8. Model: ZE03. Available online: https://www.winsen-sensor.com/d/files/ze03-electrochemical-module-manualv2_8(3).pdf (accessed on 14 March 2025).
  84. Espressif Systems. ESP32-S3–Wi-Fi and Bluetooth Microcontroller. Available online: https://www.espressif.com/en/products/socs/esp32-s3 (accessed on 14 March 2025).
  85. Draper, N.R.; Smith, H. Applied Regression Analysis, 3rd ed.; John Wiley & Sons: New York, NY, USA, 1998. [Google Scholar] [CrossRef]
  86. Mada Sanjaya, W.S.; Roziqin, A.; Purnama Sari, N.A.M.; Sintia, P.; Alamsyah, F.; Putra, T.F.; Taqwim, A.; Mubasyir, F.H.; Temiesela, A.W.; Zaman, M.F.B.; et al. Chicken Meat Freshness Evaluation using Support Vector Machine Method and Raspberry Pi 4 based Electronic Nose. In Proceedings of the 2023 IEEE 9th Information Technology International Seminar (ITIS), Batu Malang, Indonesia, 18–20 October 2023; pp. 1–6. [Google Scholar] [CrossRef]
  87. Arduino. Arduino Mega 2560 Rev3–Microcontroller Board. Available online: https://docs.arduino.cc/hardware/mega-2560/?_gl=1*11yinyr*_up*MQ..*_ga*MTQyNTM5MTk0NC4xNzQyMTI2Mzk4*_ga_NEXN8H46L5*MTc0MjEyNjM5Ni4xLjAuMTc0MjEyNjM5Ni4wLjAuMjAwODkwNTQ4NQ (accessed on 14 March 2025).
  88. The Raspberry Pi Foundation. Raspberry Pi 4 Model B–Single Board Computer. Available online: https://www.raspberrypi.com/products/raspberry-pi-4-model-b/ (accessed on 14 March 2025).
  89. Anwar, H.; Anwar, T. Quality assessment of chicken using machine learning and electronic nose. Sens. Bio-Sens. Res. 2025, 47, 100739. [Google Scholar] [CrossRef]
  90. Kuongshun Electronic, Ltd. DHT22/AM2302 Digital Temperature Humidity Sensor. Available online: https://www.kuongshun-ks.com/uno/uno-sensor/dht22-am2302-digital-temperature-humidity-sensor.html (accessed on 14 March 2025).
  91. Breiman, L. Random Forests. Mach. Learn. 2001, 45, 5–32. [Google Scholar] [CrossRef]
  92. Ren, X.; Wang, Y.; Huang, Y.; Sun, D.; Xu, L.; Wu, F. E-nose System using CNN and Abstract Odor Map to Classify Meat Freshness. In Proceedings of the 2023 IEEE 6th Information Technology, Networking, Electronic and Automation Control Conference (ITNEC), Chongqing, China, 24–26 February 2023; Volume 6, pp. 1200–1204. [Google Scholar] [CrossRef]
  93. Hefei Micro Nano Sensing Technology Co., Ltd. HGS1000, HGS1001, HGS1002 MEMS Gas Sensors. Available online: https://www.pilotagemicrosystem.com/ (accessed on 14 March 2025).
  94. O’Shea, K.; Nash, R. An Introduction to Convolutional Neural Networks. arXiv 2015. [Google Scholar] [CrossRef]
  95. Li, L.; Wu, Y.; Ou, Y.; Li, Q.; Zhou, Y.; Chen, D. Research on machine learning algorithms and feature extraction for time series. In Proceedings of the 2017 IEEE 28th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC), Montreal, QC, Canada, 8–13 October 2017; pp. 1–5. [Google Scholar] [CrossRef]
  96. Laga, S.A.; Sarno, R. Optimal Sample Temperature of Electronic Nose For Detecting Beef and Pork Mixture. In Proceedings of the 2019 International Conference on Information and Communications Technology (ICOIACT), Yogyakarta, Indonesia, 24–25 July 2019; pp. 398–402. [Google Scholar] [CrossRef]
  97. Zakaria, E.; Wijaya, D.R.; Gunawan, T. Predicting Microbial Populations in Seafood using Support Vector Regression Algorithms and Electronic Nose. In Proceedings of the 2023 International Conference on Artificial Intelligence Robotics, Signal and Image Processing (AIRoSIP), Yogyakarta, Indonesia, 9–10 August 2023; pp. 280–284. [Google Scholar] [CrossRef]
  98. Drucker, H.; Burges, C.J.C.; Kaufman, L.; Smola, A.; Vapnik, V. Support Vector Regression Machines. Adv. Neural Inf. Process. Syst. (NIPS) 1997, 9, 155–161. Available online: https://proceedings.neurips.cc/paper_files/paper/1996/file/d38901788c533e8286cb6400b40b386d-Paper.pdf (accessed on 16 March 2025).
  99. Wang, X.; Yu, G.; Liu, R.P.; Zhang, J.; Wu, Q.; Su, S.W.; He, Y.; Zhang, Z.; Yu, L.; Liu, T.; et al. Blockchain-Enabled Fish Provenance and Quality Tracking System. IEEE Internet Things J. 2022, 9, 8130–8142. [Google Scholar] [CrossRef]
  100. Hermawan, L.M.; Novamizanti, L.; Wijaya, D.R. Crab Quality Detection with Gas Sensors Using a Machine Learning. In Proceedings of the 2024 IEEE International Conference on Internet of Things and Intelligence Systems (IoTaIS), Bali, Indonesia, 28–30 November 2024; pp. 270–275. [Google Scholar] [CrossRef]
  101. Duda, R.O.; Hart, P.E. Pattern Classification and Scene Analysis; John Wiley & Sons: New York, NY, USA, 1973. [Google Scholar] [CrossRef]
  102. Astuti, S.D.; Isyrofie, A.I.F.A.; Nashichah, R.; Kashif, M.; Mujiwati, T.; Susilo, Y.; Winarno.; Syahrom, A. Gas Array Sensors based on Electronic Nose for Detection of Tuna (Euthynnus Affinis) Contaminated by Pseudomonas aeruginosa. J. Med. Signals Sens. 2022, 12, 306–316. [Google Scholar] [CrossRef] [PubMed]
  103. Jiang, C.; Ning, J.; Mei, Z.; Chen, J.; Gao, Y.; Yi, X.; Wu, P. Development of food electronic nose for prawn (macrobrachium rosenbergii) quality rapid assessment and their relationship with the physicochemical index. Int. J. Food Prop. 2021, 24, 346–353. [Google Scholar] [CrossRef]
  104. Padilla, D.A.; Krystine, D.; Anticuando, M.; Directo, C.K.R. An Electronic Nose Approach in Identifying Ripe Tomato Fruit Using LeNet-5 Structure. In Proceedings of the 2024 7th International Conference on Information and Computer Technologies (ICICT), Honolulu, HI, USA, 15–17 March 2024; pp. 223–227. [Google Scholar] [CrossRef]
  105. Anticuando, M.K.D.; Directo, C.K.R.; Padilla, D.A. Electronic Nose and Deep Learning Approach in Identifying Ripe Lycopersicum esculentum L. TomatoFruit. In Proceedings of the 2022 13th International Conference on Computing Communication and Networking Technologies (ICCCNT), Kharagpur, India, 3–5 October 2022; pp. 1–6. [Google Scholar] [CrossRef]
  106. Affognon, L.; Diallo, A.; Diallo, C.; Ezin, E.C. Supervised Learning Models for Tomato Quality Prediction Using Electronic Nose Data. In Proceedings of the 2023 International Conference on Electrical, Computer and Energy Technologies (ICECET), Cape Town, South Africa, 16–17 November 2023; pp. 1–6. [Google Scholar] [CrossRef]
  107. Concepcion, R.S.; Sybingco, E.; Lauguico, S.C.; Dadios, E.P. Implementation of Multilayer Perceptron Neural Network on Quality Assessment of Tomato Puree in Aerobic Storage using Electronic Nose. In Proceedings of the 2019 IEEE International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM), Bangkok, Thailand, 18–20 November 2019; pp. 65–70. [Google Scholar] [CrossRef]
  108. Concepcion, R.S.; Bandala, A.A.; Bedruz, R.A.R.; Dadios, E.P. Fuzzy Classification Approach on Quality Deterioration Assessment of Tomato Puree in Aerobic Storage using Electronic Nose. In Proceedings of the 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), Laoag, Philippines, 29 November–1 December 2019; pp. 1–6. [Google Scholar] [CrossRef]
  109. e Gizmo Mechatronix Central. Gizduino: Arduino Compatible Kit. 2011. Available online: https://www.e-gizmo.net/oc/kits%20documents/Gizduino/Gizduino%20manual.pdf (accessed on 16 March 2025).
  110. Yen, J. Fuzzy logic-a modern perspective. IEEE Trans. Knowl. Data Eng. 1999, 11, 153–165. [Google Scholar] [CrossRef]
  111. Ebenezer L, B.; A, S.; Baskar, C. Freshness Monitoring of Broccoli Using Intelligent Electronic Nose Systems for Sustainability in Food Supply Chains. In Proceedings of the 2024 International Conference on Smart Electronics and Communication Systems (ISENSE), Kottayam, India, 6–7 December 2024; pp. 1–6. [Google Scholar] [CrossRef]
  112. van der Maaten, L.; Hinton, G. Visualizing Data using t-SNE. J. Mach. Learn. Res. 2008, 9, 2579–2605. Available online: http://jmlr.org/papers/v9/vandermaaten08a.html (accessed on 16 March 2025).
  113. MacQueen, J.B. Some Methods for Classification and Analysis of Multivariate Observations. In Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability; University of California Press: Berkeley, CA, USA, 1967; Volume 1, pp. 281–297. Available online: https://projecteuclid.org/euclid.bsmsp/1200512992 (accessed on 16 March 2025).
  114. Hochreiter, S.; Schmidhuber, J. Long Short-Term Memory. Neural Comput. 1997, 9, 1735–1780. [Google Scholar] [CrossRef] [PubMed]
  115. Ezhilan, M.; Nesakumar, N.; Babu, K.J.; Srinandan, C.; Rayappan, J.B.B. Freshness Assessment of Broccoli using Electronic Nose. Measurement 2019, 145, 735–743. [Google Scholar] [CrossRef]
  116. Microchip Technology Inc. ATmega8. Available online: https://www.microchip.com/en-us/product/atmega8 (accessed on 18 March 2025).
  117. Rana, S.; Bisht, Y.; Tyagi, G.; Badoni, A. FruitfulInsight: IOT and ML-Based System to Predict Expiry of Fruits. In Proceedings of the 2024 2nd International Conference on Device Intelligence, Computing and Communication Technologies (DICCT), Dehradun, India, 15–16 March 2024; pp. 148–153. [Google Scholar] [CrossRef]
  118. Espressif Systems. NodeMcu Connect Things EASY. Available online: https://www.nodemcu.com/index_en.html (accessed on 18 March 2025).
  119. Chen, L.Y.; Wu, C.C.; Chou, T.I.; Chiu, S.W.; Tang, K.T. Development of a Dual MOS Electronic Nose/Camera System for Improving Fruit Ripeness Classification. Sensors 2018, 18, 3256. [Google Scholar] [CrossRef] [PubMed]
  120. National Instruments. NI USB-6008/6009. 2023. Available online: https://www.ni.com/docs/en-US/bundle/usb-6008-6009-feature/page/introduction.html (accessed on 16 March 2025).
  121. Elbi, M.D.; Özgören Çapraz, E.; Şahin, E.; Koyuncuoğlu, M.U.; Tuncer, C. A classification based on support vector machines for monitoring avocado fruit quality. Pamukkale Üniversitesi Mühendislik Bilim. Derg. 2024, 30, 343–353. [Google Scholar] [CrossRef]
  122. Arduino. Arduino Due–Microcontroller Board. Available online: https://docs.arduino.cc/hardware/due/ (accessed on 18 March 2025).
