Internet of Things‐Enabled Food and Plant Sensors to Empower Sustainability

To promote sustainability, this review explores: 1) the utilization of affordable high‐performance sensors that can enhance food safety and quality, plant growth, and disease management and 2) the Internet of Things (IoT)‐supported sensors to empower farmers, stakeholders, and agro‐food industries via rapid testing and predictive analysis based on sensing generated informatics using artificial intelligence (AI). The performance of such sensors is noticeable, but this technology is still not suitable to be used in real applications owing to the lack of focus, scalability, well‐coordinated research, and regulations. To cover this gap, this review carefully and critically analyzes state‐of‐the‐art sensing technologies developed for food quality assurance (i.e., contaminants, toxins, and packaging testing) and plant growth monitoring (i.e., phenotyping, stresses, volatile organic components, nutrient levels, hormones, and pathogens) along with the possible challenges. The following has been proposed for future research: 1) promoting the optimized combination of green sensing units supported by IoT to perform testing at the location, considering the remote and urban areas as a key focus, and 2) analyzing generated informatics via AI should also be a focus for risk assessment understanding and optimizing necessary safety majors. Finally, the perspectives of IoT‐enabled sensors, along with their real‐world limitations, are discussed.


Introduction
Emerging technologies have played important roles in empowering sustainable agriculture. [1]Wearable microfabricated sensors combined with the Internet of Things (IoT) and artificial intelligence (AI) have shown emerging solutions for next-generation precision agriculture farming, and support climate-smart agriculture (CSA). [2,3]With the recently advanced manufacturing facilities, the sensor arrays in foods or plants can provide continuous large and on-site data to make a decision, prediction, and treatment. [4]otably, multifunctional wearable sensors have been researched and implemented for personalized human medicine [5][6][7][8][9][10][11] ; however, they have not been properly explored for agriculture and food sciences, except for a few applications.To mitigate both the direct and indirect negative effects on the food value chain (FVC) and promote sustainable agriculture, there is a growing need for advanced sensing technologies with enhanced capabilities that surpass the limitations of conventional detection tools.The sophistication of wearable sensors through advanced manufacturing (i.e., 3D printing and nanomanufacturing) and integration with AI and IoT devices can provide a novel sustainable solution for food science and agriculture.However, the utilization of these advanced manufacturing technologies and IoT in food sensors is currently in its early stages.To establish a common goal for solving problems, agronomists must collaborate with engineering, medicine, materials, and veterinarians.
Food sensors technology has seen remarkable development with a focus on measuring toxins, humidity, pH, freshness, temperature, contaminants, and pathogens.These sensors play a crucial role in ensuring food safety, maintaining food quality, and monitoring packaging standards. [12]Early detection of pathogens, for instance, plays a vital role in plant health monitoring.By integrating these intelligent sensors with IoT devices utilizing communication technologies such as Wireless Sensor Networks (WSN), including Wi-Fi, Bluetooth, Zigbee, and LoRA, the extensive data generated is a valuable resource for decision-makers.These data will assist in effectively managing food safety and quality, thereby protecting public health. [13]Traditional food sensing/analysis technologies, such as gas chromatography, enzyme-linked immunosorbent assay (ELISA), hyperspectral imaging, and colorimetric and spectroscopic techniques, encounter certain obstacles.These methods often rely on noninvasive procedures and can be time-consuming, leading to delays in conveying critical data about the health of food products. [14]urthermore, these techniques require expert knowledge for an accurate interpretation.Therefore, integrating advanced noninvasive deployable WSN sensors with IoT devices is crucial, as it allows for the direct collection of on-site and large-scale data across various sectors, including harvesting, packaging, and transportation management.This integration enhances decision-making.Additionally, this approach establishes an integrated network that ensures that product quality is maintained from the farm to shelf level.
Precise, timely, and on-site monitoring of plant health, physiology, and disease can significantly enhance crop production and yield, resulting in substantial economic benefits. [15]However, early monitoring of plant health using traditional laboratorybased analysis is not feasible, thus demonstrating the demand for wearable and wireless sensor networks to promote precision farming. [16,17]For instance, an electronic wearable sensor has been mounted on the leaves of plants to detect viral and fungal infections or stresses such as salinity or drought. [18]This sensor chip was tested with tomato plants infected with three different pathogens, tomato spotted wilt virus, early blight, and late blight, by exposing them to a variety of abiotic stresses. [18]Moreover, it is worth noting that wearable electronics equipped with 3D nanoprinting technology [19] for plant health monitoring are currently emerging in the field.This is mainly because of two inherent features: customizability and miniaturization. [19]The customizability of nanoprinted devices promotes the wearability of the sensor interface with plant tissues, wherein the elastic modulus of the printed structures is very close to that of plant tissues.Notably, the Si-based electronic devices have a large mismatch of elastic modulus (%five orders of magnitude) with the plant tissues which creates a potential barrier of sensor coupling with the plants.Furthermore, the wireless data conveyed from the sensor network and combined with IoT devices can empower precision and sustainable farming and build possible prediction models.This would assist farmers in the early detection of plant diseases and improve crop yields, resulting in economic benefits and food security. [20]igure 1 shows the timeline history of sensors and IoT devices applied in agriculture.In the middle of the 1300, litmus paper was invented and acted as a pH indicator that could be detected by the naked eye. [21]However, the cornerstone of modern sensing techniques was established in the previous century when Max Cremer realized the pH response of glass electrodes. [22]ince then, fundamental changes and developments in sensing science and technology have gained tremendous speed and evolution in various sciences, including advanced microfabrication, and have been integrated into this field, thus facilitating accurate and multiplex measurements.However, advancements in computer and data sciences, particularly IoT breakthroughs in the late 1990s, have opened an avenue for incorporating these concepts to modernize sensing technologies.In 2011, Hu et al. built a monitoring platform based on the IoT using video surveillance technology, sensor networks, and Global Positioning System (GPS). [21]The implementation of AI and IoT devices with sensors for the food and agriculture industries can bring innovative solutions to existing challenges.
This review carefully and critically explores the key opportunities and challenges toward the current state-of-the-art IoT-based wearable and wireless sensors for food and plant sensing applications.It covers two major areas of agriculture: a) food, and b) plant sensing, with their critical challenges and opportunities.On-site and noninvasive IoT wearable sensors are excellent candidates for bridging the gap between benchtop laboratory measurements and in-field testing for time-sensitive Figure 1.[156][157][158] decisions regarding smart farming.Furthermore, the sophistication of devices depends on sensor manufacturing modalities, the advancement of low-powered or self-powered wireless sensor networks, and data communications.We demonstrated various scenarios concerning the future integration of food and plant sensors with IoT and AI systems.