  123. Caya, M.V.C.; Cruz, F.R.G.; Fernando, C.M.N.; Lafuente, R.M.M.; Malonzo, M.B.; Chung, W.Y. Monitoring and Detection of Fruits and Vegetables Spoilage in the Refrigerator using Electronic Nose Based on Principal Component Analysis. In Proceedings of the 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), Laoag, Philippines, 29 November–1 December 2019; pp. 1–6. [Google Scholar] [CrossRef]
  124. Tămâian, A.; Folea, S. Spoiled Food Detection Using a Matrix of Gas Sensors. In Proceedings of the 2024 IEEE International Conference on Automation, Quality and Testing, Robotics (AQTR), Cluj-Napoca, Romania, 16–18 May 2024; pp. 1–5. [Google Scholar] [CrossRef]
  125. Bosch Sensortec. BME688–Digital Low Power Gas, Pressure, Temperature & Humidity Sensor with AI. Technical Report, Bosch Sensortec. 2024. Available online: https://www.bosch-sensortec.com/products/environmental-sensors/gas-sensors/bme688/ (accessed on 18 March 2025).
  126. Adafruit. Adafruit HUZZAH32–ESP32 Feather. Technical Report, Adafruit Industries. 2024. Available online: https://cdn-learn.adafruit.com/downloads/pdf/adafruit-huzzah32-esp32-feather.pdf (accessed on 18 March 2025).
  127. Omatu, S.; Yano, M. E-nose system by using neural networks. Neurocomputing 2016, 172, 394–398. [Google Scholar] [CrossRef]
  128. Bhujbal, K.; Rawat, A. Artificial Nose: Machine Learning model for categorization & identification of spices. In Proceedings of the 2023 International Conference on Computational Intelligence, Communication Technology and Networking (CICTN), Ghaziabad, India, 20–21 April 2023; pp. 596–599. [Google Scholar] [CrossRef]
  129. Kukade, M.; Karve, T.; Gharpure, D. Identification and Classification of Spices by Machine Learning. In Proceedings of the 2019 IEEE International Conference on Intelligent Systems and Green Technology (ICISGT), Visakhapatnam, India, 29–30 June 2019; pp. 1–13. [Google Scholar] [CrossRef]
  130. Yunxiang, L.; Tianwei, L. Study on electronic nose and algorithm for identification of spice. In Proceedings of the 2016 First IEEE International Conference on Computer Communication and the Internet (ICCCI), Wuhan, China, 13–15 October 2016; pp. 518–521. [Google Scholar] [CrossRef]
  131. Analog Devices. AD7606: 16-Bit, 8-Channel, Simultaneous Sampling Analog-to-Digital Converter (ADC). 2019. Available online: https://www.analog.com/en/products/ad7606.html (accessed on 18 March 2025).
  132. Samsung Electronics. S3C6410: ARM11-Based Mobile Application Processor. 2009. Available online: https://www.samsung.com/semiconductor/ (accessed on 18 March 2025).
  133. Choirul Amri, T.; Sarno, R.; Sunaryono, D.; Ahsana Putri, R. Clustering in A Sensor Array System Based on The Distribution of Volatile Compounds from Palm Oil Using Electronic Nose. In Proceedings of the 2024 16th International Conference on Information Technology and Electrical Engineering (ICITEE), Bali, Indonesia, 23–25 October 2024; pp. 189–194. [Google Scholar] [CrossRef]
  134. Ghosal, A.; Nandy, A.; Das, A.K.; Goswami, S.; Panday, M. A Short Review on Different Clustering Techniques and Their Applications. In Proceedings of the Emerging Technology in Modelling and Graphics; Mandal, J.K., Bhattacharya, D., Eds.; Springer: Singapore, 2020; pp. 69–83. [Google Scholar] [CrossRef]
  135. Ivanov, S.; Todorov, T.; Nenov, T.; Wilk-Jakubowski, J. Multisensor Sunflower Oil Quality Assessment System Based on “Electronic Nose”. In Proceedings of the 2023 International Conference Automatics and Informatics (ICAI), Varna, Bulgaria, 5–7 October 2023; pp. 290–294. [Google Scholar] [CrossRef]
  136. SGX Sensortech. MICS-6814: Multiple Gas Sensor. Available online: https://www.sgxsensortech.com/content/uploads/2015/02/1143_Datasheet-MiCS-6814-rev-8.pdf (accessed on 20 March 2025).
  137. Microchip Technology Inc. MCP9700. Available online: https://www.microchip.com/en-us/product/mcp9700#Documentation (accessed on 20 May 2025).
  138. Gonzalez Viejo, C.; Fuentes, S. Digital Detection of Olive Oil Rancidity Levels and Aroma Profiles Using Near-Infrared Spectroscopy, a Low-Cost Electronic Nose and Machine Learning Modelling. Chemosensors 2022, 10, 159. [Google Scholar] [CrossRef]
  139. Oates, M.J.; Fox, P.; Sanchez-Rodriguez, L.; Ángel, A.; Carbonell-Barrachina.; Ruiz-Canales, A. DFT based classification of olive oil type using a sinusoidally heated, low cost electronic nose. Comput. Electron. Agric. 2018, 155, 348–358. [Google Scholar] [CrossRef]
  140. Sidhu, T.; Bhajla, B.; Das, S. Numerical algorithms for protection and metering devices. In Encyclopedia of Electrical and Electronic Power Engineering; García, J., Ed.; Elsevier: Oxford, UK, 2023; pp. 45–87. [Google Scholar] [CrossRef]
  141. Zorpeykar, S.; Mirzaee-Ghaleh, E.; Karami, H.; Ramedani, Z.; Wilson, A.D. Electronic Nose Analysis and Statistical Methods for Investigating Volatile Organic Compounds and Yield of Mint Essential Oils Obtained by Hydrodistillation. Chemosensors 2022, 10, 486. [Google Scholar] [CrossRef]
  142. Rasekh, M.; Karami, H.; Wilson, A.D.; Gancarz, M. Classification and Identification of Essential Oils from Herbs and Fruits Based on a MOS Electronic-Nose Technology. Chemosensors 2021, 9, 142. [Google Scholar] [CrossRef]
  143. Graboski, A.; Feltes, G.; Zakrzevski, C.; Shimizu, F.M.; Steffens, J.; Paroul, N.; Steffens, C. Adulteration of Clove Essential Oil: Detection Using an Electronic Nose with Polymeric Gas Sensors. Food Anal. Methods 2024, 17, 296–308. [Google Scholar] [CrossRef]
  144. Chen, I.T.; Chen, C.C.; Dai, H.J.; Rianto, B.; Huang, S.K.; Lee, C.H. An Incremental Learning Method for Preserving World Coffee Aromas by Using an Electronic Nose and Accumulated Specialty Coffee Datasets. IEEE Trans. Agrifood Electron. 2024, 2, 12–27. [Google Scholar] [CrossRef]
  145. Chen, T.; Guestrin, C. XGBoost: A Scalable Tree Boosting System. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, 13–17 August 2016; pp. 785–794. [Google Scholar] [CrossRef]
  146. Greff, K.; Srivastava, R.K.; Koutník, J.; Steunebrink, B.R.; Schmidhuber, J. LSTM: A Search Space Odyssey. IEEE Trans. Neural Netw. Learn. Syst. 2017, 28, 2222–2232. [Google Scholar] [CrossRef] [PubMed]
  147. Huang, S.; Riemenschneider, L.; Panes-Ruiz, L.; Ibarlucea, B.; Cuniberti, G. Discrimination of Complex Mixtures Using Carbon Nanotubes-based Multichannel Electronic Nose: Coffee Aromas. In Proceedings of the 2023 IEEE Nanotechnology Materials and Devices Conference (NMDC), Paestum, Italy, 22–25 October 2023; pp. 1–4. [Google Scholar] [CrossRef]
  148. Sberveglieri, V.; Pulvirenti, A.; Comini, E.; Carmona, E.N. What happens at the aroma of coffee beans after roasting? MOX nanowire technology by Novel Electronic Nose to discover the fingerprint. Int. J. Smart Sens. Intell. Syst. 2019, 12, 1–4. [Google Scholar] [CrossRef]
  149. Thazin, Y.; Pobkrut, T.; Kerdcharoen, T. Prediction of Acidity Levels of Fresh Roasted Coffees Using E-nose and Artificial Neural Network. In Proceedings of the 2018 10th International Conference on Knowledge and Smart Technology (KST), Chiang Mai, Thailand, 31 January–3 February 2018; pp. 210–215. [Google Scholar] [CrossRef]
  150. Broomhead, D.S.; Lowe, D. Multivariable functional interpolation and adaptive networks. Complex Syst. 1988, 2, 321–355. Available online: https://sci2s.ugr.es/keel/pdf/algorithm/articulo/1988-Broomhead-CS.pdf (accessed on 18 March 2025).
  151. Andre, R.S.; Campaner, K.; Facure, M.H.M.; Mercante, L.A.; Bogusz, S.; Correa, D.S. Nanocomposite-Based Chemiresistive Electronic Nose and Application in Coffee Analysis. ACS Food Sci. Technol. 2021, 1, 1464–1471. [Google Scholar] [CrossRef]
  152. Sanislav, T.; Sipos, D.A.; Mois, G.; Folea, S. Intelligent System for Coffee Odor Identification and Classification. In Proceedings of the 2024 32nd Telecommunications Forum (TELFOR), Belgrade, Serbia, 26–27 November 2024; pp. 1–4. [Google Scholar] [CrossRef]
  153. Ketkar, N. Stochastic Gradient Descent. In Deep Learning with Python: A Hands-On Introduction; Apress: Berkeley, CA, USA, 2017; pp. 113–132. [Google Scholar] [CrossRef]
  154. Kingma, D.; Ba, J. Adam: A Method for Stochastic Optimization. In Proceedings of the International Conference on Learning Representations (ICLR), San Diega, CA, USA, 7–9 May 2015. [Google Scholar] [CrossRef]
  155. Banerjee(Roy), R.; Chattopadhyay, P.; Tudu, B.; Bhattacharyya, N.; Bandyopadhyay, R. Artificial flavor perception of black tea using fusion of electronic nose and tongue response: A Bayesian statistical approach. J. Food Eng. 2014, 142, 87–93. [Google Scholar] [CrossRef]
  156. National Instruments. PCI-6035E: 16-Bit, 200 kS/s, 32-Channel, Analog I/O Device. 2024. Available online: https://www.artisantg.com/TestMeasurement/73100-4/National-Instruments-PCI-6035E-200-kS-s-16-Bit-Multifunction-DAQ?srsltid=AfmBOoq-LdUEsAT7NfmWSD1hvS7BW2hKCeG_cP2spv1Q5ImNk-P02smn (accessed on 18 March 2025).
  157. Zhang, D. Bayesian Classification. In Fundamentals of Image Data Mining: Analysis, Features, Classification and Retrieval; Springer International Publishing: Cham, Switzerland, 2019; pp. 161–178. [Google Scholar] [CrossRef]
  158. Mu, F.; Gu, Y.; Zhang, J.; Zhang, L. Milk Source Identification and Milk Quality Estimation Using an Electronic Nose and Machine Learning Techniques. Sensors 2020, 20, 4238. [Google Scholar] [CrossRef] [PubMed]
  159. LaValley, M.P. Logistic Regression. Circulation 2008, 117, 2395–2399. [Google Scholar] [CrossRef] [PubMed]
  160. Zheng, W.; Wang, Y.; Liang, X.; Zhang, A. Origin identification for rice wines based on an electronic nose and convolution dot-product attention mechanism. Sens. Actuators A Phys. 2024, 375, 115521. [Google Scholar] [CrossRef]
  161. Zhang, K.; Sun, M.; Han, T.X.; Yuan, X.; Guo, L.; Liu, T. Residual Networks of Residual Networks: Multilevel Residual Networks. IEEE Trans. Circuits Syst. Video Technol. 2018, 28, 1303–1314. [Google Scholar] [CrossRef]
  162. Sun, J.T.; Lee, C.H.; Rianto, B.; Huang, S.K. Combining an Electronic Nose System with Artificial Intelligence Methods for Odor Identification of Spirits. In Proceedings of the 2024 10th International Conference on Applied System Innovation (ICASI), Kyoto, Japan, 17–21 April 2024; pp. 332–334. [Google Scholar] [CrossRef]
  163. Zhang, W.; Liu, T.; Brown, A.; Ueland, M.; Forbes, S.L.; Su, S.W. The Use of Electronic Nose for the Classification of Blended and Single Malt Scotch Whisky. IEEE Sens. J. 2022, 22, 7015–7021. [Google Scholar] [CrossRef]
  164. Hastie, T.; Tibshirani, R.; Friedman, J. The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd ed.; Springer: New York, NY, USA, 2009; Available online: https://hastie.su.domains/ElemStatLearn/ (accessed on 18 March 2025).
  165. Sun, J.T.; Lee, C.H. AI-driven Sensor Array Electronic Nose System for Authenticating and Recognizing Aromas in Spirit Samples. Sens. Mater. 2025, 37, 23–40. [Google Scholar] [CrossRef]
  166. Caterini, A.L.; Chang, D.E. Recurrent Neural Networks. In Deep Neural Networks in a Mathematical Framework; Springer International Publishing: Cham, Switzerland, 2018; pp. 59–79. [Google Scholar] [CrossRef]
  167. Liang, K.; Wang, Y.; He, J.Q.; Meng, H.; Chen, S.Q.; Chang, J.Y.; Gao, J.M.; Wang, J.H.; Feng, L. An Electronic Nose Based on Copper Oxide Heterojunctions for Rapid Assessment of Liquor. Chin. J. Anal. Chem. 2019, 47, e19073–e19080. [Google Scholar] [CrossRef]
  168. da Silva Torres, E.A.F.; Garbelotti, M.L.; Moita Neto, J.M. The application of hierarchical clusters analysis to the study of the composition of foods. Food Chem. 2006, 99, 622–629. [Google Scholar] [CrossRef]
  169. Jin, J.; Tang, X.; Cai, Y.; Han, Y.; Wang, M.; Zheng, H.; Hui, G. Optimization of eigenvalue selection in Chinese liquors discrimination based on electronic nose. Food Bioprocess Technol. 2014, 8, 270–276. [Google Scholar] [CrossRef]
  170. Harnsoongnoen, S.; Babpan, N.; Srisai, S.; Kongkeaw, P.; Srisongkram, N. A Portable Electronic Nose Coupled with Deep Learning for Enhanced Detection and Differentiation of Local Thai Craft Spirits. Chemosensors 2024, 12, 221. [Google Scholar] [CrossRef]
  171. Sensirion. SHT15 Digital Temperature and Humidity Sensor. Available online: https://sensirion.com/media/documents/BD45ECB5/61642783/Sensirion_Humidity_Sensors_SHT1x_Datasheet.pdf (accessed on 20 March 2025).
  172. Bourlard, H.A.; Morgan, N. Multilayer Perceptrons. In Connectionist Speech Recognition: A Hybrid Approach; Springer: Boston, MA, USA, 1994; pp. 59–80. [Google Scholar] [CrossRef]
  173. Viejo, C.G.; Fuentes, S.; Godbole, A.; Widdicombe, B.; Unnithan, R.R. Development of a low-cost e-nose to assess aroma profiles: An artificial intelligence application to assess beer quality. Sens. Actuators B Chem. 2020, 308, 127688. [Google Scholar] [CrossRef]
  174. Hanwei Electronics Co., Ltd. MG811 Carbon Dioxide (CO2) Sensor Module. Available online: https://www.yumpu.com/en/document/read/33225829/mg811-datasheet (accessed on 20 March 2025).
  175. Aosong Electronics Co., Ltd. AM2320 Digital Temperature and Humidity Sensor. Available online: https://cdn-shop.adafruit.com/product-files/3721/AM2320.pdf (accessed on 20 March 2025).
  176. McNair, H.M.; Miller, J.M.; Snow, N.H. Basic Gas Chromatography, 3rd ed.; Wiley: Hoboken, NJ, USA, 2019; p. 288. [Google Scholar]
  177. AIRSENSE Analytics GmbH. PEN3–Portable Electronic Nose Technical Data. Available online: https://airsense.com/sites/default/files/flyer_pen.pdf (accessed on 20 March 2025).
  178. Hazarika, S.; Choudhury, R.; Montazer, B.; Medhi, S.; Goswami, M.P.; Sarma, U. Detection of Citrus Tristeza Virus in Mandarin Orange Using a Custom-Developed Electronic Nose System. IEEE Trans. Instrum. Meas. 2020, 69, 9010–9018. [Google Scholar] [CrossRef]
  179. Esfahani, S.; Wicaksono, A.; Mozdiak, E.; Arasaradnam, R.P.; Covington, J.A. Non-Invasive Diagnosis of Diabetes by Volatile Organic Compounds in Urine Using FAIMS and Fox4000 Electronic Nose. Biosensors 2018, 8, 121. [Google Scholar] [CrossRef] [PubMed]
  180. FOODSniffer. The Foodsniffer. Available online: https://www.myfoodsniffer.com/product.html (accessed on 23 March 2025).
  181. Aryballe. NeOse Advance Datasheet. Available online: http://aryballe.com/wp-content/uploads/2021/07/NeOse_Advance_DS_072121.pdf (accessed on 20 March 2025).
  182. Sensigent Intelligent Sensing Solution. Cyranose Electronic Nose. Available online: https://www.sensigent.com/cyranose-320.html (accessed on 24 March 2025).
  183. Yu, Y.; Li, Q.; Shi, Y. Integrating Criss-Cross Attention Mechanism for Origin Traceability of Peanuts Using Electronic Nose Technology. In Proceedings of the 2024 7th International Symposium on Autonomous Systems (ISAS), Chongqing, China, 7–9 May 2024; pp. 1–5. [Google Scholar] [CrossRef]
  184. Huang, Z.; Wang, X.; Wei, Y.; Huang, L.; Shi, H.; Liu, W.; Huang, T.S. CCNet: Criss-Cross Attention for Semantic Segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 2020, 45, 6896–6908. [Google Scholar] [CrossRef] [PubMed]
  185. Kodithuwakku, P.; Jayasundara, D.; Munaweera, I.; Jayasinghe, R.; Thoradeniya, T.; Bogahawatta, A.; Manuda, K.R.J.; Weerasekera, M.; Kottegoda, N. Ilmenite-Grafted Graphene Oxide as an Antimicrobial Coating for Fruit Peels. ACS Omega 2024, 9, 26568–26581. [Google Scholar] [CrossRef] [PubMed]
  186. Zhang, P.P.; Gui, X.J.; Fan, X.H.; Han, L.; Li, H.Y.; Li, X.P.; Dong, F.Y.; Wang, Y.L.; Jing, Y.; Shi, J.H.; et al. Quality identification of Amomi fructus using E-nose, HS-GC-IMS, and intelligent data fusion methods. Front. Chem. 2025, 13, 1544743. [Google Scholar] [CrossRef] [PubMed]
  187. Lee, L.C.; Liong, C.Y.; Jemain, A.A. Partial least squares-discriminant analysis (PLS-DA) for classification of high-dimensional (HD) data: A review of contemporary practice strategies and knowledge gaps. Analyst 2018, 143, 3526–3539. [Google Scholar] [CrossRef] [PubMed]
  188. Yang, Y.; Wei, L. Application of E-nose technology combined with artificial neural network to predict total bacterial count in milk. J. Dairy Sci. 2021, 104, 10558–10565. [Google Scholar] [CrossRef] [PubMed]
  189. Mu, W.; Zhao, Y.; Wang, Z.; He, Y.; Yang, C.; Wang, J. Combining lipase enzymatic techniques and antioxidants on the flavor of structured lipids (SLs) prepared from goat butter and coconut oil. Food Biosci. 2024, 60, 104332. [Google Scholar] [CrossRef]
  190. Boyacı, E.; Rodríguez-Lafuente, Á.; Gorynski, K.; Mirnaghi, F.; Souza-Silva, É.A.; Hein, D.; Pawliszyn, J. Sample preparation with solid phase microextraction and exhaustive extraction approaches: Comparison for challenging cases. Anal. Chim. Acta 2015, 873, 14–30. [Google Scholar] [CrossRef] [PubMed]
  191. Dai, J.; Li, W.; Dong, G. Dung Beetle Optimizer Algorithm and Machine Learning-Based Genome Analysis of Lactococcus lactis: Predicting Electronic Sensory Properties of Fermented Milk. Foods 2024, 13, 1958. [Google Scholar] [CrossRef] [PubMed]
  192. Xue, J.; Shen, B. Dung beetle optimizer: A new meta-heuristic algorithm for global optimization. J. Supercomput. 2023, 79, 7305–7336. [Google Scholar] [CrossRef]
  193. Zheng, W.; Yuan, Q.; Zhang, A.; Lei, Y.; Pan, G. Data augmentation of flavor information for electronic nose and electronic tongue: An olfactory-taste synesthesia model combined with multiblock reconstruction method. Expert Syst. Appl. 2025, 272, 126810. [Google Scholar] [CrossRef]
  194. Knežević-Jugović, Z.; Culetu, A.; Mijalković, J.; Duta, D.; Stefanović, A.; Šekuljica, N.; Đorđević, V.; Antov, M. Impact of Different Enzymatic Processes on Antioxidant, Nutritional and Functional Properties of Soy Protein Hydrolysates Incorporated into Novel Cookies. Foods 2023, 12, 24. [Google Scholar] [CrossRef] [PubMed]
  195. Wang, Z.; Shi, Y.; Zhou, K.; Zhou, H.; Li, X.; Li, C.; Wang, Z.; Xu, B. Effects of different thermal temperatures on the shelf life and microbial diversity of Dezhou-braised chicken. Food Res. Int. 2020, 136, 109471. [Google Scholar] [CrossRef] [PubMed]
  196. Culetu, A.; Manolache, F.; Duta, D. Exploratory Study of Physicochemical, Textural and Sensory Characteristics of Sugar-Free Traditional Plum Jams. J. Texture Stud. 2014, 45, 138–147. [Google Scholar] [CrossRef]
  197. Gu, S.-q.; Wang, X.-c.; Tao, N.-p.; Wu, N. Characterization of volatile compounds in different edible parts of steamed Chinese mitten crab (Eriocheir sinensis). Food Res. Int. 2013, 54, 81–92. [Google Scholar] [CrossRef]
  198. Wold, S.; Sjöström, M.; Eriksson, L. PLS-regression: A basic tool of chemometrics. Chemom. Intell. Lab. Syst. 2001, 58, 109–130. [Google Scholar] [CrossRef]
  199. Castrica, M.; Chiesa, L.M.; Nobile, M.; De Battisti, F.; Siletti, E.; Pessina, D.; Panseri, S.; Balzaretti, C.M. Rapid safety and quality control during fish shelf-life by using a portable device. J. Sci. Food Agric. 2021, 101, 315–326. [Google Scholar] [CrossRef] [PubMed]
  200. Ramírez, H.L.; Soriano, A.; Gómez, S.; Iranzo, J.U.; Briones, A.I. Evaluation of the Food Sniffer electronic nose for assessing the shelf life of fresh pork meat compared to physicochemical measurements of meat quality. Eur. Food Res. Technol. 2018, 244, 1047–1055. [Google Scholar] [CrossRef]
  201. Papp, Z.; Nemeth, L.G.; Nzetchouang Siyapndjeu, S.; Bufa, A.; Marosvölgyi, T.; Gyöngyi, Z. Classification of Plant-Based Drinks Based on Volatile Compounds. Foods 2024, 13, 4086. [Google Scholar] [CrossRef] [PubMed]
  202. Ferreira, I.; Dias, T.; Mouazen, A.M.; Cruz, C. Using Science and Technology to Unveil The Hidden Delicacy Terfezia arenaria, a Desert Truffle. Foods 2023, 12, 3527. [Google Scholar] [CrossRef] [PubMed]
  203. Di Natale, C.; Macagnano, A.; Davide, F.; D’Amico, A.; Paolesse, R.; Boschi, T.; Faccio, M.; Ferri, G. An electronic nose for food analysis. Sens. Actuators B Chem. 1997, 44, 521–526. [Google Scholar] [CrossRef]