IoT-Assisted Food Sensors
The market size of food sensing will be %$21.2 billion in 2022 and will have a growth rate of 9.7% in 2023-2028 across the globe. [23]Thus, rapid and on-site food sensing has a potential demand for food safety and quality to maintain public health. [24]illions of people suffer from foodborne illnesses, and may even die.Food contamination with disease-causing bacteria/viruses, toxins, chemical contaminants, and microbes can cause a variety of disorders such as diarrhea and cancer. [25]In addition, spread of foodborne pathogens has led to a multitude of issues regarding food safety; thus, to ensure that food products are free from harmful contaminants and pathogens, rapid and accurate testing using point-of-use detection systems is required. [26]Several point-of-use detection systems with various sensing modalities are being explored for the on-site screening of these contaminants.[34][35] Further, on-site analysis of food pathogens can prevent tedious culturing and sample processing times, making them suitable for real applications. [32]Each modality has its own issues regarding integration with IoT devices for onsite monitoring.In this context, electrochemical biosensors have shown excellent improvements in the on-site monitoring of food pathogens, toxins, and other contaminants.To avoid the use of traditional electrochemical computer workstations, many portable electrochemical sensors integrated with Bluetooth and smartphones have been developed to analyze foodborne pathogens. [36]An IoT-assisted electrochemical sensor was used for the on-site detection of heavy metals, such as Cd 2þ , Pb 2þ , and Hg 2þ , in cow milk, orange juice, and apple juice.This stripping voltammetry-based sensor provides a sensitive response to ≥2 μg L À1 of a target ion and shows a good accuracy of device functionality within 72 h. [37]ood sensors that provide indirect indications of quality by screening storage parameters such as time, temperature, and humidity are best positioned for monitoring food quality with complex and heterogeneous constitutions.Time-temperature indicators (TTI) are useful for indicating exposure to excessive temperatures. [38]This sensor records the food temperature history when mounted on food packaging rather than on the food body.These can be applied to predict whether food is likely to spoil without direct contact with it. [39]Choi et al. developed opaque/transparent TTIs based on nanofibers, which irreversibly changed their transparency upon exposure to room temperature. [40]ditionally, cutting-edge advancements are transforming time-temperature history (TTH)-based near-field communication NFC food-monitoring systems, making them accessible through smartphones. [41]By leveraging NFC technology a new and convenient method combined with TTH monitoring for tracking and monitoring food temperature history has been developed.To capture and store real-time temperature data, NFC tags and sensors have been integrated into food packaging.When a smartphone with NFC capabilities interacts with the packaging, TTH information is instantly retrieved and displayed.This technology provides valuable insights into the complete journey of food products, including storage, transportation, and handling.Users can quickly determine whether the food is exposed to unfavorable temperatures or has compromised quality.Smartphone applications utilize these data to offer real-time alerts and guidance for optimal food handling practices.The integration of TTH monitoring with NFC and smartphone technology presents a user-friendly operational solution to ensure food safety, minimize food waste, and empower consumers with reliable information on food quality and freshness.
One of the simplest indicators of the health of food products is humidity, which guarantees the survival of foodborne pathogens on the food product.Moisture-dependent materials such as organic polymers generally respond to moisture by altering some of their measurable physical properties.This sensing strategy facilitates the inclusion of multiplex sensors, which helps integrate and miniaturize several sensing platforms into one. [42]or example, Yuan et al. developed a multiplex detection system through which moisture, ammonia, and hydrogen sulfide could be measured simultaneously. [43]A printed silicon-based fieldeffect transistor (FET) can accurately and precisely measure the concentration of these gases as an indicator of pork quality. [43]wing to its multiplex detection capacity, this device can significantly reduce monitoring costs.
Additionally, efforts are required to produce sufficient food, and manage food quality.This is because of world population growth, which is expected to exceed 10 billion by 2050, and will require 50% more food to maintain the global food supply chain.Thus, the quality of food at every step of the food cycle is emerging as a key concern owing to unhealthy foods. [44]To achieve this, it is important to understand each food cycle, including crop cultivation, harvesting, packaging, transportation, consumer adoption, and utilization.Figure 2 depicts a typical food cycle from cultivation to consumption, with challenges.These challenges can be overcome through the implementation of biological, chemical, and physical sensors.Food contamination not only depends on food degradation due to food age and environmental factors, such as humidity and temperature (toxins and pathogens), but also on 1) scenarios of the utilization of fertilizers, pesticides, and irrigation systems (toxic materials); 2) packaging materials, preservatives, and transportation (toxic materials, toxins, and pathogens); and 3) storage conditions at the industrial and consumer levels (toxic materials, toxins, and pathogens). [45]o address these potential challenges, the demand for IoT-based biosensors with rapid, accurate, exhaustive, and low-cost monitoring for the best practices in smart farming is growing. [46]ifferent types of food contaminants, toxins, and foodborne pathogens are listed in Figure 2. It should be noted that only a few commercial biosensors are available to detect these markers, owing to their complex food matrices. [47]To date, several solutions have been suggested, such as incorporating food sensors into food packaging to ensure contact with the food's chemical environment, or sampling the food body.Nonetheless, the major concern is the contamination of the food product with the sensor-active material or inconsistent sampling, specifically for heterogeneous food products. [48]Subsequently, various categories of food sensors and emerging technologies in this field are discussed in detail.

Pesticide Sensing
Pesticide sensors play a crucial role in the food industry by enabling the rapid and accurate detection of pesticide residues in agricultural products.These sensors utilize advanced technologies to detect and quantify the presence of various pesticides, thereby ensuring food safety and compliance with regulatory standards. [49]With increasing global demand for safe and contaminant-free foods, the development of efficient pesticide sensors has become a priority for food manufacturers and regulatory authorities.These sensors not only help monitor and control pesticide usage during cultivation but also aid in quality assurance processes along the food supply chain.By providing real-time data on pesticide levels, these sensors can significantly contribute to enhancing consumer confidence and promoting healthier food environments.National regulatory authorities have indicated a maximum residue limit (MRL) ranging from 10 À3 to more than 1 mg of pesticide per kilogram of food, depending on pesticide toxicity. [50]This low MRL urged researchers to develop smart sensors to be used in the supply chain from farm to shelf, while combining IoT and AI to make time-sensitive decisions. [51]Further, it is noted that the wearable sensors embedded with IoT devices make the screening procedure easier and user-friendly for growers and stakeholders. [52]Zhang et al. introduced flexible sensors combined with smartphones and wireless transmission for the detection of chiral (þ)/(À)methamidophos with an LoD of 0.3 μg L À1 . [53]In this study, the sensor's transducer was acetylcholinesterase-modified graphene, which works under an operating potential of 1 V, and provided a sensitivity of 0.34 and 0.32 μg L À1 for (þ)/(À)methamidophos, respectively. [53]Ongoing research and innovation in this field is expected to further improve the sensitivity and specificity of pesticide sensors, making them indispensable tools for ensuring the safety and integrity of food products.Farmers can precisely target specific locations in fields affected by pathogens and insects by combining the IoT technology with pesticide sensors.This targeted approach results in a decreased reliance on pesticides, ensuring the health and quality of harvested crops while promoting environmentally friendly farming practices.

Foodborne Pathogens
Foodborne diseases are of significant concern because of their life-threatening and widespread effects. [54]The Centers for Disease Control and Prevention (CDC) reported approximately 26 000 cases of infection, 6000 hospitalizations, and 122 deaths caused by foodborne pathogens. [55]These pathogens can be detected in a wide range of foods in the early stages, including fresh products, such as milk, fruits, vegetables, meat, chicken, and water, to prevent disease outbreaks.However, the variety of bacterial strains (Escherichia coli and Salmonella spp), parasites, and viruses makes detection strategies more challenging.Although the most recent detectors are based on antibodies, bacteriophages, aptamers, [56] and electrostatic probes, [57] their integration with the IoT and deployment in the food production sector is still limited.In a recent study, smartphone technology was integrated with a 3D printed microfluidic system for detecting food pathogens (specifically DNA analysis) [58] ; however, this requires sample pretreatment, signal amplification, and processing.Further, the integration of a laser-induced fluorescence detector with a microchip capillary electrophoresis, and then utilizing electrophoresis to separate the reacted and unreacted aptamer exhibited a LoD as 3.4 Â 102 and 3.7 Â 102 CFU mL À1 for Salmonella [59] and E. coli, [60] respectively.Although, commonly used recognition elements, such as aptamers and enzymes, exhibit excellent selectivity and sensitivity in their response, small-molecule probes, including lectins, carbohydrates, and vancomycin, are stable, reusable, and cheap alternatives. [61]In this regard, small-molecule probes can be anchored on a variety of transducers, including graphene or silicone, thus enabling different detection modalities, such as electrochemical [62] and optical methods. [61]In addition to E. coli, Salmonella enterica (S. enterica) is a common cause of foodborne infections due to its multiple pathogenic islands (SPI-1 and SPI-2). [63,64]A microbial culture method (golden standard), PCR, and ELISA are common tools for the detection of S. enterica in food samples. [64]A new molecular assay based on clustered regularly interspaced short palindromic repeats (CRISPR)-Cas12 was developed for the detection of S. enteric in fresh eggs without nucleic acid amplification. [64]Specifically, CRISPR-Cas12a has a unique targeting ability owing to the accessory cleavage activity of the Cas12a protein and can be used as a diagnostic tool.Furthermore, a smartphone-integrated colorimetric bioassay based on CRISPR-Cas12a was developed for the ultrasensitive detection of Salmonella in food samples using a G-rich DNAzyme reporting system. [65]This G-quadruplex-based CRISPR-Cas12a system allows on-site and selective detection of foodborne pathogens, which promotes food safety and public health.

Microfluidic Systems
Microfluidic systems have brought about a significant revolution in the domain of sensors and biosensors by offering precise manipulation of fluids and facilitating the integration of multiple analytical processes on a single-chip platform.These compact devices enable the control of small fluid volumes, typically ranging from microliters to nanoliters, within a network of channels and chambers. [66]The exploitation of microfluidics in sensing technologies has led to improved sensitivity, reduced consumption of samples and reagents, and accelerated analyses.To achieve the fast and efficient detection of targets, the distinctive characteristics of microscale flows, such as laminar flow, rapid mixing, and enhanced mass transport, are harnessed.Moreover, to provide versatile sensing capabilities, microfluidic sensors can be combined with various transduction mechanisms, including optical, electrochemical, and acoustic methods.Microfluidic systems enhance sensing characteristics and reduce the cost per test.Owing to their compact size, portability, multiplexing, and ability to handle complex sample matrices, microfluidic systems have gained popularity across diverse applications, such as point-of-care diagnostics and environmental, food, and agriculture monitoring. [67]Furthermore, such devices can be integrated with sensing platforms to directly extract samples from food products on-site for food safety monitoring. [68]hese miniature chips are utilized in lab-on-a-chip systems containing etched microchannels that exploit the properties of microfluidics for sample pumping, mixing, and purification.Recent food sensors that detect targets such as pathogens, [69] pesticides, [48] and toxins, [70] have been scaled up from lab benches to point-of-use settings by incorporation into microfluidic devices.Furthermore, the integration of advanced micro/nanofabrication and 3D printing as a new class of food sensors is expected to improve the functionality of microfluidic devices by enhancing the design complexity. [71]Recently, Kim et al. developed a silk microneedle food sensor for bacterial detection. [72]This system significantly reduced the risk of food contamination by nanoparticles used in the sensing system.In addition, it exploits the capillary action of microneedles for sampling and transports them to a polydiacetylene (PDA) bioink sensor platform functionalized with E. coli targeting antibodies.detection. [72]This sensor has been used to analyze salmon by direct attachment to the salmon fillet, where the microneedles pierce the fillet packaging and deliver fluids from the sample to the sensor by capillary action.The sensor operability was enhanced by integrating a smartphone device. [73]

Smart Sensors for Food Safety and Quality
Food safety is important in the food industry (food production and supply chains) for improving food management. [74]Food sensors can detect quality markers or targets that are indicative of food safety via specific biological interactions. [74]Measuring pathogens, toxins, allergens, fungi, pesticides, contaminants, and heavy metals can aid food safety management systems. [75]hus, there is high demand for smart biosensors that can trace contaminants, toxins, and harmful substances.To meet this demand, emerging technologies including 3D printing, smartphones, AI and IoT, and biosensing technologies are gradually increasing. [76]Smartphone-enabled sensors interfaced with complex algorithms can be utilized for automatic data processing and to improve the detection accuracy.Furthermore, wearable sensors and smartphones can be connected to the cloud to generate sensing, information, and process data for decision-making.The high speed of data transmission with the help of a five-generation (5G) wireless network can accelerate data analysis obtained from biosensors. [77]Subsequently, several studies have been conducted.
RipeSense is an intelligent sensor, as shown in Figure 3a, that can detect fruit ripeness due to colorimetric changes based on the Figure 3. a) Naked-eye detection of fruit ripeness with a commercial indicator, RipeSense.Adapted with permission. [159]Copyright 2016, Hindawi Publishing Corporation.b) A commercial lateral flow test, i.e., RIDA SMART APP for measuring mycotoxins.Adapted with permission. [79]Copyright 2023, R-Biopharm AG.c) A tuneable TTI indicator constructed from plasmonic metal nanocrystals that exhibit a range of time-and temperaturedependent colors.Adapted with permission. [80]Copyright 2013, American Chemical Society.d) A label-free fluorescent aptamer sensor based on graphene oxide layer that detects aflatoxin B1 in food samples (corn, milk, and rice).Adapted with permission. [81]Copyright 2019, Elsevier.
release of aromas from fruits. [78]Additionally, the RIDASMART APP shown in Figure 3b is an evaluation method for the quantification of mycotoxins that evaluates lateral flow tests (LFDs) utilized for immunochromatographic analysis of mycotoxins. [79]he smartphone application supplies reliable results that can be forwarded via e-mail and exported to a cloud server. [79]Moreover, Figure 3c illustrates the low-cost flexible programmability of the TTI combined with its substantial promise of general applicability to each single-packaged item of a plethora of perishable packed products. [80]Besides, Jia et al. developed a label-free fluorescent aptasensor, based on quaternized tetraphenylethene salt (TPE-Z), graphene oxide (GO), and aflatoxin B1 (AFB1) aptamer, which has been constructed to detect AFB1. [81]5.Food Contaminants Food contaminants are substances that unintentionally enter the food supply and pose potential risks to human health.These contaminants originate from various sources, such as environmental pollutants, agricultural practices, processing techniques, and packaging materials.These include chemical contaminants (e.g., pesticides, heavy metals, and food additives), microbial contaminants (e.g., bacteria, viruses, and parasites), and physical contaminants (e.g., glass, metal fragments, and plastic particles).Exposure to these contaminants can result in many health problems, ranging from immediate foodborne illnesses to long-term chronic conditions (Figure 4b).One of the best methods to ensure food safety and quality is to use sensors and monitoring systems to minimize the presence of contaminants throughout the food chain.This necessitates rigorous testing, proper handling, and adherence to good manufacturing practices (Figure 4a).Once contaminated food enters the life cycle, although, in a slow process, each contaminant can affect cellular function and eventually lead to health problems.The negative effects of food contaminants on biological cells and organs are shown in Figure 4b.The resistance induced by human defense mechanisms makes these health consequences unnoticeable.However, contaminants alter the cellular genome and cause permanent alterations in cellular and tissue functions.Over time, if contaminated food is consumed, health deterioration appears to be out of control, and such individuals require immediate medical attention.The effects of food contaminants on the human organs are shown in Figure 4b.Temporary and nonserious side effects include vomiting, diarrhea, and headache, as shown in Figure 4b.
Overall, as a final product, food reaches consumers via a sequential process involving production, processing, packaging, safe storage, and timely transport.Maintaining pace with global demand while considering the management and overuse of agrochemicals, the associated environmental and health risks, food loss, and timely quality assessment of food have emerged as a serious concern.Similarly, maintaining food safety is not the only factor that must be tracked; knowing the effects of contaminated food on humans, animals, and animals as part of food and the possibility of altering food cycle-based ecosystems should be focused on managing food safety and quality monitoring.
Compared to the traditional sensors, innovative IoT-enabled sensors are used to detect food contaminants with many features, including early and rapid detection and continuous tracking, as shown in Figure 4a, thus significantly improving food safety.Such systems utilize the capabilities of the Internet of Things together, transmit, and analyze data in real time.These features are important for making rapid decisions and predictions.By integrating advanced sensors with IoT platforms, valuable data can be continuously collected at various stages of food production, processing, and storage.These comprehensive data enable the proactive identification of potential contamination risks, allowing timely intervention and preventive actions.Additionally, IoT-based sensing technologies enable remote monitoring and control, thus providing instant alerts and notifications in the event of deviations or abnormalities in food quality and safety parameters.This technology-driven approach enhances the efficiency and accuracy of food contaminant detection, while promoting traceability and transparency in the food industry, ultimately ensuring the delivery of safe and high-quality food to consumers.