  204. Muñoz-Castells, R.; Modesti, M.; Moreno-García, J.; Rodríguez-Moreno, M.; Catini, A.; Capuano, R.; Di Natale, C.; Bellincontro, A.; Moreno, J. Differentiation through E-nose and GC-FID data modeling of rosé sparkling wines elaborated via traditional and Charmat methods. J. Sci. Food Agric. 2025, 105, 1439–1447. [Google Scholar] [CrossRef] [PubMed]
  205. Capuano, R.; Paba, E.; Mansi, A.; Marcelloni, A.M.; Chiominto, A.; Proietto, A.R.; Zampetti, E.; Macagnano, A.; Lvova, L.; Catini, A.; et al. Aspergillus Species Discrimination Using a Gas Sensor Array. Sensors 2020, 20, 4004. [Google Scholar] [CrossRef] [PubMed]
  206. Barea-Ramos, J.D.; Cascos, G.; Mesías, M.; Lozano, J.; Martín-Vertedor, D. Evaluation of the Olfactory Quality of Roasted Coffee Beans Using a Digital Nose. Sensors 2022, 22, 8654. [Google Scholar] [CrossRef] [PubMed]
  207. Gancarz, M.; Rusinek, R.; Nawrocka, A.; Tadla, M.; Gawrysiak-Witulska, M.; Wawrzyniak, J. Electronic Sensor Array of the Electronic Nose. P.419383, 7 November 2016. [Google Scholar]
  208. Makarichian, A.; Amiri Chayjan, R.; Ahmadi, E.; Mohtasebi, S.S. Assessment the influence of different drying methods and pre-storage periods on garlic (Allium sativum L.) aroma using electronic nose. Food Bioprod. Process. 2021, 127, 198–211. [Google Scholar] [CrossRef]
  209. Estelles-Lopez, L.; Ropodi, A.; Pavlidis, D.; Fotopoulou, J.; Gkousari, C.; Peyrodie, A.; Panagou, E.; Nychas, G.J.; Mohareb, F. An automated ranking platform for machine learning regression models for meat spoilage prediction using multi-spectral imaging and metabolic profiling. Food Res. Int. 2017, 99, 206–215. [Google Scholar] [CrossRef] [PubMed]
  210. Kodogiannis, V.S.; Alshejari, A. Neuro-fuzzy based identification of meat spoilage using an electronic nose. In Proceedings of the 2016 IEEE 8th International Conference on Intelligent Systems (IS), Sofia, Bulgaria, 4–6 September 2016; pp. 96–103. [Google Scholar] [CrossRef]
  211. Kodogiannis, V. Application of an Electronic Nose Coupled with Fuzzy-Wavelet Network for the Detection of Meat Spoilage. Food Bioprocess Technol. 2017, 10, 730–749. [Google Scholar] [CrossRef]
  212. Mohareb, F.; Papadopoulou, O.; Panagou, E.; Nychas, G.J.; Bessant, C. Ensemble-based support vector machine classifiers as an efficient tool for quality assessment of beef fillets from electronic nose data. Anal. Methods 2016, 8, 3711–3721. [Google Scholar] [CrossRef]
  213. Modesti, M.; Alfieri, G.; Chieffo, C.; Mencarelli, F.; Vannini, A.; Catalani, A.; Chilosi, G.; Bellincontro, A. Destructive and non-destructive early detection of postharvest noble rot (Botrytis cinerea) in wine grapes aimed at producing high-quality wines. J. Sci. Food Agric. 2024, 104, 2314–2325. [Google Scholar] [CrossRef] [PubMed]
  214. Bosch Sensortec. Gas Sensor BME680. Available online: https://www.bosch-sensortec.com/products/environmental-sensors/gas-sensors/bme680/#technical (accessed on 29 May 2025).
  215. Sensirion. Datasheet SGP30. Available online: https://sensirion.com/media/documents/984E0DD5/61644B8B/Sensirion_Gas_Sensors_Datasheet_SGP30.pdf (accessed on 29 May 2025).
  216. Adafruit CCS811 Air Quality Sensor. Available online: https://cdn-learn.adafruit.com/downloads/pdf/adafruit-ccs811-air-quality-sensor.pdf (accessed on 29 May 2025).
  217. iAQ-Core Indoor Air Quality Sensor Module. Available online: https://www.es.co.th/Schemetic/PDF/IAQ-CORE-C.PDF (accessed on 29 May 2025).
  218. Mesías, M.; Barea-Ramos, J.D.; Lozano, J.; Morales, F.J.; Martín-Vertedor, D. Application of an Electronic Nose Technology for the Prediction of Chemical Process Contaminants in Roasted Almonds. Chemosensors 2023, 11, 287. [Google Scholar] [CrossRef]
  219. AS-MLV-P2 Air Quality Sensor. Available online: https://ro.mouser.com/datasheet/2/1081/AS_MLV_P2_Datasheet-1890483.pdf (accessed on 29 May 2025).
  220. Figaro USA, Inc. TGS 8100–For the Detection of Air Contaminants. Available online: www.figarosensor.com/product/docs/TGS8100%280914%29.pdf (accessed on 29 May 2025).
  221. Rusinek, R.; Gancarz, M.; Nawrocka, A. Application of an electronic nose with novel method for generation of smellprints for testing the suitability for consumption of wheat bread during 4-day storage. LWT 2020, 117, 108665. [Google Scholar] [CrossRef]
  222. Rusinek, R.; Gancarz, M.; Krekora, M.; Nawrocka, A. A Novel Method for Generation of a Fingerprint Using Electronic Nose on the Example of Rapeseed Spoilage. J. Food Sci. 2019, 84, 51–58. [Google Scholar] [CrossRef] [PubMed]
  223. Berthomieu, C.; Hienerwadel, R. Fourier transform infrared (FTIR) spectroscopy. Photosynth. Res. 2009, 101, 157–170. [Google Scholar] [CrossRef] [PubMed]
  224. Makarichian, A.; Chayjan, R.A.; Ahmadi, E.; Mohtasebi, S.S.; Zafari, D. Use of E-Nose in inspecting the effect of processing type on the aroma of garlic (Allium sativum L.): A critical hint in the quality assessment. Food Prod. Process. Nutr. 2024, 6, 52. [Google Scholar] [CrossRef]
  225. Hecht-Nielsen. Theory of the backpropagation neural network. In Proceedings of the International 1989 Joint Conference on Neural Networks, Washington, DC, USA, 18–22 June 1989; Volume 1, pp. 593–605. [Google Scholar] [CrossRef]
  226. Linevych, Y.; Koval, V.; Dusheiko, M.; Lakyda, M.; Kavraska, N. Odor Sensors Based on Silicon Nanowires With a Modified Surface for Electronic Nose Application. IEEE Sens. Lett. 2025, 9, 1–4. [Google Scholar] [CrossRef]
  227. Mo, F.; Ala, W.; Spano, C.E.; Piccinini, G.; Graziano, M.; Ardesi, Y. Single-Molecule Graphene Junction Aflatoxin Sensors: Chemo-Physical Insights from Ab Initio Simulations. In Proceedings of the 2024 IEEE 24th International Conference on Nanotechnology (NANO), Gijon, Spain, 8–11 July 2024; pp. 420–423. [Google Scholar] [CrossRef]
  228. Qu, C.; Fang, H.; Yu, F.; Chen, J.; Su, M.; Liu, H. Artificial nose of scalable plasmonic array gas sensor for Multi-Dimensional SERS recognition of volatile organic compounds. Chem. Eng. J. 2024, 482, 148773. [Google Scholar] [CrossRef]
  229. Tan, J.; Duan, S.; Wang, L.; Yan, J. Multi-Gas Sensing Electronic Nose Using Memristor-Based In-Memory Computing. IEEE Sens. J. 2023, 23, 28526–28534. [Google Scholar] [CrossRef]
  230. Freddi, S.; Marzuoli, C.; Pagliara, S.; Drera, G.; Sangaletti, L. Targeting biomarkers in the gas phase through a chemoresistive electronic nose based on graphene functionalized with metal phthalocyanines. RSC Adv. 2023, 13, 251–263. [Google Scholar] [CrossRef] [PubMed]
  231. Wu, C.; Li, J. Portable FBAR based E-nose for cold chain real-time bananas shelf time detection. Nanotechnol. Precis. Eng. 2023, 6, 013004. [Google Scholar] [CrossRef]
  232. Sun, L.; Rotaru, A.; Garcia, Y. A non-porous Fe(II) complex for the colorimetric detection of hazardous gases and the monitoring of meat freshness. J. Hazard. Mater. 2022, 437, 129364. [Google Scholar] [CrossRef] [PubMed]
  233. Shin, N.; Lee, S.; Pham Ba, V.; Park, T.H.; Hong, S. Micelle-stabilized Olfactory Receptors for a Bioelectronic Nose Detecting Butter Flavors in Real Fermented Alcoholic Beverages. Sci. Rep. 2020, 10, 9064. [Google Scholar] [CrossRef] [PubMed]
  234. Wijaya, D.R. Dataset for Electronic Nose from Various Beef Cuts. Available online: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/XNFVTS (accessed on 23 March 2025).
  235. Wijaya, D.R.; Sarno, R.; Zulaika, E. DWTLSTM for electronic nose signal processing in beef quality monitoring. Sens. Actuators B Chem. 2021, 326, 128931. [Google Scholar] [CrossRef]
  236. Espina, K.R.; Naval, P. 1D-CNN Model for Fast Assessment of Beef Freshness from Noisy Electronic Nose Signal. In Proceedings of the 2024 6th International Conference on Computer Communication and the Internet (ICCCI), Tokyo, Japan, 14–16 June 2024; pp. 67–71. [Google Scholar] [CrossRef]
  237. Feyzioglu, A.; Taspinar, Y.S. Beef Quality Classification with Reduced E-Nose Data Features According to Beef Cut Types. Sensors 2023, 23, 2222. [Google Scholar] [CrossRef] [PubMed]
  238. Enériz, D.; Medrano, N.; Calvo, B. An FPGA-Based Machine Learning Tool for In-Situ Food Quality Tracking Using Sensor Fusion. Biosensors 2021, 11, 366. [Google Scholar] [CrossRef] [PubMed]
  239. Kaya, A.; Keçeli, A.S.; Catal, C.; Tekinerdogan, B. Sensor Failure Tolerable Machine Learning-Based Food Quality Prediction Model. Sensors 2020, 20, 3173. [Google Scholar] [CrossRef] [PubMed]
  240. Sarno, R.; Wijaya, D.R.; Sabilla, S.I. The dataset for pork adulteration from electronic nose system. IEEE Dataport 2020. [Google Scholar] [CrossRef]
  241. Avian, C.; Leu, J.; Prakosa, S.; Faisal, M. An Improved Classification of Pork Adulteration in Beef Based on Electronic Nose Using Modified Deep Extreme Learning with Principal Component Analysis as Feature Learning. Food Anal. Methods 2022, 15, 3020–3031. [Google Scholar] [CrossRef]
  242. GAMBOA, J.C.R.; Estrada, E.S.A.; da Silva, A.; Ferreira, T. Electronic Nose Dataset for Detection of Wine Spoilage Thresholds. Available online: https://data.mendeley.com/datasets/vpc887d53s/3 (accessed on 23 March 2025).
  243. Rodriguez Gamboa, J.C.; Albarracin, E.S.; da Silva, A.J.; Tiago, T.A. Electronic nose dataset for detection of wine spoilage thresholds. Data Brief 2019, 25, 104202. [Google Scholar] [CrossRef] [PubMed]
  244. Bulucu, P.; Özkurt, N.; Güzeliş, C.; Yıldız, O. Wine Quality Assessment with Application Specific 2D Single Channel Convolutional Neural Networks. In Proceedings of the 2021 13th International Conference on Electrical and Electronics Engineering (ELECO), Bursa, Turkey, 25–27 November 2021; pp. 369–372. [Google Scholar] [CrossRef]
  245. Hanif, S.; Wijaya, D.R.; Wikusna, W. K-Nearest Neighbors Algorithm for Prediction Shelf Life of Rice Based on Electronic Nose Dataset. In Proceedings of the 2021 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob), Bandung, Indonesia, 8–10 April 2021; pp. 116–120. [Google Scholar] [CrossRef]
  246. Aulia, I.; Wijaya, D.R.; Hidayat, W. Rice Quality Detection Using Gradient Tree Boosting Based On Electronic Nose Dataset. In Proceedings of the 2021 International Conference on Artificial Intelligence and Mechatronics Systems (AIMS), Bandung, Indonesia, 28–30 April 2021; pp. 1–5. [Google Scholar] [CrossRef]
  247. Sari, I.M.; Wijaya, D.R.; Hidayat, W.; Kannan, R. An Approach to Classify Rice Quality using Electronic Nose Dataset-based Naïve Bayes Classifier. In Proceedings of the 2021 International Symposium on Electronics and Smart Devices (ISESD), Bandung, Indonesia, 29–30 June 2021; pp. 1–5. [Google Scholar] [CrossRef]
  248. Erlangga, F.; Wijaya, D.R.; Wikusna, W. Electronic Nose Dataset for Classifying Rice Quality using Neural Network. In Proceedings of the 2021 9th International Conference on Information and Communication Technology (ICoICT), Yogyakarta, Indonesia, 3–5 August 2021; pp. 462–466. [Google Scholar] [CrossRef]
Figure 1. The main components of e-nose and their interactions.
Figure 1. The main components of e-nose and their interactions.
Sensors 25 04437 g001
Figure 2. The three stages of the review methodology.
Figure 2. The three stages of the review methodology.
Sensors 25 04437 g002
Figure 3. Statistics from the final list of publications.
Figure 3. Statistics from the final list of publications.
Sensors 25 04437 g003
Figure 4. (a) Taxonomy of the research literature on electronic nose for food quality and safety. (b) The distribution (by categories) of papers in the final list of publications.
Figure 4. (a) Taxonomy of the research literature on electronic nose for food quality and safety. (b) The distribution (by categories) of papers in the final list of publications.
Sensors 25 04437 g004
Table 1. E-nose sensor array types, their advantages, and application sectors.
Table 1. E-nose sensor array types, their advantages, and application sectors.
Type of SensorAdvantagesApplication Sector
Chemiresistive sensors: MOSHigh sensitivity, high selectivity, durability, long lifespan, fast response timeAir quality monitoring, food freshness detection, industrial gas sensing, medical diagnostic
Chemiresistive sensors: CNTUltra-high sensitivity, fast response time, low power consumption, miniaturization potentialBreath analysis for disease detection, air quality monitoring, workspace safety
Conductometric sensors: CPFast response time, low power consumption, tunable sensitivityMedical diagnostics, food quality assessment, environmental monitoring
Mass-sensitive sensors: QCMHigh sensitivity, ability to detect low-concentration gasesBreath analysis, detection of toxic substances, fragrance quality control
Mass-sensitive sensors: SAWFast response time, small size, high ruggednessExplosive and drug detection, environmental monitoring, workspace safety
Electrochemical sensorsHigh selectivity in terms of the electrochemical properties of target VOCs, low power consumption, reliable detection of specific gasesToxic gas detection, breath analysis, air quality monitoring
Optical sensorsNon-contact sensing, high specificity in terms of the chemical identity of VOCs, fast response timeIndustrial gas detection, medical diagnostics, hazardous material monitoring, food quality assessment
FET sensorsHigh sensitivity, fast response time, compatibility with nanomaterials and 2D materials, fast electronic responseMedical diagnostics, food quality assessment, environmental monitoring, industrial process control, security and defense
Bioelectronic sensorsHigh specificity in terms of molecular recognition of target VOCs, biomimetic functionality, potential for personalized diagnosticsDisease detection, food quality monitoring
Table 2. Pattern recognition technique, their advantages, related sensor type, and application areas.
Table 2. Pattern recognition technique, their advantages, related sensor type, and application areas.
Pattern Recognition TechniqueAdvantagesType of SensorApplication Sector
PCA
-
Reduces data dimensionality without losing key information
-
Enhances visualization and interpretation
-
Improves classification performance
-
Speeds up computational processing
-
Enables unsupervised odor classification
MOS, CP, QCM, SAW, CNT, optical sensors, electrochemical sensorsFood quality control, medical diagnostics, environmental monitoring, industrial process control
LDA
-
Maximizes odor class separation
-
Enhances classification accuracy
-
Reduces data dimensionality
-
Speeds up computational processing
-
Suitable for small and well-defined datasets
MOS, CP, QCM, SAW, optical sensors, electrochemical sensorsMedical diagnostics, food quality control, environmental monitoring, security and defense
ANNs
-
Can handle nonlinear and complex data
-
High accuracy in odor classification
-
Self-learning and adaptability
-
Real-time processing capability
-
Multi-modal data fusion
-
Noise tolerance
MOS, CP, QCM, SAW, CNT, bioelectronic sensorsMedical diagnostics, food quality control, environmental monitoring, security and defense
SVM
-
High classification accuracy
-
Handles nonlinear and high-dimensional data
-
Effective for small datasets
-
Works well with multiple sensor types
-
Suitable for binary and multi-class odor classification
-
Works well in real-time applications
MOS, CP, QCM, SAW, CNTMedical diagnostics, food quality control, environmental monitoring, security and defense, workspace safety
KNN
-
Easy to implement and interpret
-
Effective for small and medium-sized datasets
-
Works well with multiple sensor types
-
Adaptable for classification and regression
-
Suitable for real-time odor analysis
MOS, CP, QCM, SAW, CNTFood quality monitoring, environmental monitoring, medical diagnostics, industrial process control
Table 3. Preliminary search results.
Table 3. Preliminary search results.
6 March 2025Journal/
Magazine Articles
ReviewsConference PapersTotal
Scopus40111869
IEEE Xplore21294117
WoS29812047465
Total359133159651
Total without duplicates314122113549
Table 4. Final search results.
Table 4. Final search results.
6 March 2025Journal/
Magazine Articles
ReviewsConference PapersTotal
Scopus32611
IEEE Xplore503540
WoS2337835346
Final total2418076397
Table 5. Summary of electronic nose systems for food quality and safety.
Table 5. Summary of electronic nose systems for food quality and safety.
PaperDesign ArchitectureData Analysis TechniquesEvaluation and Performance MetricsApplication Area
Electronic Nose Systems for Meat Quality and Safety
[76]Sensor array: MQ-137, MQ-136 [77], TGS2602 [78]; Signal processing unit: Arduino Uno R3PCAAccuracy: 94.9% (fresh/spoiled/rotten)Beef quality assessment
PCA + Probabilistic Neural Network (PNN) [79]Accuracy: 100% (fresh/spoiled)
[80]IoT-enabled e-nose; Sensor array: AM2302 [81], one optical sensor from Winsen Electronics Technology Co., Zhengzhou, China, MH-Z19C [82], ZE03-NH3, ZE03-C2H4 [83]; Signal processing unit: ESP32-S3 controller [84]Linear Regression [85]Aerobic bacteria and Pseudomonas species play a crucial role in the production of VOCs in beefBeef quality monitoring and spoilage detection
[86]Sensor array: MQ-2, MQ-3, MQ-4, MQ-6, MQ-7, MQ-8, MQ-9, MQ-135 [77]; Signal processing unit: Arduino Mega 2560 microcontroller [87], Raspberry Pi 4 [88]PCA + SVMAccuracy: 98.49% (healthy/
compromised)
Chicken meat quality assessment
[89]Sensor array: MQ-2, MQ-3, MQ-6, MQ-7, MQ-9, MQ-135, DHT22 [90]; Signal processing unit: Arduino Uno microcontrollerSVM, Linear Regression, KNN, Random Forest [91]Best accuracy: 100% for Random Forest with random split data; 69% for Random Forest with non-randomly split data; 78.5% for SVM with group split data (fresh/semi-fresh/spoiled)Chicken meat quality assessment
[92]Sensor array: HGS1000, HGS1001, HGS1002, HGS1007 [93]; Signal processing unit: 12-bit ADC with four channels of input data; heating voltage can be set between 0 and 2.4VConvolutional Neural Network (CNN) CNN [94] + time series feature extraction [95]Accuracy: 92.1% (fresh/sub-fresh/spoiled) Accuracy: 98.4% (fresh/sub-fresh/spoiled)Pork, beef, mutton, chicken, crab, shrimp, fish meat quality assessment
[96]Sensor array: MQ-2, MQ-4, MQ-6, MQ-9, MQ-135, MQ-136, MQ-137, MQ-138,DHT22; Signal processingunit: N/AKNNAccuracy rates between 97% and 100% (variations of meat with ratio 0%, 10%, 50%, 90%, 100%)Authenticity of beef and pork meat
SVMAccuracy rates between 81.5% and 99.5% (variations of meat with ratio 0%, 10%, 50%, 90%, 100%)
Electronic Nose Systems for Seafood Quality and Safety
[97]Sensor array: MQ-136, MQ-137, MQ-5, MQ-8; Signal processing unit: N/ASupport Vector Machine Regression Technique (SVR) [98]R-squared (R2): 0.981; Root Mean Square Error (RMSE): 0.012Estimation of the microbial population in seafood
[99]IoT-enabled e-nose with image processing capabilities; Sensor array: N/A; Signal processing unit: N/AA nonparametric kernel-based modeling + hidden Markov modelQuality model indices closely align with the manual results provided by quality assurance expertsFish origin verification, fish quality assessment
[100]Sensor array: MQ-1, MQ-2 and two MQ-135; Signal processing unit: ESP32 microcontrollerKNNAccuracy: 98% (fresh/less fresh/
not fresh)
Freshness and quality of crabs
Naïve Bayes [101]Accuracy: 91% (fresh/less fresh/
not fresh)
SVMAccuracy: 87% for SVM (fresh/less fresh/not fresh)
[102]Sensor array: TGS2620, TGS2611, TGS822, TGS832, TGS2602, TGS2600, TGS826, TGS825; Signal processing unit: N/A; Preheating process before using the sensorsPCACumulative variance of the principal component: 95% (fresh/contaminated)Tuna quality assessment (Pseudomonas aeruginosa bacteria)
SVMAccuracy: 99% (fresh/contaminated)
[103]Sensor array: MQ-2, MQ-3, MQ-4, MQ-5, MQ-6, MQ-7, MQ-8, MQ-9; Signal processing unit: N/ALinear RegressionR2: 0.98; Accuracy: 93.75%Prawn quality assessment
Electronic Nose Systems for Vegetables and Fruits Quality and Safety
[104,105]Sensor array: MQ-3, MQ-6, MQ-8, MQ-135; Signal processing unit: ADC for Raspberry Pi 4/
Raspberry Pi 3
CNNAccuracy: 86% (ripe/not ripe/unknown)Identification of the ripening stage of tomato fruits
[106]Sensor array: MQ-135, MQ-136, TGS822, TGS2600, TGS2602, TGS2603, TGS2610, TGS2611, DHT22; Signal processingunit: N/ARandom ForestAccuracy: 94% (good/good/fair/