IoT-Enabled Plant Sensors
The early detection of plant diseases and physiology has the potential to improve crop yields. [82]With wearable and wireless sensor networks, emerging plant sensor technologies are effective in assisting farmers and stakeholders in making on-the-spot decisions and promoting smart agricultural practices.In particular, IoT-enabled plant sensors with flexible substrates and printed circuits have been interfaced with plant leaves and explored for monitoring plant diseases, environmental pollution, fungal infections, pesticides, and exposure to bacterial toxins. [82]][85] Here, we described several detection modalities of plant sensors and summarized their potential problems for plant sensing.

Plant Growth Monitoring
Laboratory-based analyses for plant sensing do not involve realtime measurements, which do not allow us to make timesensitive predictions and actions owing to delayed results. [86]his challenge adversely affects the productivity and economic gains.For instance, plant diseases and stresses can significantly reduce the postharvest crop yield.To address these concerns, plant sensors can be embedded in different plant organs, including leaves, stems, and roots, to wirelessly gather data on early plant health conditions. [87,88]Further, wearable interfaces of the continuously monitored sensors with plant organs provide superior data accuracy on plant health information to take action and assist to build the best prediction model.For example, bioristors have been utilized for in vivo monitoring of plant physiology, which enables precision farming, crop management, and plant phenotyping. [89]This bioristor (organic electrochemical transistor) has a natural textile fiber functionalized by a conductive polymer and implanted into the stem of a tomato plant to communicate the plant's health status in real time by detecting abiotic stress. [89]igure 5 shows several IoT-enabled plant sensors.[92] Compared with traditional tools, lightweight and leaf-implantable sensors are useful for detecting many plant parameters, including stresses (abiotic/biotic), drought, diseases, and nutrients.When combined with IoT-based devices, these sensors can remotely collect data, which is a key aspect of this promising technology. [92]In Figure 5, we demonstrate a set of IoT-enabled plant sensors for practical applications of plant monitoring to support best management practices for smart crop farming.For example, a plant monitoring system can monitor temperature, humidity, and plant water content together and can provide immediate alarms when necessary. [93]A number of sensing modalities with wireless circuits are shown in Figure 5 to demonstrate plant sensors and their utilities.
For on-site diagnosis of plant diseases, a microneedle-based integrated plant sensor was developed (Figure 5a). [94]This device rapidly extracts RNA from plant leaves using a simple press and retraction process with a microneedle array. [94]Furthermore, after inoculation in asymptomatic tomato plants, a smartphonebased 3D-printed device was integrated with this array for loop-mediated isothermal amplification reactions and on-site detection of tomato-spotted wilt virus as early as 5 days. [94]his device can be used for the on-site and qualitative detection of viral infections in plants.A carbon nanotube ink-based  [94] This device consists of a polymeric microneedle array for DNA extraction and a smartphone-based 3D-printed device for loop-mediated isothermal amplification reaction and detection.Adapted with permission. [94]Copyright 2021, Elsevier.b) A plant drought sensor based on nanoprinted ink to detect water respiration. [95]This sensor has been placed on top of the stomata underneath the plant leaves.Adapted with permission. [95]Copyright 2021, Royal Society of Chemistry.c) Biomimetic implantable in-planta sensor for long-term, continuous, and in situ monitoring of plant nitrate.Adapted with permission. [99]Copyright 2022, American Chemical Society.d) Diurnal in vivo xylem sensor based on an organic electrochemical transistor for continuous detection of glucose and sucrose.Adapted with permission. [100]Copyright 2021, Elsevier.e) Wearable plant sensor for in situ detection of pesticide. [101]This smartphone wearable sensor was interfaced with a wireless sensor node to collect data remotely.Adapted with permission. [101]opyright 2020, Elsevier.
printable plant sensor was demonstrated for the long-term monitoring of plant physiological responses to drought. [95]The sensor was printed on plant stomata (Figure 5b) and could provide early warnings to farmers when their crops are in danger. [95]his stomatal electromechanical pore size sensor allows real-time screening of the latency of a single stomatal opening and closing.Furthermore, this leaf-implantable sensor can provide useful information when dealing with water shortages and changes in environmental temperature. [95]lants require many macro-micronutrients from the soil to maintain their growth, health, and yield. [96]Macronutrients such as nitrogen (N), potassium (K), and phosphorus (P) are essential for various metabolic processes. [16]Plants absorb nutrients (inorganic nitrate, ammonia, and other nutrients) and transport them from the soil to shoot cells via root-specific transporters. [97]But, plants have reduced growth and productivity owing to these deficiencies.Implantable and wearable sensors play a major role in plant nutrient homeostasis. [83,98]esearchers have shown that needle-based plant sensors can be used to monitor plant nitrates to improve agricultural productivity, profitability, and environmental performance (Figure 5c). [99]To maintain a biocompatible sensor interface during the long-term monitoring of plant nitrate dynamics, this sensor uses an artificial enzyme (vitamin B 12 ) combined with a carbon-based nanocomposite (which converts nitrate to nitrite and produces electrons for measurement). [99]This needle-based sensor provides a novel approach for monitoring nitrate dynamics in plants without precalibration or preprocessing, thus potentially impacting plant sciences. [99]In addition, a low-cost, portable, and IoT-enabled wearable plant sensor based on an organic electrochemical transistor (OECT) was developed for the real time and in vivo screening of sugar levels and fluctuations directly in the vascular tissue of trees without any sampling (Figure 5d). [100]This bioelectronic device offers high spatiotemporal resolution for the detection and diurnal measurements of sugar homeostasis.Wearable plant sensors have been used to monitor organophosphorus pesticides on crop surfaces. [101]or instance, Zhao et al. recently introduced the simple geometry of a flexible serpentine three-electrode system which was fabricated via the laser-induced graphene (LIG) technique (Figure 5e). [101]This sensor can selectively capture and detect methyl parathion, and the sensing data are wirelessly transmitted to a smartphone device for in situ analysis of pesticides on the surface of agricultural products. [101]Despite the excellent performance of these plant sensing technologies, they are not sufficiently mature to be used by end users, such as farmers and stakeholders. [83]nother gap is that wearable electronics is a very new concept in agriculture and plant sciences, and its implications are in their infancy.New innovative approaches may motivate the development of wearable sensors with many functionalities to track and predict plant health with their stresses in real time.To monitor plant health instantly or continuously, an ideal wearable sensor can be attached to plant organs (stalks, leaves, and roots).