poor)
Identification of the ripening stage of tomato fruits
KNNAccuracy: 83% (good/good/fair/
poor)
ANNsAccuracy: 79% (good/good/fair/
poor)
SVMAccuracy: 64% (good/good/fair/
poor)
[107,108]Sensor array: MQ-135, MQ-4; Signal processing unit: Gizduino micro- controller [109]ANNsAccuracy: 93.33% (not spoiled/
partially spoiled/
spoiled)
Tomato puree quality assessment
Fuzzy logic technique [110]Accuracy: 90% (not spoiled/
partially spoiled/
spoiled)
[111]Sensor array: TGS2620, TGS823, DHT22; Signal processing unit: Arduino microcontrollerPCA + t-distributed Stochastic Neighbor Embedding [112] + k-Means [113] + Long Short-Term Model [114]Accuracy: 99.46% (fresh/half-spoiled/spoiled)Broccoli quality assessment
[115]Sensor array: TGS880, TGS822, TGS826, TGS2602, TGS2600; Signal processing unit: ATmega8 [116] microcontroller with an integrated ADCPCA + Centroid link- and completely-link cluster analysesSimilarity levels >93% for 3 / 4 of the samples tested (fresh/half/
completely contaminated)
Broccoli quality assessment (Staphylococcus, Salmonella and Shigella)
[117]Sensor array: MQ-2, DHT11; Signal processing unit: Arduino Uno microcontroller and Node MCU [118]
IoT platform
Linear Regression, Random Forest, SVRBest performance with a value of Mean Squared Error (MSE): 0.1207 for Random ForestBanana freshness assessment
[119]Sensor array: TGS2600, TGS2602,TGS2603, TGS2610, TGS2611, TGS2612, TGS2620; Signal processing unit: NI DAQ card, USB-6009 [120]PCA + KNNAccuracy: 98.10% (unripe/half-ripe/fully ripe/overripe)Identification of the ripening stage of banana
PCA + SVMAccuracy: 95.24% (unripe/half-ripe/fully ripe/overripe)
LDA + KNNAccuracy: 90.48% (unripe/half-ripe/fully ripe/overripe)
LDA + SVMAccuracy: 86.67% (unripe/half-ripe/fully ripe/overripe)
[121]Sensor array: MQ-2, MQ-3, MQ-4, MQ-5, MQ-7, MQ-8, MQ-135 sensors; Signal processing unit: Arduino Due [122] microcontrollerSVMAccuracy: 99% (rotten/fresh)Avocado fruits quality assessment
[123]Sensor array: MQ-136, MQ-4, MQ-137, MQ-3, MQ-2, MQ-135, MQ-131, MQ-8, MQ-9; Signal processing unit: Raspberry Pi computerPCA + KNNAccuracy: 92% (spoiled/not spoiled)Fruits (banana, pechay, carrot, grape) quality assessment
[124]Sensor array: eight BME688 gas sensors [125]; Signal processing unit: Adafruit Huzzah32 (ESP32) development board [126]Neural Networks [127]Accuracy: 76% (spoiled/not spoiled)Fruits and vegetables quality assessment
Electronic Nose Systems for Spices’ Quality and Safety
[128,129]Sensor array: TGS800, TGS813, TGS823, TGS2602, TGS2610, TGS2611, TGS2620, MQ-135; Signal processing unit: ADCs of a Programmable Interface Controller (PIC) microcontrollerRandom ForestAccuracy: 100% (nutmeg/
clove/cinnamon)
Identification of nutmeg, clove, and cinnamon
[130]Sensor array: TGS2600, TGS2602, TGS2610, TGS813, TGS822, MQ-138, MQ-2 MQ-8; Signal proc. unit: AD7606 analog-to- digital data acquisition system [131] and S3C6410-based Linux platform [132]PCA + SVMAccuracy: 95%Authenticity of star anise
Electronic Nose Systems for Oils’ Quality and Safety
[133]Sensor array: eight TGS and MQ sensors, and one temperature and relative humidity sensor; Signal proc. unit: RPi computerClustering technique [134]Identification of three classes of palm oil: never heated/heated for 10 to 30 h/heated for 40 to 60 hPalm oil quality assessment
[135]Sensor array: MICS-6814 MOS sensor [136], MCP9700 temperature sensor [137]; Signal processing unit: ADS1015, PIC18F45K22, FT230XS USB/UART converterANNsAccuracy: 99.49% (degummed/extraction/filtered/marketed)Sunflower oil quality assessment
[138]Sensor array: MQ-3, MQ-4, MQ-7, MQ-8, MQ-135, MQ-137, MQ-138, MG-811; Signal processing unit: N/AANNs (classification)Accuracy: 86.5%Extra-virgin olive oil quality assessment
ANNs (regression)Correlation coefficient: 0.93; Slope 0.90
[139]Sensor array: MQ-2, MQ-3, MQ-4, MQ-5, MQ-7, MQ-8, MQ-9, MQ-135; Sensor processing unit: Arduino Nano microcontrollerDiscrete Fourier transform data analysis [140]Accuracy: 91% (extra virgin/virgin); Accuracy: between 67% and 77% (extra virgin/virgin/blend /pomace/fresh air)Olive oil quality assessment
[141]Sensor array: MQ-3, TGS822, MQ-136, MQ-9, TGS813, MQ-135, TGS2602, TGS2620; Sensor processingunit: N/APCATotal variance of the data for distilled water extracts from mint plants: 95%; Total variance of the data for mint essential oil: 89%Mint essential oil and mint distilled water extracts quality assessment
LDAAccuracy for the classification of mint essential oil: 91.33%; Accuracy for mint distilled water extracts: 86.67%
ANNsAccuracy for the classification of distilled water extracts 100%; Accuracy for the classification of mint essential oil 96.7%
[142]Sensor array: MQ-9, MQ-4, MQ-135, MQ-8, TGS2620, MQ-136, TGS813, TGS822, MQ-3; Signal proc. unit: N/APCAAccuracy: 98%Identification of essential oils from herbs and fruits
LDA and Quadratic Discriminant AnalysisAccuracy: 100% (essential oil emissions from herbal leaves/fruits); Accuracy: 100% for Quadratic Discriminant Analysis and 98.9% for LDA (mango/lemon/orange /mint/tarragon/thyme)
SVMAccuracy: 100% (essential oil emissions from herbal leaves/fruits); Accuracy: 98.9% (mango/lemon/orange /mint/tarragon/thyme)
[143]Sensor array: six different polymeric gas sensors (polymeric nanocom- posites of polyaniline with multiwalled carbon nanotubes and graphene oxide); Signal processing unit: N/APCAAccuracy: 99.85%Authenticity of clove oil
Interactive Document Map multivariate projection techniquesAccuracy: 99.81%
LDAAccuracy: 98.30%
Electronic Nose Systems for Coffee and Tea Quality and Safety
[144]Sensor array: MQ-7, MQ-3, MQ-135, TGS2600, TGS2602, TGS2610, TGS2611, TGS2620, DHT22; Signal processing unit: N/AExtreme Gradient Boosting [145]Accuracy rates between 82% and 95% (sixteen classes of coffee)Authenticity of coffee
SVMAccuracy rates between 81% and 95% (sixteen types of coffee)
CNNAccuracy rates between 86% and 98% (sixteen types of coffee)
CNN + Long Short-Term Memory (LSTM) [146]Accuracy rates between 83% and 98% (sixteen types of coffee)
[147]Sensor array: carbon nanotube-based multichannel with 64 interdigital electrodes; Sensor processing unit: N/ALDAAccuracy: 97.4% (three different coffee aromas)Authenticity of coffee
[148]Sensor array: four sensors (SnO2_bs1, ZH0504, SnO2 Au_bs2, SU0303) and two nanowire sensors (Sn-NW1, Sn-NW2); Sensor processing unit: N/APCAN/A (four classes of roasted coffee beans)Analysis of different methods of coffee roasting
[149]Sensor array: TGS821, TGS2444, TGS823, TGS2600, TGS2602, TGS2610, TGS826, TGS2620; Signal processing unit: NI DAQ card, USB-6009PCAAccuracy: 95% (four acidity levels of coffee drinks)Coffee drinks quality assessment
Radial Basis Function Neural Network [150]Accuracy: 94.75% (to predict the scores of acidity level)
[151]Sensor array: six sensing units (nanocomposites that stem from the combination of ZnO, In2O3, and ZnO/In2O3 nanoparticles with polypyrrole and poly(styrenesulfonate)); Signal processing unit: N/APCA + Euclidean distances by dendrogramsN/A (seventeen classes of coffee)Authenticity of coffee
[152]Sensor array: eight BME688 sensors; Signal processing unit: Adafruit Huzzah32 (ESP32) development boardRandom ForestMSE: 0.062Authenticity of coffee
Stochastic Gradient Descent [153]Accuracy: 70.10% (two classes of coffee)
Adam Optimizer [154]Accuracy: 67.70% (two classes of coffee)
[155]Sensor array: TGS832, TGS823, TGS2600, TGS2610, TGS2611; Signal processing unit: PCI6035E data acquisition card [156]Bayesian classification [157]Classification error in percentage 30.91% (four classes of tea)Black tea quality assessment
Electronic Nose Systems for Diary Quality and Safety
[158]Sensor array: TGS2600, TGS822, TGS2611, TGS826, TGS2602, TGS832, TGS2620; Signal processing unit: Arduino microcontrollerPCA + LDA + SVMAccuracy: 85%Identification of milk source
PCA + LDA + Logistic Regression [159]Accuracy: 81.50%
PCA + LDA + Random ForestAccuracy: 80.50%
Electronic Nose Systems for Alcoholic Beverage Quality and Safety
[160]Sensor array: MQ-2, MQ-135, TGS825, WSP-2110, MP-503, TGS2602, WSP-1110, MQ-138, MQ-137, MQ-136; Sensor processing unit: N/AConvolution Dot-Product Attention Mechanism [160], Residual network (ResNet50 mode) [161]Accuracy: 98.47% (ten production origins of rice wines)Identification of the origins of rice wines
[162]Sensor array: TGS2600, TGS2602, TGS2603, TGS2610, TGS2611, TGS2620, TGS813, TGS822; Sensor processing unit: N/ALDA + PCA + CNN-LSTMAccuracy: 98% (whiskey/brandy /gin/vodka/tequila)Identification of various types of spirit samples
[163]Sensor array: TGS2600, TGS2603, TGS2610D, TGS2611C, TGS2620; Signal processing unit: N/ALinear DiscriminantAccuracy: 69.23% (six brands of whiskey); Accuracy: 100% (whiskey regions of origin)Authenticity of whiskey
SVMAccuracy: 82.05% (six brands of whiskey); Accuracy: 98.72% (whiskey regions of origin)
KNNAccuracy: 61.54% (six brands of whiskey); Accuracy: 92.31% (whiskey regions of origin)
Bagged Tree [91]Accuracy: 74.36% (six brands of whiskey); Accuracy: 94.87% (whiskey regions of origin)
Subspace Discriminant [164]Accuracy: 70.51% (six brands of whiskey); Accuracy: 100% (whiskey regions of origin)
[165]Sensor array: TGS2600, TGS2602, TGS2603, TGS2610, TGS2611, TGS2620, TGS813, TGS822, DHT22; Signal processing unit: N/A;CNN-LSTMAccuracy: 93% (three whiskey types)Authentication of whiskey
CNNAccuracy: 91% (three different types of whiskey)
LSTMAccuracy: 91% (three different types of whiskey)
Recurrent Neural Networks [166]Accuracy: 89% (three whiskey types)
[167]Sensor array: eight MOS sensors with two different types of copper oxide heterojunctions, ZnO–CuO and NiO–CuO; Signal processing unit: N/AHierarchical Clustering Analysis [168]Euclidean distance: 0.5 (four samples of Chinese Jing Wine)Identification of the same liquors manufactured in different years
[169]Sensor array: TGS825, TGS821, TGS826, TGS822, TGS842, TGS813, TGS2610, TGS2201; Signal processing unit: N/APCA + Signal-to-Noise RatioFirst two principal components captured 92.47% of data variance (thirteen varieties of Chinese liquor)Identification of several types of liquors
[170]Sensor array: MQ-3, MQ-6, MQ-9, MQ-135, MQ-136, MQ-137, MQ-138, MQ-139, SHT15 [171]; Signal processing unit: NI DAQ card, USB-6009PCA + Multi-Layer Perceptron (MLP) [172]Accuracy: 100% (three distinct local Thai spirits)Identification of local Thai craft spirits
PCA + k-MeansAccuracy: 72.23% (three distinct local Thai spirits)
[173]Sensor array: MQ-3, MQ-4, MQ-7, MQ-8, MQ-135, MQ-136, MQ-137, MQ-138, MG811 [174], AM2320 [175]; Signal processing unit: microcontroller with an onboard ADCANNsCorrelation coefficient: 0.97 (to predict seventeen volatile aromatic compounds)Beer quality assessment
Table 6. Summary of gas sensors used in e-nose systems and their key features (response time—the time a sensor takes to reach a certain percentage of its final output signal after exposure to a target gas; resume time—the time the sensor takes to return to its baseline signal after the removal of the target gas).