Plant VOCs Sensing
Figure 6 shows the wearable plant sensors used to detect volatile organic compounds (VOCs), plant phenotyping, stress, and other parameters at different plant locations.Plant wearable sensors are small devices mounted on plant leaves or implanted into plant stalks that are useful for the continuous measurement of these parameters.These sensors provide information regarding plant VOCs, diseases, or stresses, and can also be useful for tracking plant growth and health.For continuous and remote monitoring, these sensors must be connected to both the readout conditioning circuits and wireless sensor nodes.The miniaturization of such sensors with circuits is another aspect that needs to be significantly improved to mount sensors onto plant leaves.Furthermore, long-term monitoring of VOCs is problematic because of the limited power of low-power batteries.We critically summarize the progress and limitations of wearable plant sensors with a set of recent applications.
Plant volatile organic compounds (VOCs) are crucial for studying how genotypes interact with environmental stressors. [102,103]any wearable VOCs are released from plants because of their interaction with biotic (temperature and high light) and abiotic (water, salt, and oxidative) stresses.These VOCs are indicators of inaccurate agrochemicals, nutrient deficiencies, mineral toxicities, a lack of soil moisture, and pathogen infection. [102]lthough, the proton-transfer-reaction mass spectrometry [104] and gas chromatography-mass spectrometry [105] are used to detect plant VOCs, they cannot be used for in-field instant measurements.An inexpensive wearable sensor installed on the surface of leaves can detect VOCs (e.g., methanol) without expensive laboratory analysis. [106]Wearable plant sensors measured methanol emissions directly from the leaves of two different genotypes of corn plants (Mo17 and B73) under field conditions. [103]uring detection, a polymer such as poly(2-amino-1,3,4thiadiazole) used in this sensor performed the electrochemical oxidation of methanol gas molecules released from plant leaves.Furthermore, a machine learning-assisted electronic nose (e-nose) consisting of titanium dioxide (TiO 2 ) nanostructures combined with polymers was used to monitor VOCs. [107]herefore, these sensing technologies could be implemented for detecting other VOCs to know the status of plant growth, health, stress, and physiology.
Figure 6a shows a lightweight, wireless field-effect transistor sensor interfaced with the leaf of a plant to monitor dimethyl methylphosphonate (DMMP) emission. [108]This flexible transistor uses an array of carbon nanotube channels and graphitic electrodes. [108]The sensor output signal shown in Figure 6b is the change in normalized resistance, ΔR/R 0 = (RÀR 0 )/R 0 , where R 0 and R are the resistance values before and during the DMMP exposure, respectively, and is directly proportional to the concentration of DMMP with a resolution of ppm concentration. [108]In this device, wireless measurements by implantation on plant leaves were possible because of the sensor structures imprinted onto a flexible PET substrate and the electronics. [108]

High-Throughput Plant Phenotyping and Stress Sensing
Plant phenotyping is a promising technique that connects genomics, plant physiology, and agronomy.Plant phenotyping systems can be described in terms of resolution, throughput, and dimensionality. [109]With the advancement of recent imaging/sensor technologies (unmanned aerial vehicles, remote sensing via drones, postharvest phenotyping, phenotyping, field scanning, etc.), plant phenotype screening is widely used for decision-making in agricultural farming. [109,110]The highthroughput image-based phenotyping can image nondestructively plant physiology, and growth status and can collect environmental (light intensity, water supply, temperature, humidity), disease tolerance, and physical (plant weight) data to quantify genotype in the presence of ecological interactions.Robot-assisted and satellite-based imaging platforms, highperformance drowning with high-resolution cameras, and other sophisticated tools have been employed for high-throughput plant phenotyping.When combined with these tools, automated sensor systems for accurate determination of plant nutrients, often integrated with high-throughput phenotyping platforms, are critical for plant health.Jiang et al. developed a novel microplatform for high-throughput phenotyping of Arabidopsis plants that can monitor phenotypic changes in different plant organs, including both root and shoot systems. [111]Figure 6c shows a vehicle-based robotic system for high-throughput field phenotyping. [112]Also, the successful application of plant phenotyping to monitor nutrients, plant diseases, and other environmental factors requires nutrient and environmental sensor networks with imaging systems, quality data management and analysis, remote sensing, and IoT devices.On the phenotyping platform, data were captured using different types of sensors, monitoring changes in both plant tissues and environmental variables. [113]hese sensors can be divided into two categories based on their mobility and the location of the target plant.In the first approach, these sensors can be simply mounted on the plant tissue (wearable plant sensors) to collect the required information and transfer the collected data to the processor.In the second method, sensors can be attached to a drone or any other mobile carrier to move over a farm or greenhouse for data collection. [114]ext, the information stored in the data storage is combined and analyzed by the defined algorithms to extract the health status and features of the plates.These features have been  [108] Copyright 2014, American Chemical Society.b) Real-time sensing of different concentrations of dimethyl methylphosphonate (DMMP).Adapted with permission. [108]Copyright 2014, American Chemical Society.c) A vehicle-based robotic system for high throughput field phenotyping.Adapted with permission. [112]Copyright 2019, J. Murman.d) Schematic of hooked miniature machines for on-leaf sensing and delivery.Adapted with permission. [119]Copyright 2021, Springer Nature.e) Schematic of the experimental setup for a nanobionic sensing plant comprising of a nanosensor embedded on the plant leaf to a portable Raspberry Pi-based electronic device.Adapted with permission. [121]Copyright 2016, Springer Nature Ltd. f ) A multimodal flexible sensor attached on the bottom side of a leaf to detect temperature and humidity, and a schematic of the device structure with the functional components.Adapted with permission. [160]Copyright 2020, American Chemical Society.investigated using data science and/or statistical techniques. [115]n contrast, images were captured by drone-mounted sensors and analyzed using image processing procedures, artificial intelligence, deep learning, and machine learning. [116]biotic stressors, including salinity, heat, and drought, as well as biotic stressors, such as insects, nematodes, and pathogens (fungi, bacteria, and viruses), can negatively affect crop plant growth and development, and consequently increase crop losses. [117]Furthermore, continuous global climate change causes more frequent experiences of extreme abiotic stress in plants, which demands more effort to develop advanced sensing technologies to promote climate-smart agriculture. [118]Figure 6d shows a prototype of microneedle array sensors used as plant wearables to monitor plants experiencing drought and rehydration. [119]Furthermore, hyperspectral imaging is a promising platform for detecting abiotic and biotic stresses in plants.A multisensor system combined with canopy data processing was developed for high-throughput field phenotyping in soybean and wheat breeding, including canopy height, temperature, NDVI, reflectance, and RGB images. [120]A nanobionic-sensing plant comprising carbon nanotubes conjugated to a peptide Bombolitin II-based fluorescence sensor was used to recognize nitroaromatics in spinach plants via infrared fluorescence emission (Figure 6e).The sensor was interfaced with IoT devices (smartphones and charge-coupled devices) to collect plant chemical sensing data. [121]A flexible multimodal sensor system was used to detect plant health status and external abiotic stresses (Figure 6f ). [121]This leaf-implantable sensor uses functional ZnIn 2 S 4 (ZIS) nanosheets that can respond to both light illumination and the relative humidity of a plant.This sensor addresses global water resource deficits during farming and monitors the health status of plants. [121]

Plant Pathogens Sensing
Plant diseases caused by pathogens have substantial effects on crop health, quality, and yield in the United States, and lead to significant economic losses.Pathogens such as bacteria, viruses, fungi, phytoplasmas, nematodes, and parasitic plants are threatening to agricultural productivity and sustainability.Early and accurate measurement of these pathogens is crucial for improving plant health and preventing the further spread of infections.The integration of biosensors with IoT devices plays a critical role in screening symptomless plants, detecting early infections, and facilitating swift intervention measures to control and manage diseases.The combination of these technologies empowers farmers with real-time monitoring capabilities, enables data-driven decisions, and promoting precision agriculture practices.By leveraging advanced technologies, the agricultural sector can enhance disease management, minimize economic losses, and foster a more resilient and productive farming environment.Almond plants infected with X. arboricola pv.pruni (bacteria), which cause leaf, stem, and trunk injuries, defoliation, and fruit drop.Citrus plants can be infected by two common viruses and bacteria, Citrus tristeza and Candidatus Liberibacter which cause a decline in plants, yellowing of leaves, yellowing of shoots, leaf spots, decrease in size, and fruit deformity. [122]rect methods for detecting plant pathogens include enzymelinked immunosorbent assays (ELISA), flow cytometry, DNA microarray, polymerase chain reaction (PCR), and fluorescence in situ hybridization, whereas indirect methods include spectroscopic and imaging techniques, plant metabolite profiling, and gas metabolite profiling, such as VOCs.Recently, Thaitrong et al. combined traditional ELISA with a microfluidic device for the detection of Acidovorax citrulli (Ac), watermelon silver mottle virus (WSMoV), and melon yellow spot virus (MYSV) to improve the sensitivity by 12.5-, 2-, and 4-folds, respectively. [123]A miniaturized paper-based gene sensor was developed for the rapid and sensitive identification of a contagious plant virus (banana bunchy top virus). [124]A recent study developed a stand-alone real-time microchip PCR system combining a PCR reaction chamber microchip coupled with a thin-film heater, a fluorescence detector to quantify amplified DNA, a microcontroller controlling thermocycling with a data acquisition system, and a battery.This real-time microchip PCR system combined with DNA extraction protocol was used to detect several fungal and bacterial plant pathogens and showed a low limit of detection (5 ng/8 μL) and the success rate is 100%. [125]urther, VOCs released by plants owing to pathogen infection can be detected indirectly.Cevallos-Cevallos et al. extracted multiple compounds from the leaves, such as hesperidin, naringenin, and quercetin, which can be used as biomarkers to detect huanglongbing (HLB) disease in citrus trees. [126]An electronic nose system was developed to detect VOCs produced by tomato, cucumber, and pepper plants under healthy, infected, and infested conditions.This electronic nose system contains 13 conducting polymeric gas sensors to differentiate the VOCs patterns emitted by plants. [127]Table 1 presents a list of pathogens that cause plant infections.Indirect methods are a promising modality for determining the status of plant diseases caused by pathogens.Direct sensing of plant pathogens for in-field measurements is still not a mature technology that can be used by farmers.

Plant Hormones
Hormones regulate plant functions, such as growth, development, and stress responses.These hormones include jasmonic acid, ethylene, brassinosteroids, abscisic acid, strigolactone, small peptides, salicylic acid, gibberellins, and auxins, and their concentrations are low (nano-to femtomolar) in plants. [128]ormones regulate the cellular processes in target cells and are transported to different locations in plants.Abnormal levels of these hormones in plants negatively affect growth, behavior, adaptation to the environment, and yield.Sensing and detecting plant hormones is essential for understanding plant biology and optimizing agricultural practices.Bioassays, immunoassays, electroanalysis, chromatography, capillary electrophoresis, and high-performance liquid chromatography (HPLC) are commonly used for the quantification of plant hormones. [129]These methods have advantages such as high sensitivity, selectivity, accuracy, and throughput for the detection of hormones in plant tissues.However, sample preparation, purification, and extraction are the challenging aspects of these methods.Pan et al. described the quantification of plant hormones from crude plant extracts using reverse-phase liquid chromatography-tandem mass spectrometry by enabling multiple reaction monitoring within a signal run from 50 mg of fresh plant tissue.The detection and extraction of major hormones from 40 samples for their analysis took 2-3 days, which is long and requires sample pretreatment. [130]A wireless monitoring of jasmonate was developed by the direct growth of pyrite iron sulfide (FeS 2 ) on cellulose paper and integrated with a microcontroller; sensing data were transported remotely to a smartphone via Bluetooth. [131]Still hormones analysis (sampling, extraction, purification, and concentration) in plants is challenging because of the minute concentrations of hormones in plant leaves.Advanced sensing technologies have been developed for the accurate and timely estimation of plant hormones.Biosensors integrated with microfluidic systems, for instance, employ biological recognition elements such as enzymes or antibodies to selectively interact with plant hormones, enabling the real-time and label-free detection of hormones. [132]utomated point-of-use sensors with IoT for hormone detection may not only provide a sustainable solution, but promote smart farming.Such sensors are designed for continuous and remote monitoring, and provide detailed information on hormone dynamics and their influence on plant health, growth, and development.These sensors accelerate data analysis, visualization, and decision-making processes through the wireless transmission of data to cloud-based platforms.The combination of IoT and hormone-sensing technologies enables the implementation of precise and targeted strategies for managing crops, thereby optimizing resource utilization, crop production, and environmental sustainability.Additionally, these systems allow for immediate adjustments in response to hormonal imbalances or stress conditions, leading to enhanced crop health, yield, and quality.Overall, IoT-based sensors dedicated to plant hormone sensing hold significant promise for advancing precision agriculture and plant science research while promoting sustainable and efficient agricultural practices.

Challenges, Critics, and Limitations
IoT-enabled food and agro-sensors have shown significant potential for revolutionizing agricultural industries, food production, and distribution.Nonetheless, there are many challenges, criticisms, and limitations to the integration of sensors with IoT technologies.One of the primary concerns is the security and privacy of the sensitive data collected by IoT sensors.Unauthorized access to farmer and consumer data can have severe consequences.However, ensuring the accuracy and reliability of IoT-enabled sensors is crucial for obtaining meaningful insights.Inaccurate readings can result from variations in the sensor quality and calibration issues, leading to incorrect decisions and resource wastage.
Furthermore, the implementation of IoT-based sensors can be expensive, particularly for small-scale farming or agricultural operations.For some users, the costs of the sensors, network infrastructure, and data management systems can be unaffordable.In addition, IoT sensors relying on batteries may face energy-efficiency challenges, requiring frequent replacement or recharging, particularly in remote areas, which can increase their real-world application costs.Additionally, users may require training to understand how to effectively use and interpret IoT sensor data, which can increase the total cost of their applications.Reliable internet connectivity is essential for IoT sensors to transmit data; however, it can be limited to rural or remote areas where most farms are located.Thus, it is essential to establish internet facilities and related infrastructure that could further increase the final cost.
The deployment of IoT sensors and their close contact with plants and farm animals can have environmental or biosafety impacts.The disposal of IoT sensors can contribute to electronic waste and environmental concerns if not managed responsibly.Addressing these challenges, criticisms, and limitations is vital for ensuring that IoT-based food and plant sensors are implemented sustainably, securely, and beneficially for all agro-users.Plum pox virus (PPV) [161] Stone fruit trees Electrolyte-gated organic field-effect transistor Anti-plum pox virus polyclonal Gold (Au) and pentacene films VOCs exhaled by Aspergillus and Rhizopus fungi [162] Strawberry Electronic nose -Carbone nanotubes VOCs exhaled by Phytophthora infestans infection [163] Tomato Chemiresistive sensor array -Graphene and Au nanoparticles (NPs) Citrus tristeza virus (CTV) [122] Citrus Electrochemical impedance spectroscopy Thiolated single-strain DNA Au NPs Phytophthora cactorum fungus [164] -Cyclic voltammetry/differential pulse voltammetry -Titanium dioxide and Tin oxide NPs Pantoea stewartii sbusp.stewartia (PSS) [165] -Linear sweep voltammetry HRP Au NPs Cucumber mosaic virus [166] Cucumber Chronoamperometry -Au NPs Citrus tristeza virus (CTV) [167] Sweet orange trees Amperometry Monoclonal antibodies Ab1 and Ab2 Au NPs Rice tungro disease [168] Rice Cyclic voltammetry anti-RTBV/RTSV Au NPs Xanthomonas axonopodis [169] Citrus Anti-PthA Glassy carbon electrode Au NPs Integrating food and plant sensors with IoT devices can revolutionize agriculture and the food industry, but various aspects must be considered for successful implementation.Several crucial aspects for the integration of each sensor with IoT devices are as follows: 1) Data collection and connectivity are required because these sensors need to collect data continuously or at predefined intervals.They should be equipped with communication capabilities (e.g., Wi-Fi, Bluetooth, LoRaWAN, Zigbee) to transmit data to the cloud or a central server for analysis.
2) The integration of sensors with IoT devices allows real-time monitoring.
When the system detects any anomalies, such as diseases, stresses, deficiencies, or other adverse conditions, it can send alerts to farmers or relevant stakeholders.3) Integration with smartphones is essential because farmers can benefit from mobile applications that provide easy access to sensor data, AI-driven insights, and the ability to remotely control and monitor crops.4) Scalability and flexibility are important aspects of sensors because the system should be scalable to accommodate a growing number of sensors and plants.In addition, it should be sufficiently flexible to adapt to different crops and farming practices.Based on these four criteria, a list of recent food and plant sensors is included in Table 2 using the specified criteria, and each study's adherence to these criteria is assessed".The results listed in Table 2 indicate that many of these studies have made significant progress in terms of scalability and flexibility.However, they are hindered by inadequate data collection, connectivity, and integration with smartphones, which hamper the creation of a robust sensor network, and consequently, the integration of IoT and AI in current food and agriculture sensors.This highlights the importance of prioritizing these two factors.

Future Directions
[135][136] With this in mind, AI is the next-generation solution for analyzing data and making quick decisions for smart farming and food analysis.Recently, the incorporation of new concepts including inject 3D-printing, [137] 5G/6G cellular systems, [138] and cyber-physical systems [139] with output feedback control into artificial intelligence has dramatically produced metamorphosis in the manner of food production. [140,141]emarkable changes in the food industry have been inserted into the global economy, and challenges such as food security and sustainability have become tightly linked.This incorporation of the food supply chain is moving toward a productive and interconnected food industry.Nowadays, the use of smartphones, the connection among different players in the supply chain, the IoT, and the improvement in computing process power and facile access to data and information are unprecedented. [142]This technology infuses food production, packaging, and storage, and the entire supply chain is defined by the Fourth Revolution.In the conventional approach, analytical techniques consistently evaluate food quality using single-parameter measurements.However, in the Fourth Revolution era, a new paradigm focused on complex, diverse, and interconnected food and agriculture systems to ensure food quality, traceability, and safety is in demand.
As indicated in Figure 7, the Fourth Revolution research effort is to combine both the food and agro-industries with IoT and AI using accurate state-of-the-art intelligence sensors and telecommunication systems for data acquisition and conveyance.Multiplex, accurate data acquisition can be achieved through a network of wearable food and plant sensors in direct contact with their targets or through modern imaging systems mounted on robots and UAVs.Telecommunication systems, including satellites, convey encrypted information for data processing.The final comprehensive information is transferred to stakeholders and decision-makers.The research scope for each step is illustrated in Figure 7.
Next-generation wearable sensors (all-in-one-chip) are promising plant health monitoring systems owing to their excellent sensing resolution, wireless networking capability, flexibility, real-time screening, integration with existing smart technologies, including AI and IoT, and widespread accessibility. [143,144]imilar to sensing technology in other areas, such as human healthcare devices, plant sensors face common challenges, including selectivity, sensitivity, operational stability, and scalable manufacturing.As a long-standing challenge, the integration of additional technologies, including power supplies, data acquisition devices, and wireless transmission, with sensors and biosensors to make them independent in applications is significant.For example, human healthcare devices utilize various technologies rather than large batteries for their power supply, such as redox reaction energy generation, mechanical energy harvesting, [145] thermoelectric local energy supply, [146] epidermal biofuels, [147] and wearable fuel-cell energy conversion. [148]owever, the nonidentical nature of plants makes their energy harvesting different from that of wearable human sensors.Solar energy harvesting is an alternative to batteries for long-term measurement of on-plant wearable sensors.Furthermore, the integration of solar energy conversion devices, such as photovoltaic (PV) panels, with plant wearables is a promising solution. [149]oreover, each wearable plant system mounted on an individual plant generates a large amount of data.The sensor density in each unit area of a farm should be optimized to fulfill specific applications, such as the spread range of a disease.This can also be multiplied by the amount of data generated by the network of sensors.Therefore, these sensing systems could be combined with AI for analyzing the big data and achieve the "Smart Farming'' concept. [150]s the health status of a plant can be evaluated comprehensively through the simultaneous measurement of different biomarkers and pathogens, it is important that wearable devices can measure various biomarkers or parameters simultaneously. [151]his can be achieved by developing a sensor with multiplexed operating functionalities, known as sensor arrays, which perform similarly to the mammalian olfactory system. [152]Moreover, to investigate plant responses to both abiotic and biotic stresses simultaneously, other types of sensor modalities, such as stress-strain, humidity, and temperature sensors, can be integrated with biomarker sensing. [153]As a result, both the multimodal and array sensors for wearable plant sensor applications could help farmers gain a broader view of their farm's health and assist in the best-practice management of crop production and sustainability.
In summary, IoT-enabled food and plant sensors show promise for improving food safety, health, productivity, and sustainability in agriculture.We summarized some key aspects that highlight their perspectives, such as, 1) predictive analytics and AI: the convergence of IoT sensors with AI algorithms offers the potential for predictive analytics, thus empowering farmers and food producers to proactively anticipate challenges such as disease outbreaks, weather patterns, and market demands.This forward-thinking approach enhances risk management and boosts the overall productivity and yield.2) Integrated smart supply chains: incorporating sensors into food packaging and storage facilities enables the continuous monitoring of crucial environmental factors such as temperature and humidity.Real-time data ensures the freshness, quality, and safety of food  [37]   Opaque/Transparent Ttis Food packaging division TTI b) [40]   Chemical-sensitive field effect transistors (CSFETs) Pork quality Moisture, ammonia, and hydrogen sulfide c) [43]   Electrical resistance sensor Pesticide detection (þ)/(À)-Methamidophos [53]   Point-of-care kit for entire test (POCKET) Environment, food, and agriculture Multiple types of DNA [58]   Capillary electrophoresis microchip -coupled laserinduced fluorescence (MCE-LIF) Beverage sector Escherichia coli [60]   Clustered Regularly interspaced short palindromic repeats (CRISPR)-Cas12 Fresh egg S. enteric [64]   Colorimetric bioassay Foodborne Bacteria Salmonella [65]   Microneedle-based sampler/ sensor Food packaging/supply chain Escherichia coli [72]   Microneedle-based plant sensor Agriculture sector Tomato spotted wilt virous (TSWV) [94]   Microfluidic-printed electromechanical sensor of stomata

Plant and agriculture sector
Methyl jasmonate [132]   Portable photothermal immunosensing supported by MEMS Food storage division Mycotoxins (aflatoxin B1, AFB1) [171]   Photothermal immunosensing Food industries Mycotoxins (aflatoxin B1, AFB1) [172]   a) This study has reported this criterion.b) The study has not reported this criterion.c) It is not clear if the study includes this criterion.
throughout the supply chain, minimizes waste, reduces spoilage, and enhances food traceability.3) Data-driven decision-making: the abundance of data gathered through IoT sensors provides valuable insights for farmers, food producers, and policymakers.By analyzing these data, stakeholders can identify trends, optimize production processes, and implement targeted interventions to improve efficiency, sustainability, and profitability.4) Indoor gardening and urban farming: as urbanization surges, the demand for locally grown food has increased.IoT sensors facilitate efficient monitoring and control of indoor farming systems, ensuring optimal growth conditions for plants in confined spaces.This technology has significantly contributed to sustainable food production in urban areas.5) Integration with distributed databases and blockchain: integrating IoT sensors with blockchain technology enhances transparency and trust in food systems.By recording sensor data in an immutable ledger, consumers gain access to accurate information about the origin, quality, and safety of the food they consume.Overall, the future of IoT-enabled food and plant sensors lies in enhanced automation, data-driven decision-making, and sustainable practices.As technology progresses and costs decrease, the widespread adoption of these sensors is poised to revolutionize the agriculture and food industries, fostering more efficient, resilient, and sustainable food systems.
Ajeet Kaushik, Fellow-ICS, is working as an assistant professor of chemistry at the Department of Environmental Engineering of Florida Polytechnic University, USA.He is exploring nano-enabled technologies for health wellness and environmental monitoring, involving efficient sensing and nanomedicine.He is an accomplished scholar (supported by over 250 publications, editorial roles, 10 edited books, 3 patents, and international collaborations) and the recipient of several international awards in support of his credentials.His research interests include green chemistry, electrochemistry, chemical sensors, biosensors, nanomedicine, point-of-care sensing, and personalized sensing.To achieve goals, Dr. Kaushik is focused on cutting-edge research and seeking collaborations.

Figure 2 .
Figure 2. Food cycle and associated challenges (contaminants, toxins, and pathogens) to maintain safety and quality, needed for health wellness.

Figure 4 .
Figure 4. a) Toward food-on-a-chip: the illustration of sensing technology supported by IoT and AI for food quality and safety assessment at the point of testing location, even in a personalized manner.b) Effect on food contaminants in biological cells and organs.

Figure 5 .
Figure 5. IoT-enabled plant sensors.a) Integrated microneedle-smartphone nucleic acid amplification platform for on-site diagnosis of plant diseases.[94]This device consists of a polymeric microneedle array for DNA extraction and a smartphone-based 3D-printed device for loop-mediated isothermal amplification reaction and detection.Adapted with permission.[94]Copyright 2021, Elsevier.b) A plant drought sensor based on nanoprinted ink to detect water respiration.[95]This sensor has been placed on top of the stomata underneath the plant leaves.Adapted with permission.[95]Copyright 2021, Royal Society of Chemistry.c) Biomimetic implantable in-planta sensor for long-term, continuous, and in situ monitoring of plant nitrate.Adapted with permission.[99]Copyright 2022, American Chemical Society.d) Diurnal in vivo xylem sensor based on an organic electrochemical transistor for continuous detection of glucose and sucrose.Adapted with permission.[100]Copyright 2021, Elsevier.e) Wearable plant sensor for in situ detection of pesticide.[101]This smartphone wearable sensor was interfaced with a wireless sensor node to collect data remotely.Adapted with permission.[101]Copyright 2020, Elsevier.

Figure 6 .
Figure 6.Plant wearable sensors.a) Carbon nanotube-graphite-based FET sensor interfaced onto a plant leaf to detect plant VOC.Adapted with permission.[108]Copyright 2014, American Chemical Society.b) Real-time sensing of different concentrations of dimethyl methylphosphonate (DMMP).Adapted with permission.[108]Copyright 2014, American Chemical Society.c) A vehicle-based robotic system for high throughput field phenotyping.Adapted with permission.[112]Copyright 2019, J. Murman.d) Schematic of hooked miniature machines for on-leaf sensing and delivery.Adapted with permission.[119]Copyright 2021, Springer Nature.e) Schematic of the experimental setup for a nanobionic sensing plant comprising of a nanosensor embedded on the plant leaf to a portable Raspberry Pi-based electronic device.Adapted with permission.[121]Copyright 2016, Springer Nature Ltd. f ) A multimodal flexible sensor attached on the bottom side of a leaf to detect temperature and humidity, and a schematic of the device structure with the functional components.Adapted with permission.[160]Copyright 2020, American Chemical Society.
Md. Azahar Ali is an assistant professor in the School of Animal Sciences at Virginia Tech and an affiliate faculty of Biological Systems Engineering at Virginia Tech, Blacksburg, VA.Earlier, he was a postdoc in the Department of Mechanical Engineering at Carnegie Mellon University, Pittsburgh, PA.He also did another postdoc in the Department of Electrical and Computer Engineering at Iowa State University, Ames, IA, where he received an excellent research award.He obtained his Ph.D. degree from the Indian Institute of Technology Hyderabad.He is actively engaged in the area of bioMEMS and 3Dprinted biosensors for point-of-care livestock sensing, wearable sensing, and precision agriculture.His research also focuses on nanomaterials-based biosensors for implantable, pathogens, and biomarker sensing and soil/plant sensing.He has 70 peer-reviewed publications with a Google Scholar citation of 4075, seven patents, one textbook, two book chapters and many conference papers.

Table 1 .
Quantification of plant pathogens using electrochemical methods combined with nanomaterials.

Table 2 .
IoT and AI integration capabilities of various food and agro sensors.