Table 6. Summary of gas sensors used in e-nose systems and their key features (response time—the time a sensor takes to reach a certain percentage of its final output signal after exposure to a target gas; resume time—the time the sensor takes to return to its baseline signal after the removal of the target gas).
#SensorTarget GasDetection Range [ppm]Response and Resume Time [s]Heater Consumption [mW]Preheat Time [min/h/day]
1BME688IAQ, bVOC, eCO2 bVOC: (5 ppm Ethane, 10 ppm Isoprene/2-methyl-1,3 Butadiene, 10 ppm Ethanol, 50 ppm Acetone, 15 ppm Carbon monoxide)0–500 (IAQ), bVOC, CO2
P: 300–100 hPa,
H: 0–100%
T: −40–85 °C
1/3/3000.16–21.630 min
2MQ-2Flammable gas, smoke300–10,000 ppm (Flammable gas)6095048 h
3MQ-3Alcohol, Benzine0.05–10 mg/L Alcohol-75024 h
4MQ-4Methane300–10,000 ppm (CH4)6095048 h
5MQ-5Liquefied petroleum gas, Methane300–10,000 ppm (CH4, C3H8)6095048 h
6MQ-6Liquefied petroleum gas300–10,000 ppm (Propane)6095048 h
7MQ-7Carbon monoxide20–2000 ppm (CO)6035048 h
8MQ-8Hydrogen gas100–1000 ppm (H2 gas)6095048 h
9MQ-9Carbon monoxide and Combustible gas (Methane and Liquefied petroleum gas)10–1000 ppm (CO)
100–10,000 ppm
(Combustible gas)
6035048 h
10MQ-131Ozone10–1000 ppm (O3)11095048 h
11MQ-135Ammonia gas, Sulfide, Benzene series steam10–1000 ppm (Ammonia gas, Toluene, Hydrogen, smoke)6095048 h
12MQ-136Hydrogen sulfide gas1–200 ppm (H2S gas)6095048 h
13MQ-137Ammonia gas5–500 ppm (NH3 gas)6090048 h
14MQ-138Toluene, acetone, alcohol, hydrogen5–500 ppm6090048 h
15MQ-139Freon10–1000 ppm180–30090048 h
16ZE03-NH3CO, O2, NH3, H2S, NO2, O3, SO2, CL2, HF0–1000 ppm (CO), 0–25% vol (O2), 0–100 ppm (NH3),
0–100 ppm (H2S),
0–20 ppm (NO2),
0–10 ppm (HF),
0–20 ppm (SO2),
0–10 ppm (CL2),
0–20 ppm (O3)
15–15020-
17ZE03-C2H4CO, O2, NH3, H2S, NO2, O3, SO2, CL2, HF, H2, PH3, HCL, C2H4-15–15020-
18MICS-6814Carbon monoxide, Nitrogen dioxide, Ethanol, Hydrogen, Ammonia, Methane, Propane, Iso-butane1–1000 ppm (CO),
0.05–10 ppm (NO2),
10–500 ppm (C2H5OH),
1–1000 ppm (H2),
1–500 ppm (NH3),
CH4 > 1000 ppm,
C3H8 > 1000 ppm,
C4H10 > 1000 ppm
-43–76-
19TGS2201Diesel exhaust, Gasoline exhaust0.1–10 ppm (NO, NO2) 10–1000 ppm (CO, H2, HC)-5057 d
20TGS2444Ammonia gas, Hydrogen sulfide gas, Ethanol10–300 ppm of NH3, 10–100 ppm of H2S, 300–1000 ppm of Ethanol60–1805648 h
21TGS2600Hydrogen, Ethanol1–30 ppm of H2-2107 d
22TGS2602VOCs, Ammonia, Hydrogen sulfide gas1–30 ppm of EtOH-2807 d
23TGS2603Trimethylamine, Methyl mercaptan1–30 ppm of EtOH-24096 h
24TGS2610Butane, Liquefied petroleum gas1–25 % LEL-2807 d
25TGS2611Methane, Natural gas500–10,000 ppm-2807 d
26TGS2612Methane, Propane, Iso-butane1–25 % LEL of each gas-2807 d
27TGS2620Alcohol, Organic solvent vapors50–5000 ppm EtOH-2107 d
28TGS800General air contaminants1–30 ppm-660-
29TGS813Combustible gases500–10,000 ppm of Methane-835-
30TGS821Hydrogen30–1000 ppm of H2-660-
31TGS822/823Alcohol, Organic solvents50–5000 ppm of Ethanol-660-
32TGS825Hydrogen sulfide gas5–100 ppm of (H2)-660-
33TGS826Ammonia gas30–300 ppm of NH3-835-
34TGS832R-134a100–3000 ppm of R-134a-835-
35TGS842Methane natural gas500–10,000 ppm of CH4-835-
36TGS880Fumes from food (alcohol, odor)10–1000 ppm (Air and Ethanol)-835-
37WSP1110 ObsoleteNO2 sensor0.1–10 ppm NO2---
38WSP2110Toluene, Methanal, Benzene, Alcohol, Acetone1–50 ppm NO270300120 h
39MP503Alcohol, Smoke, Iso-butane, Methanal10–1000 ppm (Alcohol)6030048 h
40MG811Carbon dioxide350–10,000 ppm (CO2)201200-
Note: No documentation was found for the MQ-1, HGS1000, HGS1001, HGS1002, and HGS1007 sensors; therefore, they are not included in this table.
Table 7. Summary of commercial electronic nose systems for food quality and safety.
Table 7. Summary of commercial electronic nose systems for food quality and safety.
Commercial E-NosePaperData Analysis TechniquesEvaluation and Performance MetricsApplication Area
PEN3: A sensor array of ten different metal oxides single thick film sensors [177][183]Recurrent Criss-Cross Attention Network [184]Accuracy: 98%Peanuts quality assessment
[185]Statistical analysis on data collected by PEN3 (weight loss measurements and firmness analysis also performed)Prove that ilmenite-grafted graphene oxide coating reduces postharvest lossesPostharvest preservation of fruits (bananas)
[186]PCA (e-nose and Headspace-Gas Chromatography-Ion Mobility Spectrometry)Accuracy: 100% (genuine/fake)Amomi fructus authenticity
Partial Least Squares-Discriminant Analysis (PLS-DA) [187] (e-nose and Headspace-Gas Chromatography-Ion Mobility Spectrometry)Accuracy: 97.96% (origin identification)Amomi fructus origin identification
[188]PCA + ANNsAccuracy: 99%Milk safety assessment
[189]Solid-Phase Microextraction [190] coupled with GC–MS and e-nose analysisNot discussedDevelopment of structured lipids with enhanced flavor profiles for dairy products and functional food
[191]Dung Beetle Optimizer algorithm [192] combined with 10 different machine learning methodsCoefficient of determination > 0.895Prediction of the electronic sensory characteristics of fermented milk
[193]Proposed data augmentation model (e-nose + e-tongue) + CNNAccuracy: 95.34% (four types of mixed solution); Accuracy: 97.78% (five brands of beer); Accuracy: 97.37% (five kinds of apple)Quality of different food
Fox 3000: Two sensor chambers equipped with twelve MOS sensors [178][178]Random ForestAccuracy: 95.30%Mandarin orange quality assessment
Fox 4000: An injection system, sensor chambers with eighteen MOS sensors, a mass flow controller, and a micro-controller acquisition board [179][194]PCADiscrimination index: 93 (seven batches of hydrolysate)Quality of baked goods (effects of enzymatic hydrolysis on soy protein concentrate)
[195]PCAPCA1: 94.54%, PCA2: 3.38% of the total variance (untreated sample/pasteurized sample/treated sample/sterilized sample in 0, 30 and 60 days of storage)Shelf life of chicken products quality assessment
[196]PCADiscrimination index: 90 (eight types of plum jam samples)Evaluation of the characteristics of sugar-free plum jams
[197]PCA + CA + Partial Least Squares regression [198] (GC–MS and e-nose data)Correlation coefficients > 0.98 (for 14 characteristic aroma-active compounds)Mitten crab quality assessment
FOODsniffer [180][199]E-nose data analysis compared with microbiological and GC–MS analysesFOODsniffer can anticipate the unacceptability conditions of salmon (at 22 °C, 10% of samples are `Not satisfactory’ when FOODSniffer is `Green’)Salmon fillet and burger quality and safety assessment
[200]E-nose data analysis (PCA) compared with physicochemical measurements of meat qualityPC1–71.13%, PC2: 83.70% of total varianceMeat quality assessment
NeOse Pro: A gold-layer-based optoelectronic sensor array featuring sixty-three non-specific peptides [181][201]PCA + Gas Chromatography with Ion Mobility Spectrometry (GC–IMS)Completely separate one samplePlant-based beverage quality assessment
PCA + e-noseCompletely separate seven samples
LDA + GC–IMSAccuracy between 15.4% and 100%
LDA + e-noseAccuracy between 96.2% and 100%
Cyranose@ 320: An array of thirty-two nanocomposite sensors [182][202]Proposed e-nose pattern recognition algorithmAccuracy: 80% at room temperatureIdentification of Terfezia arenaria truffle
Table 8. Summary of electronic noses developed by academic research groups for food quality and safety.
Table 8. Summary of electronic noses developed by academic research groups for food quality and safety.
E-NosePaperData Analysis TechniquesEvaluation and Performance MetricsApplication Area
LibraNose: An array of eight QCM non-selective sensors coated with different polypyrrole derivatives—University of Rome Tor Vergata, Italy [203][209]E-nose data analysis (PCA + Random Forest regression)Accuracy: 92.8% (for predictions of B. thermosphacta)Meat quality assessment
High Performance Liquid Chromatography + Random Forest regressionAccuracy: 100% (for predictions of Lactobacilli)
GC–MS + Random Forest regressionAccuracy: 93.9% (for predictions of Enterobacteriaceae)
GC–MS + kNN-RAccuracy: 96.0% (for predictions of Pseudomonads)
[210]PCA + Proposed data model based on Adaptive Fuzzy Logic SystemAccuracy: 94.28% (fresh/semi-fresh/spoiled)Monitoring of meat spoiling during storage
[211]PCA + Proposed Multi-Input Multi-Output Clustering-based Fuzzy Wavelet Neural Network modelAccuracy: 95.71% (fresh/semi-fresh/spoiled); RMSE: 0.2969 to predict the microbial load on meat surfaceMeat quality assessment
[212]Proposed model based on ensemble-based (Bagging and Boosting) SVMAccuracy: 84.1% (fresh/semi-fresh/spoiled)Meat quality assessment
Proposed model based on ensemble-based (Bagging and Boosting) SVM-regressionAccuracy: 85% to predict bacterial species counts
E-nose with an array of twelve QCM sensors—University of Rome Tor Vergata, Italy [204][204]PLS-DA94% of the original data’s variation can be represented in a reduced-dimensional space; Accuracy: 100% (five different classes of sparkling wines)Identification of rosé sparkling wines
[213]PLS-DA85% of the original data’s variation can be represented in a reduced-dimensional space; Accuracy between 60% and 100% for the first stages of Botrytis cinerea infection (1, 2, 3 days)Identification of noble rot (a fungus also known as Botrytis cinerea) in postharvest wine grapes
E-nose with an array of eight QCM sensors—University of Rome Tor Vergata, Italy [205][205]PCA + LDAAccuracy: 71.4% (Aspergillus niger/
Aspergillus fumigatus/
Aspergillus flavus)
Identification of Aspergillus Species
E-nose with four gas sensors (BME680 [214], SGP30 [215], CCS811 [216], iAQ-Core [217]) —Industrial Engineering School of the University of Extremadura, Spain [206][206]E-nose data analysis (PCA + PLS-DA) compared with GC–MSPC1–83.5%, PC2–12.3% of the total variance; Accuracy: 100% (six classes of roasted coffee beans exposed to different heat treatment conditions)Roasted coffee quality assessment
[218]PCA + PLS-DAWhole roasted almonds–R2: 0.89 for acrylamide and furfural, R2: 0.83 for hydroxymethylfurfural Ground roasted almonds–R2: 0.99 for acrylamide, R2: 0.98 for hydroxymethylfurfural, R2: 0.88 for furfuralPrediction of contaminants in roasted almonds
Agrinose: An array of eight MOS sensors (AS-MLV-P2 [219], TGS2602, TGS2600, TGS2603, TGS2610, TGS2611, TGS8100 [220], TGS2620) —Institute of Agrophysics IA PAS, Poland [207][221]Proposed model based on a three-parameter method based on the impregnation time, cleaning time, and maximum response of chemically sensing sensors + PCA compared with GC–MSPC1 + PC2 describe 72.64% of the total variance and enable clear separation of different sample classesAssessment of the suitability of bread for consumption after storage
[222]Proposed model based on a three-parameter method based on the impregnation time, cleaning time, and maximum response of chemically sensing sensors + PCA compared with Fourier Transform Infrared Spectroscopy [223] and GC–MSPC1 + PC2 describe 79.25% of the total variance and enable clear separation of different sample classesIdentification of rapeseed spoilage
E-nose with nine MOS sensors (MQ-2, MQ-3, MQ-4, MQ-5, MQ-6, MQ-7, MQ-8, MQ-9, MQ-135)—Department of Biosystems Engineering, Bu-Ali Sina University, Iran [208][224]PCAIncluded 55%, 75%, 47%, and 53% of data for unprocessed whole/dried slices/powder/tabletGarlic quality assessment
LDAAccuracy: 90%, 93.33%, 88.89%, 60% (unprocessed whole/dried slices/powder/tablet)
SVMAccuracy: 72.22%, 80.00%, 75.55%, 40% (unprocessed whole/dried slices/powder/tablet)
Backpropagation Neural Network [225]Accuracy: 100%, 97.80%, 92.2%, 77.78% (unprocessed whole/dried slices/powder/tablet)
Table 9. Summary of gas sensors used in e-nose systems and their key features.
Table 9. Summary of gas sensors used in e-nose systems and their key features.
#SensorTarget GasDetection Range [ppm]Response and Resume Time [s]Heater Consumption [mW]Preheat Time [min/h/Day]
1BME688IAQ, bVOC, eCO2 bVOC: (5 ppm Ethane, 10 ppm Isoprene/2-methyl-1,3 Butadiene, 10 ppm Ethanol, 50 ppm Acetone, 15 ppm Carbon monoxide)0–500 (IAQ), bVOC, CO2 P: 300–1100 hPa, H: 0–100% T: −40–85°C1/3/3000.16–21.630 min
2SGP30 End of lifeVOC, eCO2, Ethanol, Hydrogen sulfide gas0–60,000 ppb (VOC), 400–60,000 ppm (eCO2) 0– 1000 ppm (Ethanol, H2S)186.424 h
3CCS811TVOC, eCO20–1187 ppb (TVOC), 400–8192 ppm (eCO2)0.25/1/10/601.2–4648 h
4iAQ-Core ObsoleteeCO2, TVOC450–2000 ppm (eCO2), 125–600 ppb (TVOC )1/119–69-
5AS-MLV-P2 ObsoleteCO, Butane, Methane, Ethanol, Hydrogen30–500 ppm (CO), 15–150 ppm (Butane), 250–4500 ppm (CH4), 10–200 ppm (Ethanol), 25–500 ppm (H)1/10505 d
6TGS8100Methane, Iso-butane, CO, Hydrogen, Ethanol1–100 ppm, 1–30 ppm (H2)-15>1 h
Table 10. Summary of new gas sensors for the food industry.
Table 10. Summary of new gas sensors for the food industry.
Reference and Sensor Type/TechnologyTarget Compounds/
Application
Data Analysis TechniquesKey Features/Results
[226]: Silicon NanoWires + multi-walled carbon nanotubeEssential oils, alcoholic beverages, general foodPCAFast response time (20–30 s), high selectivity, dual surface, and chemical modification
[227]: Graphene junctionsAflatoxin B1N/A1.2 V bias yields >3 µA current change; suitable for rapid e-nose integration
[228]: Plasmonic arrays + chemometrics + machine learningMultiple VOCs in foodPCA + LDAUses Surface-Enhanced Raman Spectroscopy, mimics animal olfaction; machine learning enables multi-analyte detection
[229]: Memristor-based in-memory computing + MOS sensor arrayVarious gases (15 sensors)CNN94% classification accuracy; 20.2 mW power; fast response time (<0.4 ms inference time); compact processing scheme
[230]: Graphene + Metal PhthalocyaninesAmmonia gas, interfering gasesPCA5-sensor array (Co-Pc, Ni-Pc, Zn-Pc, Fe-Pc, pristine); promising for food quality monitoring
[231]: Film Bulk Acoustic Resonator sensorsGeneral gases; example: banana freshnessReal-time signal processing and pre-processing + Discriminative analysisMiniaturized portable e-nose; 6–8× more sensitive than polymer-coated Film Bulk Acoustic Resonator; drift-compensated
[232]: Colorimetric Fe(II) complexAmmonia gasPCA + Hierarchical cluster analysis, SVMDetects 105 ppb at room temp; reusable; no external energy needed
[233]: CNT + olfactory receptor (ODR-10)Diacetyl in alcoholic beveragesSensitivity and selectivity analysisDetection limit of 10 fM; better than fluorescence assays and GC–IMS in classification
Table 11. Summary of datasets used in e-nose research for food quality and safety.
Table 11. Summary of datasets used in e-nose research for food quality and safety.
DatasetDescriptionPaperData Analysis TechniquesEvaluation and Performance Metrics
[234,235]2220 sensor signal responses collected from twelve cuts of beef samples in four different degrees of freshness using eleven gas sensors[236]Proposed model based on 1D-CNNAccuracy: 97.22% (excellent/good/ acceptable/spoiled)
[237]ANNsAccuracy: 99.9% (excellent/good/ acceptable/spoiled)
Linear RegressionAccuracy: 98.9% (excellent/good/ acceptable/spoiled))
KNNAccuracy: 98.8% (excellent/good/ acceptable/spoiled)
[238]Proposed MLP model on Field Programmable Gate ArrayAccuracy: 92.72% (excellent/good/ acceptable/spoiled)
[239]Proposed approach based on Single Plurality Voting System model + Decision TreeAccuracy: 91.13% (excellent/good/ acceptable/spoiled)
Proposed approach based on Single Plurality Voting System model + KNNAccuracy: 88.69% (excellent/good/ acceptable/spoiled)
Proposed approach based on Single Plurality Voting System model + LDAAccuracy: 80.73% (excellent/good/ acceptable/spoiled)
[240]420 samples for seven different mixtures of beef and pork collected from eight gas sensors[241]Proposed model based on a conventional Deep Extreme Learning Machine with an autoencoder for feature learningAccuracy: 99.85% (seven combination mixtures of meat)
Proposed model based on SVM with a Radial Basis Function kernelAccuracy: 93.48% (seven combination mixtures of meat)
Proposed model based on a conventional Deep Extreme Learning Machine with PCA for feature learningAccuracy: 99.97% (seven combination mixtures of meat)
Proposed model based on PCA + SVM with a Radial Basis Function kernelAccuracy: 96.88% (seven combination mixtures of meat)
[242,243]Time series data for 235 wine samples collected from six gas sensors[244]Proposed model based on CNNAccuracy: 99.2% (low quality/average quality/high quality)
[245]48,846 rows for rice quality acquired from nine gas sensors and four other sensors for related data[245]Proposed model based on KNNR2: 0.7217; RMSE: 3.8043
[246]Gradient Tree BoostingAccuracy: 96% (expired/non-expired)
[247]Complement Naïve Bayes classifierAccuracy: 98% (expired/non-expired)
Multinomial Naïve Bayes classifierAccuracy: 97% (expired/non-expired)
Gaussian Naïve Bayes classifierAccuracy: 82% (expired/non-expired)
Bernoulli Naïve Bayes classifierAccuracy: 52% (expired/non-expired)
[248]MLPAccuracy: 99.84% (expired/non-expired)
Table 12. Summary of gaps and corresponding solutions in this work.
Table 12. Summary of gaps and corresponding solutions in this work.
Identified GapRecommended Solution in This Review
Low real-world deployment despite high lab accuracyProvide case studies and benchmarking tables to bridge lab-to-field gaps
Sensor selectivity and sensitivity challengesDevelop gas-sensitive materials with enhanced selectivity and sensitivity for food VOCs, or design bio-inspired or biomimetic sensors that mimic natural senses to improve detection accuracy in food analysis
Sensor response time challengesDesign gas sensors with fast response times
Sensor drift and calibration challengesIntroduce adaptive/recalibrating machine learning models and emphasize real-time feedback control
Black-box nature of machine learning models used in classificationRecommend interpretable machine learning models and alignment with food safety regulations (i.e., Codex/ISO)
Lack of efficient multi-sensor (e-nose, e-tongue, e-eye) data fusion algorithms results in incomplete food profilingDevelop advanced data fusion frameworks using machine learning, hybrid fusion techniques, and synchronized pre-processing
Miniaturization and portability challengesRecommend integration of micro-electro-mechanical systems/
nano-electro-mechanical systems technologies for compact, low-power devices
Lack of standardization in methodology and validationRecommend universal protocols for data collection, validation, and sensor benchmarking
Lack support in user-friendly visualization of the odor classification and identification resultsIntroduce real-time data analysis platforms for on-site decision making
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Sanislav, T.; Mois, G.D.; Zeadally, S.; Folea, S.; Radoni, T.C.; Al-Suhaimi, E.A. A Comprehensive Review on Sensor-Based Electronic Nose for Food Quality and Safety. Sensors 2025, 25, 4437. https://doi.org/10.3390/s25144437

AMA Style

Sanislav T, Mois GD, Zeadally S, Folea S, Radoni TC, Al-Suhaimi EA. A Comprehensive Review on Sensor-Based Electronic Nose for Food Quality and Safety. Sensors. 2025; 25(14):4437. https://doi.org/10.3390/s25144437

Chicago/Turabian Style

Sanislav, Teodora, George D. Mois, Sherali Zeadally, Silviu Folea, Tudor C. Radoni, and Ebtesam A. Al-Suhaimi. 2025. "A Comprehensive Review on Sensor-Based Electronic Nose for Food Quality and Safety" Sensors 25, no. 14: 4437. https://doi.org/10.3390/s25144437

APA Style

Sanislav, T., Mois, G. D., Zeadally, S., Folea, S., Radoni, T. C., & Al-Suhaimi, E. A. (2025). A Comprehensive Review on Sensor-Based Electronic Nose for Food Quality and Safety. Sensors, 25(14), 4437. https://doi.org/10.3390/s25144437

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop