REVIEW article

Front. For. Glob. Change, 08 April 2026

Sec. Pests, Pathogens and Invasions

Volume 9 - 2026 | https://doi.org/10.3389/ffgc.2026.1777311

Revolutionizing pest control: harnessing cutting-edge technologies in controlling forest pests

  • 1. Faculty of Computer Science, Department of Information and Communications Engineering, University of Murcia, Murcia, Spain

  • 2. University of Agriculture Faisalabad, Constituent College Burewala, Punjab, Pakistan

Abstract

Global climate change and increased anthropogenic activities are responsible for outbreaks of invasive and endemic pests, diseases, and pathogens. Forest pests pose serious threats to the growth, productivity, and resilience of natural forests and green urban spaces globally. To cope with the hazardous impacts of trunk-boring pests on forest degradation, it is crucial to detect their infestations at early growth stages. However, the borer larvae’s hidden lifestyle and delayed apparent symptoms lead to widespread infestation, resulting in large-scale tree mortality. The development of smart systems is critical to protecting natural forests and forest plantations from the fatal impacts of boring pests. Applications of artificial intelligence (AI), Internet of Things (IoT), and remote sensing (RS) technologies are transforming traditional pest management strategies by developing more sustainable and robust solutions. The integration of these cutting-edge technologies is paving the way for early pest detection, identification, and outbreak prediction by implementing on-site decision support systems (DSS) and remote monitoring of agricultural fields and large-scale forests. Advanced IoT acoustic sensors can record vibrational signals of boring pest larvae activities inside the tree trunk to confirm pest infestations at early stages, assisting forest management stakeholders in taking preventive measures to suppress the pests’ outbreaks in a timely manner. Auditory signals recorded by piezoelectric sensors are affected by external environmental noise. To improve pest detection ability, different AI-based signal enhancement algorithms are deployed to suppress noisy signals and process enhanced signals. Enormously improved RS technologies can monitor dynamic structural changes in forests by capturing multispectral images through satellites and drones equipped with high-resolution cameras, and then the RS data are incorporated into advanced AI algorithms to analyze the pests and disease-induced stress in the forest ecosystem to develop smart forest protection systems. The integration of X-ray imaging and deep learning models is another achievement in the non-destructive management of trunk-boring pests.

1 Introduction

According to reports by the Food and Agriculture Organization (FAO) and United Nations Environment Program (UNEP), world forests cover up to 31% of the total Earth’s land surface, of which more than one-third is primary forest, and the remaining half is comparatively undamaged by human disruptions (Guégan et al., 2023). The role of forests and green urban spaces is of great importance due to their ecological and socio-economic services in sustainable ecosystem development. Different climatic variables, such as high temperatures, windthrows, drought stress, and pest attacks, are responsible for unprecedented deterioration in conifer forests. High temperature is a main contributing factor for most insect pest outbreaks, as the severity, cyclicity, and frequency of pests’ abundance are now climate-driven. Forest pests have extended their geographical distribution from native hosts to global non-native hosts (Ryalls et al., 2024). Due to anthropogenically driven climate change, high temperatures have disturbed European forests over the last two decades, most dominantly through the upsurging attacks of pathogens and insects, as there is a direct link between high temperature and increasing pest outbreaks. The rapid spread of non-native menacing pests and pathogens is a consequence of globalized fast trade, human activities, and a lack of defensive adaptations in trees, resulting in the fast range extension of notorious pests due to climate change (Hartmann et al., 2025). Forest plantations dominated by monoculture trees have less potential to cope with diseases and pests, leading to increased susceptibility to pest occurrences (Liu et al., 2025). Due to these substantial climatic changes, both local and non-native invasive pests and diseases are impacting approximately 65.3% of the forest area in Asia and Europe (Hartmann et al., 2025; Liu et al., 2025). Forest pests and diseases are also continuously degrading timber quality, resulting in a reduction in the economic viability of the wood production industry (Sevinç, 2025).

1.1 Boring pests: a comprehensive background

Boring pests are responsible for forest devastation as they cause severe damage to tree growth. Due to their rapid spread, they harm tree health destructively, resulting in early tree mortality and reducing the ecological efficacy of forests and urban green spaces. The term “boring insects” refers to those insects that create trunk holes as they bore into trees for feeding; in most cases, they live inside the trees. Popular examples of boring pests are the emerald ash borer (EAB) Agrilus planipennis Fairmaire, small carpenter moth (SCM) Streltzoviella insularis Staudinger, red palm weevil (Rhynchophorus ferrugineus), juniper bark borer (Semanotus bifasciatus), longhorn beetles (Cerambycidae family), mango palm weevil (Sternochetus mangiferae), and Asian longhorn beetle (ALB) Anoplophora glabripennis Motschulsky (Jiang et al., 2024a; Çimen et al., 2025; Jiang et al., 2022; Potamitis and Rigakis, 2024; Banlawe and dela Cruz, 2023; Wang et al., 2023). Forest beetles damage forest resources, and trees damaged by pests and diseases increase the risk of forest fires because the thin branches and leaves of dead trees are easily flammable (Sevinç, 2023).

Boring pests are devastating and insidious, making their early detection a challenging task for forest managers. Larvae of these boring pests cause damage to trees as they burrow into the trunk’s wood. The tunnels they develop hinder nutrient and water transport, resulting in plants withering and drying. The larvae of boring pests also bore into the internal organs of trees and damage the conductive tissues, i.e., xylem and phloem, leading to the weakening of trees, growth retardation, and, in some cases, tree mortality (Li et al., 2024a; Jiang et al., 2022). Due to the feeding behavior of boring pests, lesions develop inside trees, which act as entry points for different pathogens, e.g., bacteria and fungi (Cappaert et al., 2005; Aukema et al., 2010). Boring pests reproduce on a large scale, and their quick spread has devastating impacts on forest ecosystems. Tree mortality on a large scale not only disrupts forest structure and species composition but also disturbs the habitats and food sources of other animals and plant species residing in forests. After a tree’s death, the risk of forest fire increases due to the fallen deadwood on the forest floor (Li et al., 2024b). Unlike other pests, the larval stages of boring pests are invisible due to delayed apparent symptoms. Infected trees can only be diagnosed in the middle and late stages of infestation, and at that time, it becomes challenging for forest managers to take preventive measures for their population control (Jiang et al., 2022).

1.2 Importance of boring pests’ surveillance

Boring pests concealed infestations inside tree trunks, logs, and other wood-packing materials pose a significant challenge for all pest management stakeholders, including pest managers, researchers, and regulators (Jiang et al., 2021a). Traditional methods for controlling boring pest populations are often inefficient, as the most common approach for their management is the utilization of sex pheromones (Jiang et al., 2024b). For early pest detection, the commonly adopted approaches are manual field inspections, which demand higher expertise in pest detection, as it is a laborious, costly, time-consuming practice and limited in accuracy (Ahmed et al., 2024). Manual inspections for wood-boring pests are not suitable for efficient large-scale pest monitoring. The use of pesticides to suppress pest populations leads to environmental pollution, kills beneficial insects, has hazardous impacts on human health, and develops pest resistance against chemicals (Li et al., 2024a). Utilizing different traps such as color-attractant traps, sex-pheromone traps, aggregation-pheromone traps, and girdled-tree semichemical attractant traps, as well as employing image detection and remote sensing technologies to control pest outbreaks, are utilized for early pest detection, but these approaches have proved insufficient in detecting boring pest population occurrences due to their hidden lifestyles. These techniques provide valuable insights to timely forewarn growers about the expected pest outbreaks. To control the damage caused by boring pest larvae inside host trees, there is an urgent need to develop innovative methods for detecting pest infestations. It is quite challenging to detect concealed pest incidences without conspicuous symptoms of their infestations. Earlier detection challenges accompanied by delayed infestation symptoms can lead to irreversible losses (Jiang et al., 2024b).

This alarming situation highlights the critical requirement for improved strategies that combine traditional pest management practices with advanced technological innovations to combat the devastating impacts of pests (Vidya Madhuri et al., 2025). To mitigate the losses caused by threatening xylophagous insects, the implementation of intervention strategies is crucial, but locating the exact location and infestation severity, along with surveillance of the gradual evolution of pest populations and accompanying infestations, is also beneficial in the planning of management strategies. Visual symptoms of infestations, e.g., leaf discoloration and falling, and exit tunnels typically appear late, at which time controlling larval infestation becomes irrepressible (Potamitis and Rigakis, 2024). Substantial progress has been observed regarding the development of enhanced procedures for the detection and surveillance of boring pests in the recent decade (Jiang et al., 2021b). Recent advancements in the field of sound recognition technologies are providing new avenues for pest recognition by processing their sound signals to confirm their occurrences (Jiang et al., 2024b). The purpose of advanced monitoring methods is to detect boring pests in their early phases of infestation, aiming to facilitate informed decision-making (Potamitis and Rigakis, 2024). Improvement in boring pest management accuracy can serve to protect agriculture and forestry resources, as well as achieve high economic and ecological benefits (Li et al., 2024b).

1.3 Adoption of innovative technologies for boring pest management

Utilization of modern technologies is revolutionizing plant protection, which includes real-time monitoring of crops, precise pesticide applications, and enhancements in pest and disease detection approaches. These innovative solutions are not only improving the efficacy of pest management but also protecting the environment by introducing sustainable farming practices. Among these technologies, pest detection and identification, early pest population prediction, targeted spraying, and DSSs are ensuring crop and environmental safety. Recently, the modernization of technological solutions has introduced innovative pest population tracking and scouting approaches encompassing AI, IoT, and RS technologies that assist in smart pest management. These technological innovations facilitate pest infestation prevention and notable reductions in crop yield losses, as well as solve problems related to dynamic climate patterns (Nahiyoon et al., 2024).

To overcome the losses caused by boring pests in forests, researchers are leveraging IoT technologies and AI-based intelligent algorithms. Researchers are utilizing AED-2000 and AED-2010 equipment along with the SP-1 L piezoelectric sensor probe for the surveillance of boring pest populations by recording and interpreting vibrational signals produced by their locomotion and feeding activities. The latest advancement in the early detection of larval infestations inside the trunk involves recording their vibrational signals using IoT sensors, incorporating these signals into an AI model, e.g., deep learning (DL) networks, and leveraging the automatic pattern extraction capabilities of these networks for accurate pest larvae detection instead of manual tree inspections. Earlier detection of larvae inside wood can help in developing customized management solutions, such as injecting insecticides only into target plants instead of random applications of pesticides to all trees, which have consequences of increasing environmental pollution and its toxic impacts on human health (Wang et al., 2021; Jiang et al., 2024a). Improvements in neural network architectures are paving the way for smart pest monitoring and sound recognition by employing speech recognition technologies. Sensor deployment for vibrational signal recording could be fixed on trees, as well as using portable sensors such as piezoelectric sensors, which are suitable due to their sensitivity and better, less noisy signal recordings. Furthermore, these signals are recorded in audio format and then processed by AI algorithms for larvae detection and classification (Kahl et al., 2018; Liu et al., 2022; Jiang et al., 2024a).

1.4 Purpose and area of investigation in the review

Numerous review studies have underscored the potential of cutting-edge technologies in the realm of smart pest management, but there remains a critical research gap regarding the use of AI, IoT, RS, and X-ray imaging in trunk-boring pest management. This study makes a novel contribution by reviewing the potential benefits of cutting-edge technologies in trunk-boring pest management, with particular focus on bioacoustics and vibrational signal analysis, thereby distinguishing it from previously published work.

The purpose of this study is to investigate the utilization of advanced technologies such as AI data analytics, IoT, and remote sensing tools in the management of boring pests. This study highlights the transition from traditional boring pest management strategies, e.g., manual field observations, trapping adult pests, and destructive dissection approaches, to modern pest management approaches. This comprehensive review also entails the potential benefits of AI and IoT integration in controlling boring pest populations by forewarning about larval infestations at earlier stages.

By combining the latest research, technological advancements, and their practicality, this study focuses on the benefits of using these technological solutions to address key global challenges related to forest protection, sustainable ecosystem development, and climate change. It also provides a deeper understanding of modern technologies for solving prevailing boring pest management problems and the transformation of traditional forest ecosystems into more sustainable and resilient systems. This detailed document provides a roadmap for future directions and research in sustainable boring pest management by integrating innovative technologies, as well as highlighting their efficacy in integrated pest management (IPM) programs to pursue the objectives of long-term environmental sustainability.

2 Overview of AI and IoT technologies

2.1 Exploration of AI within technology landscapes

AI is an interdisciplinary field integrating computer science, cognitive psychology, statistics, and mathematics, and is gaining much popularity with its innovative and smart applications in diverse fields (Jaboob et al., 2024). AI exhibits machine-displayed intelligence that mimics human cognitive abilities, e.g., problem-solving and decision-making. Among the other subfields, machine learning, deep learning (DL), computer vision, and natural language processing (NLP) are the key branches of AI (Figure 1). By combining the predictive capabilities of machine learning and DL approaches, AI algorithms are fitted on large-scale datasets to make data-driven and informed decisions autonomously. AI algorithms perform data analysis and extract meaningful patterns from data to replicate human intelligence. Pattern matching and using cognitive computing for explaining data are central to AI functioning. The aim behind the development of intelligent machines is to simulate human cognitive abilities, e.g., learning, perceiving, and reacting according to the prevailing situations. It has been forecasted that the global AI software market is expected to reach $126 billion in 2025 (Rashid and Kausik, 2024). Advancements in AI algorithms provide opportunities in science and technology fields, and these technological inventions are popular among billions of people. Individuals who interact with a smartphone can leverage the potential of computer vision, NLP, and communication network technologies (Zincir-Heywood et al., 2021). An introduction to the major subfields of AI is given in the next section.

Figure 1

2.1.1 Machine learning

Machine learning (ML) is a subfield of AI that utilizes statistical methods to strengthen computing machines with learning and autonomous decision-making capabilities without explicit and extensive programming. It is based on the theory that machines can extract knowledge from given data, recognize patterns, and draw conclusions with minimal or no human intervention. The major categories of ML are supervised, semi-supervised, unsupervised, and reinforcement learning. They focus on continuous improvement in machine performance by reflecting upon the previously stored and newly generated data. Supervised learning models are incorporated with labeled datasets, and this category is primarily divided into two forms: regression and classification. Regression models predict continuous results, while classification models produce class labels or categorical outputs. Supervised learning models facilitate the analysis of different types of statistical data, and they can be applied in a diverse range of applications such as predictive modeling, data mining, image classification, and anomaly detection (Alnuaimi and Albaldawi, 2024; Farina et al., 2022). In unsupervised learning, models are trained only on input parameters, and they learn insights from the underlying structures of data instances, thereby extracting patterns from the given data. Its major categories are clustering and association (Sahu et al., 2023). In semi-supervised learning, during model training, class labels are not given for the whole training data as some training data are labeled while the remaining is unlabeled, and algorithms try to extract meaningful patterns from both types of given data to make predictions on unseen testing data, such as positive-unlabeled learning (Yan and Wang, 2022).

2.1.2 Deep learning

DL is a leading class of ML that emerged recently with the aim of providing more computing power at less cost. In DL, mathematical algorithms are employed using neural networks, as these networks have been built by inspiring the neurological structure of the human brain. The motivation behind DL algorithms is to simulate the human brain’s functioning for data learning and interpretation. DL models interpret low-level features of given data and build more abstract and high-level features by determining complex underlying patterns. Due to their automatic feature extractions, these algorithms facilitate solving complicated and unstructured data problems such as image and sound recognition. DL algorithms can also be supervised, unsupervised, and semi-supervised; an analogy with ML algorithms exists (Yan and Wang, 2022). There are different types of neural architectures, including artificial neural network (ANN), deep neural network (DNN), recurrent neural network (RNN), and convolutional neural network (CNN), as well as different variants of these networks (Farina et al., 2022). In recent years, self-supervised learning has emerged as an important DL approach, attracting extensive public attention. Major ML approaches such as reinforcement learning (RL) and transfer learning (TL) have progressed with the adoption of DL models (Yan and Wang, 2022; Esteva et al., 2019).

2.1.3 Generative AI

Now, generative AI (GenAI) is standing at the forefront of AI, depicting the capabilities of creating novel content by surpassing the boundaries of computational creativity, as AI has earned much public attention with the introduction of popular models of generative AI, e.g., ChatGPT by OpenAI and generative adversarial networks (GANs). Generative AI models follow innovative training procedures to create realistic content such as images, contextual information, and even original music. Tremendous success has been made in this field, including text and image generation, music composition, interactive application development and designing video games, etc. It concentrates on interactive, personalized, and immersive user experiences by utilizing multimodal approaches (Glotzbach, 2024). It has the potential to create new content, just like human-generated content. Continuous research and development are underway in GenAI to build faster, better, and more capable models. GenAI provides innovative solutions for diverse fields, but at the same time, it suffers from problems of data misuse, bias, and transparency, which must be addressed for the successful implementation of GenAI models (Banh and Strobel, 2023).

2.1.4 Introduction to internet of things (IoT) technologies

A continuous evolution in technologies has been recorded in the last few years, among which IoT is another growing field, encompassing a universal computing network in which everything and everyone is connected to the Internet. IoT is a continually growing network with infinite opportunities, making it a hot research topic. IoT is defined as the integration of high-tech technologies by employing “smart” sensors and gadgets in the form of a single network for the sake of data collection. All connected devices, including sensors, computers, and network devices, can transfer data and information by utilizing different network protocols (Farooq et al., 2015; Salama et al., 2023; Mansour et al., 2023). This network can connect multiple devices due to recent advancements in communication technologies, including wired, wireless, and hybrid media. Lightweight IoT protocols facilitate IoT devices with limited hardware resources in terms of computation, energy, and storage, but it is necessary to opt for the best communications protocol compatible with the system architecture (Mansour et al., 2023). The IoT ecosystem assists the reliable transmission of information between physical and digital components, fostering the development of creative applications, automation, and data-driven decision-making. These networks are becoming popular in transportation, agriculture, medical, and industrial sectors with the introduction of smart systems such as smart cities and home projects (Gupta and Nagpal, 2024). Despite their applications in diverse areas, machines cannot surpass human abilities, but their potential for data collection and further analysis are the strongest aspects of IoT (Salama et al., 2023).

2.1.5 Major components of IoT network

A diverse set of components belongs to IoT frameworks, which facilitate devices to connect and exchange data and enable the seamless integration of applications (Figure 2). Sensors are actual physical components that can sense and collect environmental data. Actuators are also a type of sensor that can perceive and react to environmental conditions. Popular categories of sensors include wearable sensors, environmental sensors, smart cameras, motion detectors, and smart meters (Qazi et al., 2022; Gupta and Nagpal, 2024). With an increase in things or devices, their connectivity is of utmost importance. For IoT applications operating in small coverage areas, Bluetooth, Zigbee, Optical Wireless Communication (OWC), and Wi-Fi protocols are deployed for connectivity. In the case of wide-range IoT applications, such as unmanned aerial systems (UAS) and the implementation of environmental sensors in outdoor environments for monitoring, long-range connectivity for data exchange to a central server is required. Wide-range communication technologies utilized include LoRaWAN, satellite, Sigfox, and cellular networks (2G-5G) (Ding et al., 2020). Gateways connect physical devices and central systems, gathering data from different connected devices, performing data preprocessing, and transferring this information to cloud computing or edge computing servers. Gateways also facilitate local data storage, protocol translation, and security. Cloud servers deal with massive amounts of data produced by IoT devices and analyze the data. Important characteristics of cloud infrastructure include computing power, networking, and storage resources (Gupta and Nagpal, 2024). Edge computing expands the abilities of computation, network connectivity, and storage from cloud infrastructure to edge devices, resulting in a reduction of massive data traffic. This component of the IoT framework provides additional benefits of agility and real-time data processing at edge devices instead of burdening the central server (Chen et al., 2018). IoT platforms offer middleware and software tools to administer, track, and assess the data and devices of the IoT framework. These tools provide vertical services to the front-end user by offering software tools, and they also perform computation for managing connected devices and data lifecycle management (Asemani et al., 2019). To avoid unauthorized access, protect confidential information, and ensure integrity, security solutions are also an important component of the IoT framework. To extract meaningful insights from data collected through IoT devices, it is crucial to utilize AI models and data analysis techniques (Gupta and Nagpal, 2024).

Figure 2

3 Integration of AI and IoT in the agriculture sector

The agricultural sector is facing critical challenges related to the production of high-quality crop yields to meet the food demands of an ever-increasing world population. It is expected that the world population will rise to 10 billion by 2050, and in the meantime, compatible growth in agricultural production is demanded. While pursuing these objectives, several obstacles such as climate change, natural disasters, labor shortages, and resource scarcity are faced by farmers and other stakeholders (Ghobadpour et al., 2022; Elbasi et al., 2023; Bellout et al., 2024a). Conventional agricultural practices rely on field experts’ knowledge, resulting in inefficient resource utilization and an increased risk of pest and disease attacks due to extreme climatic events. Employing manual labor in agricultural practices is time-intensive and laborious (Khakimov et al., 2022; Bai et al., 2023). The deployment of AI intelligent models and IoT technologies in the agricultural ecosystem offers extraordinary opportunities to enhance farming practices, optimize resource utilization, and improve resistance against climatic stressors (Figure 3). Integrating AI algorithms and IoT technologies is bringing revolutions to the agricultural landscape by introducing smart agricultural solutions that enable real-time field monitoring, precision agriculture, and data-driven decision-making. Enhanced utilization of various sensors, including weather, optical, acoustic, and soil sensors, together with AI algorithms, is executed to improve fertilization and irrigation strategies and to enable smart pest and disease management (Miller et al., 2025).

Figure 3

IoT systems have the potential to gather high-resolution and large-scale data on weather parameters, soil characteristics, crops, and pest phenological patterns to assist in data-oriented decision-making instead of relying on manual field observations to gain insights about crop health. The valuable data generated by smart sensing technologies is further integrated with AI algorithms, including ML and DL models, to analyze data and extract meaningful patterns to forecast expected outcomes and trigger automated responses according to prevailing conditions (Janiesch et al., 2021; Saleem et al., 2021). Synergizing AI models and IoT technologies enables precision agriculture, assisting farmers in utilizing resources, such as water, pesticides, and fertilizers, in an optimized manner, thus minimizing costs and reducing environmental impacts (Miller et al., 2025).

Precision farming practices supported by AI and IoT integration result in better accuracy, i.e., 98.65%, and improved efficiency of resource utilization via variable rate applications derived from real-time insights. “Vital” is a smart irrigation system that demonstrates the utilization of smart sensors and explainable AI (XAI) to irrigate land automatically based on climatic and soil parameter data. AIoT smart systems can optimize crop protection by predicting pest and disease attacks in advance, as well as enhancing animal welfare by utilizing a continuous livestock monitoring framework. These smart systems are equally beneficial for environmental protection through reductions in waste resources and greenhouse gas emissions (Bara et al., 2025). To accurately identify tomato diseases, Zarboubi et al. (2025) introduced a computationally efficient CustomBottleneck-VGGNet model that achieved 99.12% accuracy, outperforming other competitive models. The proposed model was integrated into a mobile application for the real-time detection of tomato disease. In another study, an innovative approach that combines the YOLOv10 model with a remotely controlled robot car for the smart detection of strawberry disease was proposed (Bellout et al., 2024b). Along with the tremendous benefits of AIoT smart systems, several challenges still exist in the adoption of these technologically advanced solutions by farmers due to their concerns, including high infrastructure costs, data privacy, and security problems. Farmers are reluctant to deploy these smart solutions due to a lack of proper training, as traditional farmers are not tech-savvy (Qazi et al., 2022).

4 Synergizing AI models and IoT technologies in pest management

Detection and management of pests and diseases in the early phases of infestation are significant challenges for protecting crops from devastation. Pests’ attacks are responsible for the degradation of the quality and quantity of agricultural production, resulting in disruption in the food supply chain and food insecurity. Conventional methods of pest management are laborious, time-consuming, and error-prone. To mitigate these challenges, leveraging advanced technologies such as AI, IoT, and remote sensing has the potential to compete with the environmental hazards and agricultural losses caused by pests and diseases (Pierre Nyakuri et al., 2024). This synergy of AI and IoT technologies is paving the way for more advanced research in pest management. By deploying interconnected sensor networks, advanced AI algorithms, and cloud and edge computing platforms, AIoT systems facilitate site-specific surveillance and control of critical abiotic and biotic parameters to ensure the continuity of the food supply chain, as well as protect human health and the environment from the lethal impacts of excessive pesticide applications (Zarboubi et al., 2024a; Bayar et al., 2025). For the development of pest detection and identification systems, data are collected through IoT-based smart cameras, and furthermore, these garnered data are incorporated into AI-based object detection models to confirm the presence of specific pest species in crops in order to implement mitigation measures (Hasan et al., 2024). For the development of early pest population prediction systems, IoT sensors are deployed to collect meteorological parameter data, such as relative humidity, temperature, and rainfall, as well as crop and pests’ phenological stage data, and AI algorithms are fitted on sensor-based data to forewarn the farmers about expected pest outbreaks (Saleem et al., 2023). Figure 4 highlights the major applications of AIoT systems in the pest management domain.

Figure 4

4.1 Fusing AI and IoT for pest detection and classification

Prasath and Akila (2023) proposed a pest detection and classification model by integrating agriculture-based IoT sensors for data collection, an AI-based object detection model You Only Look Once v3 (YOLOv3), and an enhanced classifier for pest detection and classification in crops, respectively. For the real-time monitoring and detection of stored-grain pests, Zarboubi et al. (2024c) integrated a YOLOv8 model with a remote-controlled robot car to offer a proactive, data-driven approach that supports farmers through smart pest control.

To automate the detection of pest infestation inside fruit orchards without farmer intervention, Albanese et al. (2021) developed an embedded sensing system combined with a neural accelerator to acquire and process pest infestation images. They deployed various ML algorithms on the collected image dataset and concluded that the proposed architecture has the potential to detect pest infestation automatically. Gokeda and Yalavarthi (2024) proposed a deep hybrid model-based pest detection (DHBPD) model for pest detection. The proposed architecture is a combination of an IoT-based system integrated with an unmanned aerial vehicle (UAV) along with the DHMPD system, resulting in good pest detection accuracy. Saranya et al. (2023) used a transfer learning approach bootstrapping along with a UAV-based IoT framework for detecting and classifying pests. The proposed architecture, VGG16 fused with enhanced and optimized fully connected layers, outperformed other competitive models by resulting in better accuracy and minimal loss values. To tackle the challenges posed by pest infestations and crop diseases, an enhanced version of the YOLOv5s model customized to work with UAV-based applications was developed. It offers a more reliable solution for real-time pest detection and classification within a drone-centric domain (Khan et al., 2024). Rahman et al. (2024) developed an automated system for the timely identification of paddy pests by integrating DL, IoT, and blockchain technologies. An additional feature of the proposed architecture was the utilization of a blockchain security mechanism to ensure the secure transmission of data in the IoT framework.

Ahmed et al. (2024) proposed a smart pest detection system by integrating DL and IoT for detecting oriental fruit flies in guava orchards. They utilized a smart pest trap embedded with pest detection and identification computation and proposed a CNN classifier that delivered the best results compared to other ML and DL models. For detecting Tuta absoluta infestation in tomato cultivation, Christakakis et al. (2024) integrated AI and smart sensing technologies with smartphone technology by highlighting the potential of these technologies in pest and disease monitoring in the agriculture sector. DL pest detection and identification models were deployed on a decision support system (DSS), and to facilitate users, a mobile application was developed through which users capture and upload pest data to the DSS for obtaining meaningful and data-driven insights for pest management. Sharada et al. (2025) introduced a GeoAgriGuard system by combining remote-sensing technologies and smart AI models to mitigate the risk of pests and disease outbreaks. A hybrid deep learning ensemble architecture was developed by integrating ResNet, DenseNet, the Transformer, and autoencoders for processing the collected data, and the proposed method achieved better accuracy than other models. Zhu et al. (2023) introduced a pest detection and classification model using Raspberry Pi for managing disease and pests’ risks of lychee and longan fruits. Image-based data collected were incorporated into the VGG16 model, which was then deployed on a Raspberry Pi for the detection and classification of pests and diseases. Kathole et al. (2024) utilized IoT sensors for collecting pest image-based data, then used the YOLOv3 object detection model for pest detection. The images detected were fed into the CNN for feature extraction. These extracted features were given as input to the convolutional neural LSTM (CNLSTM) model. Sankar et al. (2025) developed an efficient pest classification system using IoT and DL approaches to facilitate accurate pest detection in smart farming. An IoT framework was employed to collect and transfer real-time pest image data to DL models. For pest detection and classification, YOLOv8 and DNN models were deployed, respectively. To reduce redundant computation, the cross-scale attention gradient pruning method was employed to increase focus on pest detection and identification. Kapetas et al. (2025b) utilized the SpyFly robotic trap for the continuous monitoring and detection of Tuta absoluta. For pest detection, a DL model, YOLOv8x, was deployed, and for predicting population outbreaks of Tuta absoluta, different ML techniques were applied. The autoregressive integrated moving average with exogenous variables (ARIMAX) model resulted in the best prediction performance compared to other competitive models. Zarboubi et al. (2026) proposed a cost-effective smart pest detection system by integrating the YOLOv10-Evo model on a smart pest monitoring trap equipped with a Raspberry Pi 4B.

4.1.1 Evaluation of pest detection models in terms of accuracy and field deployment

Many state-of-the-art models based on CNNs and standard YOLO architectures are computationally intensive and unsuitable for real-time deployment on resource-constrained devices in traditional agricultural environments. The trade-off between detection accuracy and computational efficiency often restricts their practical deployment, particularly in rural farming communities (Bellout et al., 2025). In pest management applications, attention-based models and Vision Transformer hybrids achieve the highest performance, resulting in accuracy ranging from 97-99% (Zeng et al., 2024), outperforming typical CNN architectures, which deliver accuracy ranging from 89-99% depending on the deployed architecture and dataset (Bhujel et al., 2022). YOLO variants remain competitive in the middle, with YOLOv8 and other improved variants achieving detection accuracies of 95-99% and mAP scores of 90.57-98.6%, respectively, while earlier YOLO versions deliver slightly lower results (Arunachalam and Mohana, 2025; Zeng et al., 2024) (Table 1). However, the critical distinction lies in comparing the practical applicability of these pest detection models. Attention-based models achieve the highest accuracy in controlled environments with ample computational resources, while YOLO variants are favored in real-time environments for delivering suitable inference speed and superior detection accuracy (Mo et al., 2025). Traditional CNNs demonstrate limited effectiveness in detecting multiple small pests simultaneously or in complex agricultural backgrounds (Mabizela et al., 2025). The integration of an attention mechanism into the YOLO architecture to create a hybrid architecture is paving the way for a suitable balance between accuracy improvements and maintained computational efficiency (Lu et al., 2025). These techniques are increasingly favorable for smart pest management systems that demand high accuracy and operational feasibility simultaneously (Figure 5).

Table 1

Pest nameMethodologyResultsAuthors
Armyworm, aphids, grasshopper, sawfly, mosquito, stem borer, beetles, bollworm, mitesDHBPDAccuracy: 0.936Gokeda and Yalavarthi (2024)
Public dataset-IP102Optimized YOLOv3F1-score: 84%
Accuracy: 96%
Prasath and Akila (2023)
Crop leaf pestsVGG16, IoT-UAVAccuracy: 96.58%Saranya et al. (2023)
Palm weevils, wasps, shield bugs, grasshoppersEnhanced YOLOv5s modelAverage Precision:
96%
Recall: 93%
Mean average precision (mAP): 95%
Khan et al. (2024)
Paddy pestsVGG19, ensemble classifierAccuracy: 98.91, 99.14%
Precision: 98.12%
Rahman et al. (2024)
Oriental fruit fliesCNN classifierAccuracy: 97.5%
Precision: 92.82%
Recall: 98.92%
F1-score: 95%
Specificity: 95.90%
FNR: 5.88%
Ahmed et al. (2024)
Tuta absolutaDL-based modelsAccuracy (mAP50): 70.2%
Detection confidence level: 87%
Christakakis et al. (2024)
Pest and disease dataHybrid deep learning ensemble modelAccuracy: 97.81%
Recall: 95.83%
Precision: 96.72%
F1-score: 96.27%
Sharada et al. (2025)
Pest: Hypitima longanae Yang et Chen
Disease: Sooty blotch
VGG19-Raspberry PiAccuracy: 94.9%Zhu et al. (2023)
Tuta absolutaYOLOv8x, ARIMAXAccuracy: 97.6%
MSE: 36.69, 309.67
Kapetas et al. (2025b)
Codling mothsYOLOv10-Evo, smart trapmAP50-95: 98.11%
Average confidence level: 89%
Precision: 100%
Recall: 100%
Zarboubi et al. (2026)
Agricultural pest image datasetHCFormer (CNN + Vision Transformer)Accuracy: 98.17%, Recall: 91.98%
mAP: 90.57%
Zeng et al. (2024)

Application of AI and IoT fusion in pest detection and classification.

Figure 5

4.2 Fusing AI and IoT for earlier prediction of pests’ populations

To enhance the efficacy of IPM, researchers are focusing on the development of pest prediction models by investigating the relationship between pest population trends and meteorological variables such as temperature, rainfall, humidity, and light intensity. They are then incorporating these valuable insights into AI-based algorithms to develop smart pest detection and prediction models. Different sensors are used to gather climatic data, including temperature sensors, humidity sensors, light sensors, and pH sensors (Dai et al., 2024). To predict the population occurrence of litchi sting bugs (Tessaratoma papillosa), Chen et al. (2022) used a long short-term memory (LSTM) model for predicting weather parameters. They then deployed different ML algorithms, including K-nearest neighbors (KNN), support vector machine (SVM), and random forest regressor (RFR), to forecast the correlation between meteorological variables and pest infestations. Weather parameter data were collected through self-designed IoT systems. This study highlights the importance of meteorological variables on the occurrence of pests, as well as the synergy of AI and IoT solutions in pest prediction. To detect and predict the occurrence of black aphids in cucumber cultivation, Kapetas et al. (2025a) utilized different traps, weather sensors, and mobile phone cameras for collecting weather parameter and pest phenological stage data. The gathered image-based data were incorporated into the YOLOv10 model for pest detection. For pest population prediction, meteorological parameter data were further incorporated into different ML algorithms, including RF, XGBoost, LSTM, as well as statistical models, i.e., autoregressive integrated moving average (ARIMA), ARIMAX, and seasonal autoregressive integrated moving average with exogenous variables (SARIMAX). Chen et al. (2025) predicted the population abundance of Tuta absoluta (Meyrick) in relation to different environmental parameters using various ML algorithms. Palani et al. (2023) utilized sensor-based collected meteorological data, satellite imagery, pest historical data, and disease incidence records to develop a predictive model to control pests’ devastating impacts. Field-collected datasets were incorporated into DL models, including RNNs and CNNs for time series data and satellite image analysis, respectively. Alongside DL techniques, different ML approaches were also deployed for detecting diseases and pests. Mohanraj et al. (2025) designed a blockchain-enabled secure early pest warning system for managing pests in bamboo rice farming by integrating LSTM, IoT sensors, and an RF classifier (Table 2). Saleem et al. (2023) deployed IoT sensors’ generated weather parameter data and a DL model for developing an early pest prediction system by utilizing a dataset from 2018 to 2022. Marković et al. (2021) combined an IoT framework and ML algorithms to predict the population occurrence of Helicoverpa armigera by utilizing trap-based collected pest images and meteorological variable data. Rustia et al. (2022) combined deep learning techniques and edge computing devices to facilitate farm managers in managing different species of whiteflies and thrips in tomato cultivation in greenhouses. The proposed architecture, “Intelligent and Integrated Pest and Disease Management (I2PDM)”, demonstrated good results, resulting in an annual reduction in certain pests’ count to 50.7% on study sites. Nyakuri et al. (2025) introduced a hybrid model by combining explainable AI (XAI), a lightweight CNN model, and fuzzy logic (FL) for detecting the phenological stages of fall armyworm and their outbreak prediction based on weather parameters in maize crops. Chen et al. (2020) utilized mature IoT technologies (drones for capturing images and sensors for collecting environmental parameter data) and DL models, YOLOv3, to attain the locality of Tessaratoma papillosa, and LSTM to predict the population dynamics of pests, as precise pest positioning can minimize toxic pesticide applications, resulting in reduced soil damage. The proposed system can inform farmers about the extent and locations of pests by providing actionable insights for effective pest management. Onyejegbu (2023) introduced a pest and disease detection system by utilizing the potential of AI and IoT to monitor sweet potato pests, most prominently Cylas puncticollis. An IoT device was used to acquire images of pests from pheromone-based sticky traps and then send these images to a cloud computing platform where an AI model is implemented for predicting pest outbreaks. To monitor key pests of the tomato crop, a portable Pest Counting Device equipped with the YOLOv8 deep learning model on a Raspberry Pi 4B, coupled with the Firebase IoT platform, was used to facilitate instant surveillance of pheromone traps. This synergistic integration of DL and IoT technologies enables farmers to make informed decisions and optimize pest control (Zarboubi et al., 2024a).

Table 2

Pest nameProposed methodologyEvaluation metricsAuthors
Litchi sting bugsLSTM, SVM, KNN, and RFRR2 value: 96%
Accuracy: 85%
Chen et al. (2022)
Black aphidsYOLOv10, ARIMAXAccuracy: 89.1%
MSE: 75.61
Kapetas et al. (2025a)
Tuta absoluta (Meyrick)RFRRMSE: 1.4587Chen et al. (2025)
Bamboo rice pestsLSTM, RF ClassifierAccuracy: 98%
Precision: 92%
Yield improvement: 30%
Pesticides reduction: 25%
Mohanraj et al. (2025)
Whitefly (Bemisia tabaci)DNN modelAccuracy: 94%Saleem et al. (2023)
Helicoverpa armigeraML algorithmsAccuracy: 86.3%Marković et al. (2021)
Pepper pestDigital Twin modelAccuracy: 88.01, 85.73%Dai et al. (2024)
Thrips, whitefliesI2PDMAnnual reduction in pest count: 50.7%Rustia et al. (2022)
Fall armywormHybrid model architectureAccuracy: 98.6%
Recall: 98.6%
F1 Score: 98.5%
Nyakuri et al. (2025)
Lychee giant stink bugAIoT-based pest location detection and population prediction modelAccuracy: 90%Chen et al. (2020)

Applications of AI and IoT in early prediction of pests’ outbreaks.

5 Harnessing cutting-edge solutions in boring pest management

5.1 Traditional methods of boring pests’ management

Cultural control (landscape and crop management) practices are the commonly used traditional strategies for managing stalk and stem borers. These include destroying alternate and volunteer plants, field sanitation, crop rotation, mixed/intercropping, and alternating planting dates to minimize peak oviposition/moth activity. In the case of tree and wood systems, practices such as removing dying or dead plant branches, avoiding pruning injuries, and lessening the chances of providing shelter are essential. Surveys and field investigations have demonstrated that sanitation and mixed cropping significantly reduce stem borer occurrence and damage, while landscape-driven diversification of plants helps to avoid pest outbreaks (Reagan and Mulcahy, 2019).

Physical/mechanical controls involve the destruction and physical removal of infested plant material (such as cutting out infested branches or stems), traps, sticky barriers, utilization of stem bands, hand-picking of larger larvae, and the use of light or pheromone traps for adult borers. For structural and timber pest infestations, options such as solarization of wood, heat treatment, localized drilling, freeing, and extraction of pest galleries have historically been applied. Likewise, removal and quarantine of infested wood or logs are recommended methods to prevent pest spread. Early detection, as well as physical removal of pests, is crucial for reducing the wood-borer population growth (Li et al., 2024b).

Botanical/household remedies like extracted botanicals (pyrethrum extract, plant sap/tobacco rinses, azadirachtin/neem preparations, diatomaceous earth, or powdered ash) and oil or crude aqueous extracts have a long history of use against borers. Botanical compounds (pyrethrins, azadirachtin, and essential oils) have the potential to reduce oviposition, larval survival, and overall feeding, though their efficacy varies by timing, formulation, and pest species. Botanical-based products are sustainable and promising alternatives when properly integrated with biological and cultural measures (Civolani et al., 2025; Mahanta et al., 2025).

Biological control by natural enemies (predators, parasitoids, and entomopathogenic microbes) is an effective biological management strategy. Biological control is encouraged by providing refuges, avoiding broad-spectrum pesticides, and preserving non-crop vegetation for natural enemies. Manipulation of entomopathogenic fungi, conservation of different parasitoid species, and local releases or sprays of baculoviruses have the potential to reduce boring pest populations. However, their effectiveness is context-dependent and can provide benefits through proper integration with mechanical or cultural interventions (Mahanta et al., 2025). Entomopathogenic fungi are also very effective against borer pests, such as avocado borers and wood-boring/ambrosia beetles. These fungi are effective against both larval and adult stages of pests, and spores can be applied in galleries, on tree surfaces, or as wood/trunk treatments or into bait traps. The field efficacy of these fungi depends on factors such as temperature, application timing, formulation, and fungus strains (Shafique et al., 2025).

Structural timber/storage practices are used for finished or stored wood products, which include non-chemical traditional methods such as ventilation, stacking, regular inspection, inert dust utilization, botanical fumigants, cleaning, and drying that are used in enclosed stores. For expensive heritage/museum timber, different physical treatments like freezing, anoxia, controlled heat, and target-specific removal are chosen in comparison with chemical fumigation, while the inspection and quarantine of received wood are very crucial. Low-chemical-based options are increasingly important to address safety and resistance concerns.

Integrated pest management (IPM) is the ecology-driven integration of traditional tactics like cultural, mechanical, physical, biological, and chemical methods into IPM packages with the aim of reducing insect pests while preventing insecticide resistance, safeguarding natural enemies, and reducing environmental effects. These IPM packages depend upon pest monitoring and pest surveillance and act when the pest population exceeds already researched economic threshold levels. The IPM tactics include manual removal, pruning of insect-infested tree branches, need-derived extract dips, sanitation, wound spot treatment, conservation efforts for predators and parasitoids, and changed planting dates for orchards. The local adoption of combined pest control tactics along with microbial or targeted biological control creates a sustainable balance among efficacy, environmental safety, and management costs (Shafique et al., 2025) (Figure 6).

Figure 6

5.2 Applications of IoT in boring pest management

Effective management of trunk-boring pests’ populations necessitates the development of smart solutions that are accurate, automated, readily available, user-friendly, cost-effective, and implemented on a large scale (Çimen et al., 2025). Due to remarkable progress in sensor technologies and a reduction in sensor costs, monitoring trunk-boring pest populations is now possible by deploying acoustic detection technologies. Acoustic detection technologies have the potential to collect, record, and process signals of pest-boring activities in tree trunks. Different IoT sensors and data acquisition modules can be deployed in this domain for prompt warnings of pest infestations in their early stages (Liu et al., 2022). Another milestone in the field of electronic sensing is the provision of pest detection systems through the placement of biosensors and electronic nose arrays within the crowns of trees in commercial orchards (Omer et al., 2024; Çimen et al., 2025). These sensors can be manufactured to detect pest-specific aggregation pheromones and transmit this information to DSS through an IoT framework. These valuable insights can provide rapid and precise information about significant infestations of pests at the individual tree level (Koubaa et al., 2020; Stevanoska et al., 2020). Gradual progress is being monitored in the development of smart bioacoustic sensors for managing trunk-boring pests by analyzing their sound and locomotion signals. For the monitoring of adult pests with vocal organs or pest larvae in a soundproof environment, microphones were deployed to record vibration signals due to their low cost and high sensitivity (Liu et al., 2022). To identify the stridulation signals of four bark beetles (Dendroctonus valens, D. armandi, Tomicus yunnanensis, and T. brevipilosus), Luo et al. (2011) utilized the SM4001 microphone (BSWA TECH Inc., Beijing, China) for signal recording. To record the feeding and movement activities of Sitophilus zeamais and Prostephanus truncatus, Njoroge et al. (2016) used a 0.5″ microphone (Model 378B02, PCB Piezotronics Inc., Depew, NY, USA). A drawback of microphones is that their performance can be easily disturbed by potential sounds from the external environment, including air sounds, bird chirping, etc., as they receive vibrations from the air instead of directly sensing them from the tree trunk (Liu et al., 2022). Another limitation of microphones is that they are not suitable for detecting boring vibrational signals in solid substrates such as wood and soil. With the development of sound monitors (AED-2000 L) and an accelerometer in 2000, the range of bioacoustic analysis expanded from stored grain pests to solid substrate pests (soil and wood-boring pests) (Jiang et al., 2022).

Another advancement in recording the vibrational signals of concealed pests is the development of piezoelectric sensor probes, as these probes are embedded in the tree trunk by their pickups after drilling into the trunk section. Herrick and Mankin (2012) deployed the AED-2000 amplifier (AEC Inc., Monrovia, CA, USA) along with a Model SP-1 L sensor-preamplifier module for recording the sounds of the red palm weevil. Bu et al. (2017) utilized the AED-2010 L (AEC Inc., Monrovia, CA, USA) along with a piezoelectric sensor SP-1 L and a portable acoustic emission amplifier for recording four types of acoustic behavior signals of Anoplophora chinensis and Anoplophora glabripennis. Jiang et al. (2022) deployed the AED-2010 L to detect the timing window of Semanotus bifasciatus Motschulsky (Coleoptera: Cerambycidae) peak boring activities, as this second-generation instrument is highly sensitive to detect slight vibrations in wood substrate. They also experimented with the differentiation of early larval instars’ boring signals from external environmental sound signals by employing stepwise regression. Potamitis et al. (2017) used an accelerometer as a remotely controlled device that records vibrational signals and transmits them to a central server at scheduled intervals to process these signals and facilitate the user in automatically inferring the larval infestation level of trees. The proposed IoT framework expanded the boring pest monitoring range from individual trees to large-scale areas. Mendel et al. (2024) deployed Agrint IoTree seismic sensors to control the gravest threats of the red palm weevil in commercial palm plantations. Data collected through sensors are further translated to show the health status of palm trees and the adequate quantity of insecticides applied based on the severity of infestation. Maruthadurai et al. (2022) monitored the acoustic behavior of the stem and root borer (Neoplocaederus ferrugineus) using the TreeVibes device, as this acoustic detection device contains a piezoelectric probe, transducer, and microprocessor. The main purpose of the study was to characterize the activity signals of larval instars, and the TreeVibes device resulted in good accuracy, i.e., 91%, while detecting the infested cashew trees under field conditions. Omer et al. (2024) evaluated the efficacy of different insecticides such as Thiamethoxam, Fipronil, and Imidacloprid against red palm weevil infestation in date palm cultivation by employing the TreeVibes device over the post-treatment course of 10 months. Insecticides were injected directly into the naturally infested target trees, and by using an acoustic sensor for recording larval mean burst rates, the trees’ recovery against different insecticides was assessed. To detect bark beetle infestation, Hüttnerová and Surový (2024) evaluated three electronic noses: the Handheld VOC Detector Tiger, the Miniature Bosch sensor device, and environmental sensors. The Miniature Bosch sensor device resulted in better detection accuracy. Martínez et al. (2023) utilized microelectromechanical sensor (MEMS) microphone arrays to detect and localize the larvae of Hylotrupes bajulus L. in wood pieces of Pinus sylvestris L. based on acoustic signals produced during larval activity. Banlawe and Dela Cruz (2020) deployed different sensors (Electret microphone, MEMS microphone, and piezoelectric transducer) for capturing the acoustic emission of the mango palm weevil (MPW), a highly quarantined and dangerous pest of mango trees and fruits. The MEMS microphone resulted in better sensitivity to MPW activity and was more resistant to external noise. Bittner et al. (2018) deployed the PCB 352C15 high-frequency ceramic shear accelerometer to study the feeding vibrational signals of Callosobruchus maculatus on cowpea seeds. Hetzroni et al. (2016) used a piezoelectric sensor for the detection of acoustic sounds of the red palm weevil in Canary palm and young date trees. Mankin et al. (2018) recorded the vibrations of Mallodon dasystomus larvae by deploying the AED-2010 amplifier system in avocado trees. Pest activity was recognized by analyzing impulses in the recordings. Sutanto et al. (2023) utilized the TreeVibes device to assess the efficacy of fungal treatment applied to date palm trees to suppress the larval infestation of RPW. This acoustic sensing device was inserted on the date palm tree trunk to sense the boring activity signals of RPW (Table 3).

Table 3

Pest nameIoT sensorEvaluation metricsAuthors
Bark beetlesSM4001 microphoneAccuracy: 98.14%Luo et al. (2011)
Sitophilus zeamais and Prostephanus truncatus0.5″ microphoneMean rates of bursts: Broadband: 24 MidF2:358Njoroge et al. (2016)
Red palm weevilAED-2000 amplifierLarval burst detection: >80%Herrick and Mankin (2012)
Anoplophora chinensis and Anoplophora glabripennisAED-2010 L, piezoelectric sensor SP-1 LFeeding sound proportion: 34%
Feeding sound impulse duration: 24.28, 25.96 ms
Bu et al. (2017)
Callosobruchus maculatusPCB 352C15 high-frequency ceramic shear accelerometerRange of peak event emission frequencies: 16.4 -26.5 kHzBittner et al. (2018)
Mallodon dasystomusAED-2010 amplifier systemInfestation detection on 11 trees
Infested: 4
Uninfested: 6
Misclassified: 1
Mankin et al. (2018)
MPWMEMS microphoneHigh sensitivityBanlawe and Dela Cruz (2020)
Xylotrechus chinensis, Rhynchophorus ferrugineus OlivierAccelerometerHigh accuracy in detecting borersPotamitis et al. (2019)
Red palm weevil (Rhynchophorus ferrugineus)Agrint IoTree seismic sensorsMean (± SD) sensor value (post treatment): 12.6 ± 4.0Mendel et al. (2024)
Stem and root borer (Neoplocaederus ferrugineus)TreeVibesDetection Accuracy: 91%Maruthadurai et al. (2022)
Red palm weevilTreeVibesRPW impulse burst rate (post-treatment): dropped to zeroOmer et al. (2024)
Bark beetleMiniature Bosch sensor deviceAccuracy: 95%Hüttnerová and Surový (2024)
Hylotrupes bajulus LMEMS microphone arrayLocalization of larvae and internal trajectories: MediumMartínez et al. (2023)
Red palm weevilTreeVibesRPW impulse burst rate (post-treatment): dropped to zeroSutanto et al. (2023)

Utilization of IoT sensors in boring pests’ management.

5.2.1 Critical overview of IoT sensors in trunk-boring pest management in terms of accuracy, cost, and field deployment

Piezoelectric sensors emerged as a superior choice for boring pest detection, delivering remarkable accuracy in identifying various trunk-boring pest species, such as EAB, while demonstrating strong noise robustness in real forest datasets (Jiang et al., 2024a; Yin et al., 2025). Microphone-based acoustic systems often fail in real-time forest settings due to external environmental noise interference, resulting in significant performance degradation (Bedoya et al., 2022; Bob et al., 2025). Seismic sensors are comparatively expensive, making them unsuitable for large-scale deployments compared to alternative piezoelectric sensors. Among other critical constraints, single-channel signal loss at distance is of utmost importance. Alternatively, multi-channel piezoelectric architecture addresses this drawback, but in parallel, their cost increases substantially, requiring sophisticated deep learning noise reduction algorithms (e.g., achieving 17.45 dB signal-to-noise improvement) to maintain field accuracy (Zhang et al., 2024).

The TreeVibes device offers autonomous, remotely controlled monitoring, recording, and transmitting internal tree vibrational signals 24/7 to a cloud server without human intervention. It addresses the typical limitations of piezoelectric sensors, which depend on scheduled human inspections and often fail to detect pest activity when larvae are inactive during the inspection window (Potamitis et al., 2019). DL techniques are integrated into the TreeVibes device to automatically process incoming vibrational signals and infer infestation status, eliminating the need for manual headphone-based listening in piezoelectric sensors and thereby reducing observer fatigue and subjective interpretation errors (Rigakis et al., 2021).

5.3 Fusing AI models and IoT technologies for managing boring pests

Another progress in managing boring pest infestation is the synergy of AI, IoT, computer vision, and remote sensing technologies to monitor pest populations in real time and the development of DSS for better insect pest control. Recent advancements and developments in cutting-edge technologies are providing innovative solutions in boring pest management by deploying smart algorithms for pest detection, classification, monitoring, and management (Hessane et al., 2024). DL algorithms are frequently deployed in bioacoustics analysis due to enhanced developments in their architecture. Boring pest vibrational signals provide important bioacoustics information for pest control, biodiversity surveillance, and ecological research (Yin et al., 2025).

In the current era, researchers are deploying IoT sensors to collect tree vibration data, then incorporating these data into DL networks to identify early pest infestations during the larval stage, minimizing the need for manual forest inspections. Optimizations in neural network structures, especially advanced speech recognition technologies, are paving the way for pest scouting and detection. In comparison with early sound recording devices such as microphones, piezoelectric sensors are more robust and sensitive for recording vibrational signals generated by larval activity inside the tree trunk. The integration of IoT technologies and AI algorithms is facilitating pest management by recording vibrational signals through sensors, converting them into an audio format, and then inputting these signals into DL networks for pest detection and classification to forewarn forest managers for timely interventions (Jiang et al., 2024b; Kahl et al., 2018) (Figure 7). Two major phases are involved in the recognition of pest vibrational signals: firstly, auditory identification and evaluation of pest activity vibrational signals in the time–frequency domain. Skilled detection personnel are required to accomplish this task. However, in environments with significant surrounding noise, recognizing boring activity through short-duration segments can lead to misinterpretations. In the second phase, boring pest activity vibrational signals are collected and manipulated through IoT sensors and AI algorithms, respectively (Du, 2019; Jiang et al., 2022; Jiang et al., 2024a). ML models are among the top automated recognition approaches, capable of processing large datasets to extract meaningful patterns for classification and prediction tasks (Jiang et al., 2024a).

Figure 7

Several studies have been conducted to detect and classify concealed boring pests and their accompanying infestations by combining the potential benefits of AI and IoT technologies. Shi et al. (2023) utilized a piezoelectric sensor probe to collect boring signals of EAB larvae and environmental noise, then proposed a DL-based enhancement model called dual-vibration enhancement network (DVEN) architecture to strengthen the boring vibrational signals both in the time domain and frequency domain. After signal enhancement, several classification models were deployed to differentiate the EAB larvae activity signals from surrounding noisy signals. The proposed architecture enhanced the signal-to-noise ratio (SNR) of vibrational signals and took less time in enhancing a 5 s segment. A substantial improvement was recorded in the recognition accuracy of classification models after signal enhancement through the DVEN architecture. Muhammad Nadeem et al. (2022) deployed ML algorithms and IoT sensors for predicting the stem borer population based on above or below the economic threshold level (ETL) in sugarcane crops. Esmail Karar et al. (2021) deployed an IoT framework and transfer learning classifier InceptionResNet-V2 for the early detection of red palm weevil (RPW) larval activity based on date palm tree infestations. The proposed architecture resulted in better detection accuracy and standard deviation (SD) values. Latif et al. (2022) deployed different IoT sensors for recording weather parameters, e.g., rainfall, temperature, and relative humidity values, and then utilized an AI model to predict yellow stem borer population and produced early warnings for farmers about pests’ expected attacks. Jiang et al. (2024a) utilized piezoelectric ceramic sensors for recording drilling vibrational signals of EAB and Holcocerus insularis Staudinger pests and then proposed a novel CNN-based model called residual mixed domain attention module network (RMAMNet) to detect and classify both pests based on their boring activity. Li et al. (2025) developed the EABNet model for the early detection of EAB larvae, highlighting the importance of DL algorithms in bioacoustics analysis in forestry settings. A piezoelectric sensor probe and a data acquisition device were deployed to record vibrational signals of boring pest activity. Yazgac et al. (2016) deployed a microphone for recording Sunn pest sounds and used different K-nearest neighbors (KNN) models to automatically detect different species (Aelia rostrata, Aelia acuminata, Eurygaster austriaca, and Eurygaster integriceps) of Sunn pests through sound processing.

Liu et al. (2022) emphasized the efficacy of AI-based denoising models for the better recognition of Semanotus bifasciatus larval feeding sounds from noisy environments. A piezoelectric sensor probe SP-1 L and a data acquisition module were utilized to record Semanotus bifasciatus sounds. In that study, they designed both a frequency-domain denoising network (FDD-Net) and time-domain denoising network (TDD-Net). Different metrics such as SNR, segment SNR (SegSNR), and log spectral distance (LSD) were compared to evaluate the denoising effects of the proposed denoising model. Yin et al. (2025) proposed the VibroEABNet model by integrating novel denoising and recognition modules for detecting EAB-larvae activity. Other notable features of the model’s efficiency are minimum inference time and compact model size, proving it to be a suitable tool in resource-constrained environments. Zhang et al. (2023) used piezoelectric sensors for collecting EAB larvae activity signals and corresponding natural sounds data from forest habitats. A DL-based time-domain model with conformer called time-domain conformer-based enhancement network for vibration (T-CEVN) was proposed to enhance the vibrational signals of pests’ boring activity to detect EAB presence in a real forest environment. Both types of collected data are mixed at various SNRs to mimic the realistically collected dataset, and the SNR of improved vibrational signals reached up to 17.84 dB after being enhanced by the T-CEVN model. When the enhanced data were given to the classification model VGG16, it resulted in improved accuracy. Zhang et al. (2024) proposed dual-branch time-frequency multi-dilated dense network (DBMDNet) for environmental noise suppression while detecting EAB larvae populations through sensors. In DBMDNet, two denoising modules were decoupled, such as a magnitude masking decoder and a complex spectral decoder for coarse denoising and magnitude repair, respectively, and three different multi-dilated dense blocks to optimize global frequency modeling. In comparison with competitive models, DBMDNet resulted in improved log-likelihood ratio (LLR) and SNR values, enhancing the recognition accuracy of the larvae detection model. Li et al. (2024b) proposed a novel DL-based BorerNet model with an attention module to detect and classify the small carpenter moth (SCM) and EAB larvae with limited boring vibrational signals data collected through an acoustic sensor. Different preprocessing techniques, e.g., clipping and segmentation, were applied to extract Mel-frequency cepstral coefficients (MFCCs), and the resulting MFCCs and noisy environmental signals were then input to the BorerNet model to identify the boring activity of both species (Table 4).

Table 4

Pest nameProposed modelResultsAuthors
Stem borerNaïve Bayes modelAccuracy: 83%Muhammad Nadeem et al. (2022)
EABDVEN architectureSNR: 18.73db
Segment enhancement time: 0.46 s
Shi et al. (2023)
Red palm weevilEnhanced InceptionResNet-V2Accuracy: 94.53%
SD: ±1.69
Esmail Karar et al. (2021)
Yellow stem borerAccuracy: 85.6%Latif et al. (2022)
EAB, Holcocerus insularis StaudingerRMAMNetAccuracy: 95.34%
F1 score: 0.95
Jiang et al. (2024a)
EABEABNetAccuracy: 82.89, 79.79%Li et al. (2025)
Sunn pestsKNN variantsAccuracy: 93.6%Yazgac et al. (2016)
Juniper bark beetle (Semanotus bifasciatus)TDD-Net, FDD-NetSNR: 17.53,11.10 db
SegSNR: 18.59, 12.04
LSD: 0.85, 0.84
Liu et al. (2022)
EABVibroEABNetAccuracy:98.98, 97.5%
Inference time: 26 ms
Model size: 8.43 MB
Yin et al. (2025)
EABT-CEVN modelSNR: 17.84 dBZhang et al. (2023)
EABDBMDNetSNR: 17.45 dB
LLR: 0.14
Zhang et al. (2024)
EAB, SCMBorerNetAccuracy: 96.67%Li et al. (2024b)
EABVibDenoiserSNR: 18.57Shi et al. (2022)
Semanotus bifasciatusSR, PLSRR2: 0.71,0.97
RMSE: 0.42; 61.96
RPD: 3.38; 28.87
Jiang et al. (2022)
Parlatoria Date ScaleStacking-based weighted ensemble modelR2-nymphs: 77%
R2-female adults: 88%
R2-population: 90%
Hessane et al. (2024)
Longhorn beetleElectronic board with piezoelectric sensorMAE: 1.01
RMSE:1.91
Potamitis and Rigakis (2024)
Mango pulp weevilSVMAccuracy: 89.81%Banlawe and dela Cruz (2023)
EABMelSPPNeTAccuracy: 99.97%
IS: 0.9993
Jiang et al. (2024b)
Wood boring pestsManetAccuracy: 96.136%Min et al. (2025)
RPWAIoT-based systemControlled infestation rate: <5%Çimen et al. (2025)
EABMultiSAMSXu et al. (2023)
Semanotus bifasciatusDalPNetAccuracy: 99.5, 98.55%
CS: 2.0105
RAUC: 0.2923
Liu et al. (2023)
Longhorn beetleMobileNet
Xception
Mean accuracy/SD
Xception: 94.16/0.99
MobileNet: 93.84/1.16
Rigakis et al. (2021)
Streltzoviella insularis, EABT-CENV modelInference time: 418.43 s
SNR: 14.35
Li et al. (2024b)

Synergizing AI and IoT for effective boring pests’ management.

Shi et al. (2023) combined IoT acoustic sensors and DL speech enhancement technologies to detect EAB larvae existence by distinguishing the EAB larval activity sounds from external environmental noisy sounds. A VibDenoiser model was developed to enhance the vibrational signals of EAB larvae activity. After signal enhancement, when the data were incorporated into the classification model, a significant improvement was observed in their detection accuracy. Jiang et al. (2022) proposed an acoustic model for the early prediction of Semanotus bifasciatus attacks by identifying the timing window of peak larval activity and respective acoustic parameters. They deployed stepwise regression (SR) for larval detection and partial least squares regression (PLSR) for the prediction of larval instars and estimation of pest population size. Evaluation metrics used for the models’ performance evaluation included the coefficient of determination (R2), root mean squared error (RMSE), and relative percent deviation (RPD). To improve the efficacy of pest management in date palm cultivation, Hessane et al. (2024) developed an AI-based PDS population prediction model in correlation with climatic parameters, as this is a threatening pest of date palm trees. After data enhancement, the stacking-based weighted ensemble method resulted in better prediction performance while predicting PDS adult females, nymphs, and total population abundance. Potamitis and Rigakis (2024) employed different signal processing techniques to differentiate the borer activity (fracturing of wood fibers) sounds from external environmental sounds based on the criteria of short-duration repetitive impulses originating from larvae feeding and locomotion activities.

Banlawe and dela Cruz (2023) utilized MEMS sensors for recording the acoustic data of mango pulp weevil activities, such as mating, locomotion, and resting, inside a soundproof chamber to reduce the impact of external noise. The MFCC was deployed to extract meaningful patterns from recorded data and to train the MATLAB-based ML model. Jiang et al. (2024b) deployed piezoelectric ceramic sensors to collect the feeding vibrational signals of EAB larvae. They developed a case-based self-explainable model, MelSPPNeT, to extract more representative sound clips for better model recognition ability. Min et al. (2025) utilized previously collected datasets such as ESC-50 and TreeVibes, and then extracted two audio features, MFCC and linear FCC (LFCC), from the dataset for the training of the detection model. A light CNN-based mobile acoustic network (MAnet) was developed to accurately detect wood-boring pests’ boring activities. Çimen et al. (2025) deployed the seismic sensor system to detect the RPW presence in palm trees to control the RPW-larvae infestations. Seismic sensors detect the presence of larvae by differentiating between normal tree vibrations and infected tree vibrations. Targeted insecticide treatment was applied upon confirmation of infestation severity, and the authors concluded that it could be an effective approach for protecting palm trees.

Xu et al. (2023) developed a time-domain-based multi-channel attention mechanism-based signal separation system (MultiSAMS) to differentiate between the boring sounds of EAB larvae and external environmental sounds. Liu et al. (2023) used a piezoelectric SP-1 L probe for recording activity signals of Semanotus bifasciatus and for larvae detection. They developed dynamic acoustic larvae prototype network (DalPNet), which uses a prototype for more effective decision-making to enhance reliability and improve the model’s explainability through dynamic feature patch computation techniques. To evaluate the model interpretability, relative area under the curve (RAUC) and cumulative slope (CS) of the accuracy change curve evaluation metrics were used. Rigakis et al. (2021) utilized a remotely controlled TreeVibes device for the collection of wood-boring pest activity signals and several DL approaches to infer the infestation status of trees by analyzing the feeding and locomotion sound signals of larvae coming from probed trees to cloud servers. Li et al. (2024a) introduced a lightweight signal enhancement model, T-CENV, optimized through two lightweight optimization algorithms to effectively run on resource-constrained devices to separate the boring vibrational signals from external environmental noisy signals while monitoring the wood-boring pest larvae in real time.

5.4 Applications of remote sensing (RS) technology in boring pest management

A strategic shift is observed in pest detection, identification, and management in recent decades from traditional methods to more robust approaches by leveraging advanced information-collecting techniques and tools such as Global Positioning System (GPS), Geographic Information Systems (GIS), and remote sensors (Figure 8). RS is an innovative paradigm for collecting information about an object, phenomenon, or area by analyzing the data acquired by a device without physically touching the area or object under investigation (Lillesand et al., 2015; Aziz et al., 2025). Generally, the RS phenomenon contains several phases playing crucial roles in data acquisition and utilization. RS encompasses various technologies, e.g., GIS, satellites, IoT, DSS, cloud computing, and UAV/UAS. Properties of electromagnetic radiation (EMR) mainly impact the RS process as EMR is used for data collection and interpretation. RS is classified into two main types based on the source of EMR: active RS (which has its own source of radiation) and passive RS (which does not have its own source of radiation but rather depends on an external source for object illumination). Active RS works day and night, such as Radio Detection and Ranging (RADAR), while passive RS works only during the daytime due to the lack of radiation sources at night (Jong and van der Meer, 2006; Aziz et al., 2025).

Figure 8

RS approach offers powerful tools for the surveillance of various stages of pest infestations in a continuous and timely manner in large-scale forests. The integration of RS and expert knowledge can assist in the accurate assessment of infestation damage as well as its occurrence prediction, consequently providing more scientific foundations for sustainable forest management. The detection of wood-boring pests in the earlier stages of infestation is obviously challenging due to the lack of apparent tree symptoms, and it demands advanced RS technologies. Significant progress has been made in satellites’ spatial, spectral, and temporal resolutions, leading to the massive availability of multiscale, multisource, and multitemporal remotely sensed data. Combining UAVs with advanced lightweight sensors such as RADAR, hyperspectral imagers, and Light Detection and Ranging (LiDAR) is paving the way for the collection of multidimensional, convenient, and high-quality data that can assist in the early detection of trunk-boring pests (Luo et al., 2023).

Several studies have been conducted on boring pest detection leveraging RS technologies. To detect the precise stem-boring damage caused by pine shoot beetles, Liu et al. (2025) collected hyperspectral data with a GaiaSky-Mini2 hyperspectral imager mounted on a DJI M600 Pro UAV. They evaluated the spectral features of healthy pine trees, stem-boring infestation, and three levels of shoot-feeding infestation, and then used an RF model to detect the damage caused by pine shoot beetles. Red edge, blue edge, and chlorophyll-related indices were the most significant canopy spectral features for detecting pine shoot beetle infestation. Shao et al. (2024) developed a hyperspectral LiDAR (HSL) system based on an acoustic-optic tunable filter (AOTF) to investigate the echo intensity of tree trunks. For agile pest detection, the echo intensity ratio of the damaged part to the healthy part is taken; when the ratio is around one, an SVM classifier with reflectance is deployed to detect the infested area of bark. Georgieva et al. (2022) applied an integrated approach to detect bark beetle infestation in a pine plantation by combining data from the Information System of the Executive Forest Agency (ISEFA). Remote sensing data were collected through an eBee SQ-UAV equipped with a multispectral camera named Parrot Sequoia and subsequent terrestrial observations. To assess the damage caused by bark beetle infestation, four different Normalized Difference Vegetation Index (NDVI) categories from various infestation spots such as dead trees, living trees, grass and shrub vegetation, and stone and rock were obtained. Duarte et al. (2020b) demonstrated the accuracy of UAV-multispectral imagery for eucalyptus longhorned borers (ELB) detection and quantification in eucalyptus plantations. Otsu thresholding analysis was performed on five vegetation indices (VIs) derived from multispectral imagery to identify the spatial damage, and classification accuracy was then evaluated. After Treetops calculation by employing a local maxima filter of a sliding window algorithm, tree crown data were obtained through large-scale mean shift segmentation and then categorized with an RF classifier model. Data produced by the RF classifier was used to create forest density maps.

Arapostathi et al. (2023a) utilized RS technology for the timely identification of peach flatheaded root borer (PFRB) damage in pome fruit and stone fruit orchards. To capture aerial imagery from the respective orchards, a multispectral sensor mounted on a UAV was deployed. Different parameters, such as mean NDVI values, crown area, mean Normalized Difference Red Edge Index (NDRE) values, and two meteorological variables—mean relative humidity and mean temperature—were used to compare the ML algorithms. The mean NDVI values parameter proved to be the most critical for predicting certain pest infestations, and the XGBoost model delivered good accuracy. Kumbula et al. (2019) utilized the Sentinel-2 multispectral instrument to predict the damage caused by Coryphodema tristis in Eucalyptus plantations, as it is an emerging wood-boring pest in South Africa. Data collected through mass trapping and spectral imagery were fed to the Maxent algorithm, which delivered better results, highlighting the importance of remotely sensed data in facilitating effective decision-making while implementing forest pest management strategies. Nicoletti et al. (2024) studied the Pine Tortoise Scale (Toumeyella parvicornis) damage in stone pine forests using RS technologies, as it is a notorious pest of pine forests in Campania, Italy.

Illarionova et al. (2024) deployed open-source remote sensing images from Sentinel-2, supported by Google Earth application data, to identify the damaged forest infested by the bark beetle population through the ML algorithms YOLOv8 and HRNet. Zhou et al. (2021) introduced a novel method by integrating WV-2 images and Populus gansuensis tree physiological parameters to identify the three infestation stages of the Asian Longhorn Beetle on trees. Panopoulou et al. (2024) combined remote sensing technologies with ML algorithms for the early detection of the tiger longicorn beetle (Xylotrechus chinensis Chevrolat) infestation in mulberry trees in Greece. ML models were fitted on mean NDVI, mean EVI, mean NDRE, tree crown area, as well as temperature and pest emergence holes parameters, and the RF model outperformed its competitive models while modeling Xylotrechus chinensis infestations. Abdollahnejad et al. (2021) stated that RS plays a pivotal role in monitoring and understanding the dynamic forest structural changes triggered by pest incidences. To predict the potential outbreaks of European spruce bark beetles in healthy trees early, a possible correlation between RS data and tree health status was studied, concluding that spectral resolution is critical for the early detection of European spruce bark beetle infestation.

Shrestha et al. (2024) used multispectral image data collected through drones and employed ML algorithms to classify the trees as healthy and damaged by forest pests. Firstly, individual tree segmentation was performed using structure-from-motion (SfM) point clouds. Then, based on the multispectral reflectance of these points, healthy and damaged points were classified. A damage severity algorithm was applied to identify the damage percentage, damage severity level, and top-kill metrics. At the study site, 78.9% of damaged trees were identified as top kill. Wang et al. (2025) integrated UAV imagery and DL-based five semantic segmentation techniques for the development of an effective RS monitoring method to identify the pine trees infested by the Pine shoot beetle, as this pest poses a severe threat to pine in Yunnan Pine forests in China. Wang et al. (2022) studied the potential of thermal imagery to monitor the shoot damage ratio of the pine tree crowns caused by Tomicus spp. in Yunnan pine forests. Changes in canopy surface temperature can be monitored using thermal infrared imagery to assess forest health. Bozzini et al. (2025) utilized multispectral imagery captured by a high-resolution drone to detect the spruce trees infested by European spruce bark beetles and concluded that the following VIs, such as NDRE and Green NDVI (GNDVI), performed well while differentiating healthy and infected trees (Table 5). Nardi et al. (2023) combined satellite imagery and a volunteer-based participatory approach to assess both early and advanced stage infestation of bark beetles and highlighted the importance of this promising integration for understanding bark beetle abundance during emergency conditions.

Table 5

Pest nameRS equipmentDatasetEvaluation metricsAuthor
Pine shoot beetleMini2 hyperspectral imager, DJI M600 Pro UAVHyperspectral dataOverall accuracy: 57.38%
Highest accuracy: 86%
Liu et al. (2025)
Wood-boring pestsHSLHyperspectral dataAccuracy: 100%Shao et al. (2024)
Bark BeetleeBee SQ-UAV with Parrot SequoiaMultispectral imageryAverage plantation defoliation
Pinus sylvestris: 31.2–32.3%
Populus nigra: 47.4%
Georgieva et al. (2022)
ELBeBee SenseFly drone with Parrot Sequoia cameraMultispectral imageryRF-Accuracy:98.5%
NDVI-Kappa value: 0.96
Duarte et al. (2020b)
PFRBMultispectral RedEdge MX camera mounted on Mera UAVAerial imageryAccuracy: 85%Arapostathi et al. (2023a)
Coryphodema tristisSentinel-2 multispectral instrumentSpectral features and pest occurrence dataArea under the curve (AUC): 0.890Kumbula et al. (2019)
Pine tortoise scaleCopernicus Sentinel-2 satelliteMultispectral dataDeterminant parameter: differentiated NBR image (dNBR)Nicoletti et al. (2024)
Polygraphus proximus BlandfordSentinel-2 supported with Google EarthHigh-resolution satellite imagesF1-score: 84.56%
Accuracy: 81.26%
Illarionova et al. (2024)
Asain Longhorn BeetleYu Mavic2 professionalMultispectral WorldView-2 (WV-2) images and tree physiological parametersAccuracy: 86.2%
Kappa index value: 0.804
Zhou et al. (2021)
Tiger longicorn beetleMultispectral RedEdge MX camera mounted on Mera UAVMultispectral dataAccuracy: 0.86 Precision: 0.84 Recall: 0.81score: 0.82Panopoulou et al. (2024)
European spruce bark beetleDJI Phantom 4 Pro quadcopter with R-G-B sensorMulti-source time series RS dataSignificant impact of R-G-B bands derived from SPOT-6 (2018 and 2019) on health status prediction: 34-56%Abdollahnejad et al. (2021)
Western spruce budworm, Douglas-fir tussock moth, mountain pine beetle, Douglas-fir beetleDJI Matrice 210 V2 equipped with MX-RedEdge sensor and Zenmuse XT2 RGB sensorMultispectral dataOverall accuracy: 98.6%
High accuracy (top-kill on damaged trees): 91.8%
Shrestha et al. (2024)
Pine shoot beetleDJI Mavic 3 M UAVUAV-based imagerySegmentation precision:0.8341 F1 score: 0.8352 IoU: 0.7239
mIoU: 0.7185 Validation Accuracy: 0.9687
Wang et al. (2025)
Tomicus spp.UAV with thermal sensorThermal infrared and LiDAR dataR2 (single plant) = 0.7890
R2 (plot scale) = 0.5532
Wang et al. (2022)
European spruce bark beetlePhantom 4 RTK MS DJI equipped with an RGB camera and multispectral sensorMultispectral images, meteorological variables, and pest count dataSpot Detection: ~ 50%Bozzini et al. (2025)

Utilization of RS approaches in forest insect pests’ control.

5.4.1 Evaluating RS modalities for early detection of boring pest infestation

Among RS modalities, multispectral imagery offers the most scalable and practical RS features for the early detection of boring pests, achieving superior accuracy in identifying boring pest infestations through vegetation indices while remaining cost-effective with UAV deployment (Arapostathi et al., 2023b; Duarte et al., 2020a). Hyperspectral sensors offer superior spectral resolution for identifying subtle presymptomatic stress signals and detecting infestation 1-2 months before visual symptoms appear. However, due to their high cost and the high computational resource demand for large-scale orchard monitoring, their practical applications are limited (Näsi et al., 2015; Iordache et al., 2020). Thermal imagery is effective in damage assessment, but in cool weather conditions, thermal performance degrades and requires supplementary spectral data for reliable pest discrimination (Wang et al., 2024).

5.5 Integrating X-ray imaging and edge computing approaches to estimate wood-boring pest density

Conventional pest density detection models are heavily dependent on destructive tree sampling, which damages trees and often provides limited, incomplete data. Another novel approach to managing boring pest populations is the integration of edge computing technologies with X-ray imaging, which offers non-invasiveness and high-resolution imaging, making it a suitable alternative for non-destructive estimation of boring pest density. Despite its suitability, several challenges remain, such as capturing high-resolution X-ray images of different tree species and trunk diameters and using refined algorithms for precise pest recognition to compute reliable density estimates. Bi et al. (2025) introduced a novel boring pest density detection system by combining the potential benefits of DL models and X-ray imaging. The proposed system consists of an X-ray Exposure Optimizer (for capturing high-resolution X-ray images across different tree species and diameters to develop the boring pest dataset), a lightweight pest detection model, WPX-YOLO, and the X-BorerDensity Engine to convert pixel data into physical quantity and combine WPX-YOLO’s deliverables to accurately estimate the boring pest density. The proposed architecture can provide reliable estimates of pest density in living trees and assist in early predictions of forest pest infestations to support sustainable forest management practices.

6 Real-world deployment barriers in trunk-boring pest management and future directions

6.1 Challenges

Significant challenges exist in the implementation of these smart pest control systems, such as data availability and quality, huge computational upfront costs, technological accessibility, pest-larvae precise classification complexity, and regulatory frameworks for the implementation of AIoT-based pest management systems. Forest environments pose significant challenges for sensor reliability and longevity. Acoustic sensor measurement accuracy and reliability continuously degrade due to exposure to high humidity, fluctuating temperatures, and corrosion-prone conditions (Potamitis et al., 2019). Autonomous 24/7 field monitoring requires uninterrupted, sustained power supplies; however, IoT devices deployed in forests face severe energy consumption constraints. Wireless data transmission is the most power-intensive operation, creating a significant bottleneck and often leading to battery depletion within weeks (Borgianni et al., 2025).

The development of cost-effective, easy-to-deploy, and smart pest management systems aligned with regulatory frameworks can reduce pest damage and restore the ecological and socioeconomic services of forests, thereby supporting sustainable environmental development.

6.2 Future directions

In the recent technologically innovative era, the integration of multidisciplinary fields is paving the way for more robust research and developments in the agriculture sector. Multisource data derived from diverse sources, such as climatic parameters, satellite imagery, and host plant and pest phenological stages, are being studied to develop advanced pest management solutions. However, studying trunk-boring pest populations, either through acoustic data or satellite imagery, exhibits a significant research gap. Promising future directions in managing pest devastation involve combining data from diverse sources and further incorporating it into AI prediction models to better forecast pest detection and population levels. The integration of data collected through IoT sensors and RS approaches with AI predictive methods can model the potential risks of trunk-boring pest outbreaks to develop more focused pest management strategies. Developing more effective boring pest control systems requires the integration of multisource data to study the population trends of boring pests. Due to the cryptic lifestyle of trunk-boring pest larvae, it is quite challenging to gain early insights about expected pest outbreaks through field observations. However, their larval population could be detected by integrating pests’ boring vibrational signal data collected through acoustic sensors, as well as RS data capturing forest stress induced by hidden boring larvae activities, and incorporating these meaningful insights into AI-based predictive models. Early warnings can be issued to forest managers and other stakeholders by combining multisource data and AI-based predictive models into a centralized DSS, which can send alerts through mobile or web-based applications. Analyzing boring pest populations from diverse perspectives, such as climatic variable patterns, multispectral satellite imagery, and pests’ acoustic data, can assist in pest detection and provide a better understanding of the determinant abiotic and biotic parameters responsible for pest outbreaks. By considering the available meaningful insights about tree status, customized management interventions, such as targeted treatments, can be applied.

7 Conclusion

This article reviewed the potential benefits of AI, RS, IoT, and other innovative technologies for the timely detection, identification, prediction, and management of trunk-boring pests, which pose severe threats to forests and ornamental gardens globally. Increased developmental progress and cost reduction in sensing technologies are paving the way for the utilization of smart sensing technologies in managing the potential menace of trunk-boring pests to forest growth. The integration of AI-based smart algorithms with IoT acoustic sensors can effectively monitor the boring pest larvae population by analyzing the vibrational signals generated by boring pest activities inside the tree trunk and detecting the larvae infestation at an early growth stage. This represents a significant milestone in forest pest management, as it is difficult to detect boring pest larvae infestation due to their concealed lifestyle and delayed visual symptoms on trees. Bioacoustic analysis, collected through piezoelectric and seismic sensors and TreeVibes devices, can help model potential outbreaks of forest pests. Conventional methods of boring pest management are destructive, limited in accuracy, and highly dependent on forest experts, making them time-consuming and laborious procedures that may suffer from human error. Combining RS technologies with ML and DL algorithms can assist in monitoring dynamic structural changes in forests at a large scale to forewarn forest managers about expected pest outbreaks. The implementation of smart forest management systems has the potential to protect forests from disease and pest attacks, assisting in restoring the ecological functions of forests for a sustainable environment.

Statements

Author contributions

MB: Conceptualization, Investigation, Software, Visualization, Writing – original draft, Writing – review & editing. AS: Formal analysis, Funding acquisition, Project administration, Resources, Supervision, Validation, Writing – review & editing. SK: Conceptualization, Formal analysis, Investigation, Validation, Writing – original draft, Writing – review & editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This study was supported by the grant of project TSI-100921-2023-1 funded by the Secretary of State for Digitalization and Artificial Intelligence under the program CATEDRAS ENIA and the European Union-NextGeneration EU.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

  • 1

    AbdollahnejadA.PanagiotidisD.SurovýP.ModlingerR. (2021). Investigating the correlation between multisource remote sensing data for predicting potential spread of Ips typographus L. spots in healthy trees. Remote Sens.13:4953. doi: 10.3390/rs13234953

  • 2

    AhmedS.MarwatS. N. K.BrahimG. B.KhanW. U.KhanS.Al-FuqahaA.et al. (2024). IoT based intelligent pest management system for precision agriculture. Sci. Rep.14:31917. doi: 10.1038/s41598-024-83012-3,

  • 3

    AlbaneseA.NardelloM.BrunelliD. (2021). Automated pest detection with DNN on the edge for precision agriculture. IEEE J. Emerg. Sel. Top. Circuits Syst.11, 458–467. doi: 10.1109/JETCAS.2021.3101740

  • 4

    AlnuaimiA. F. A. H.AlbaldawiT. H. K. (2024). An overview of machine learning classification techniques. BIO Web Conf.97:00133. doi: 10.1051/bioconf/20249700133

  • 5

    ArapostathiE.PanopoulouC.AntonopoulosA.KatsilerosA.KarellasK.DimopoulosC.et al. (2023). Early detection of potential infestation by Capnodis tenebrionis (L.) (Coleoptera: Buprestidae), in stone and pome fruit orchards, using multispectral data from a UAV. Agron.14:20. doi: 10.3390/agronomy14010020

  • 6

    ArunachalamR.MohanaJ. (2025). GraphPest: a graph neural network framework for Pest detection in agricultural environment., in 2025 10th International Conference on Communication and Electronics Systems (ICCES), (Piscataway, NJ, New Jersey: IEEE), 2025–2031.

  • 7

    AsemaniM.AbdollaheiF.JabbariF. (2019). Understanding IoT platforms: towards a comprehensive definition and main characteristic description, in 2019 5th International Conference on Web Research (ICWR), (Piscataway, NJ, New Jersey: IEEE), 172–177.

  • 8

    AukemaJ. E.McCulloughD. G.Von HolleB.LiebholdA. M.BrittonK.FrankelS. J. (2010). Historical accumulation of nonindigenous forest pests in the continental United States. Bioscience60, 886–897. doi: 10.1525/bio.2010.60.11.5

  • 9

    AzizD.RafiqS.SainiP.AhadI.GonalB.RehmanS. A.et al. (2025). Remote sensing and artificial intelligence: revolutionizing pest management in agriculture. Front. Sustain. Food Syst.9:1551460. doi: 10.3389/fsufs.2025.1551460

  • 10

    BaiX.YeW.ZhouY.RuanX.WangJ.LiW.et al. (2023). Comparison between diffusive gradients in thin film technology (DGT) and traditional methods for prediction of plant available heavy metals in agricultural soil. J. Soils Sediments23, 1501–1510. doi: 10.1007/s11368-022-03410-w

  • 11

    BanhL.StrobelG. (2023). Generative artificial intelligence. Electron. Mark.33:63. doi: 10.1007/s12525-023-00680-1

  • 12

    BanlaweI. A. P.CruzJ. C.Dela (2020). Acoustic sensors for mango pulp weevil (Stretochenus Frigidus sp.) detection., in 2020 IEEE 10th International Conference on System Engineering and Technology (ICSET), Shah Alam, Malaysia. (Piscataway, NJ, New Jersey, US: IEEE), 191–195.

  • 13

    BanlaweI. A. P.dela CruzJ. C. (2023). Machine learning-based classification of mango pulp weevil activity utilizing an acoustic sensor. Micromachines14:1979. doi: 10.3390/mi14111979,

  • 14

    BaraZ. J.RiasatS.Ahsan KhanM. (2025). Integration of artificial intelligence (AI) and internet of things (IoT) in smart agriculture. Bincang Sains dan Teknologi4, 71–77. doi: 10.56741/bst.v4i02.941

  • 15

    BayarJ.AliN.CaoZ.RenY.DongY. (2025). Artificial intelligence of things (AIoT) for precision agriculture: applications in smart irrigation, nutrient and disease management. Smart Agricultural Technology12:101629. doi: 10.1016/j.atech.2025.101629

  • 16

    BedoyaC. L.NelsonX. J.BrockerhoffE. G.PawsonS.HayesM. (2022). Experimental characterization and automatic identification of stridulatory sounds inside wood. R. Soc. Open Sci.9:220217. doi: 10.1098/rsos.220217,

  • 17

    BelloutA.ZarboubiM.DliouA.LatifR.SaddikA. (2024a). Advanced YOLO models for real-time detection of tomato leaf diseases. Mathematical Modeling Computing11, 1198–1210. doi: 10.23939/mmc2024.04.1198

  • 18

    BelloutA.ZarboubiM.DliouA.LatifR.SaddikA. (2024b). YOLOv10-enabled IoT robot Car for accurate disease detection in strawberry cultivation. ITM Web of Conferences69:01011. doi: 10.1051/itmconf/20246901011

  • 19

    BelloutA.ZarboubiM.ElhosenyM.DliouA.LatifR.SaddikA. (2025). LT-YOLOv10n: a lightweight IoT-integrated deep learning model for real-time tomato leaf disease detection and management. Internet Things33:101663. doi: 10.1016/j.iot.2025.101663

  • 20

    BhujelA.KimN.-E.ArulmozhiE.BasakJ. K.KimH.-T. (2022). A lightweight attention-based convolutional neural networks for tomato leaf disease classification. Agriculture12:228. doi: 10.3390/agriculture12020228

  • 21

    BiH.LiT.XinX.ShiH.LiL.ZongS. (2025). Non-destructive estimation of wood-boring pest density in living trees using X-ray imaging and edge computing techniques. Comput. Electron. Agric.233:110183. doi: 10.1016/j.compag.2025.110183

  • 22

    BittnerJ. A.BalfeS.PittendrighB. R.PopovicsJ. S. (2018). Monitoring of the cowpea bruchid, Callosobruchus maculatus (Coleoptera: Bruchidae), feeding activity in cowpea seeds: advances in sensing technologies reveals new insights. J. Econ. Entomol.111, 1469–1475. doi: 10.1093/jee/toy086,

  • 23

    BobM. A.El-ShafieH. A. F.AmmarK. A. (2025). Evaluation and validation of acoustic sensors for detecting red palm weevil infestations in date palm trees under various insecticide treatments in the UAE. Outlook. Pest Manag.36, 113–118. doi: 10.1564/v36_jun_06

  • 24

    BorgianniL.BuaC.ColiE.AdamiD.GiordanoS. (2025). ‘‘Efficient distributed DNN for extreme edge computing in wildlife monitoring,’’ in 2025 IEEE 26th International Conference on High Performance Switching and Routing (HPSR), (IEEE), 1–6. doi: 10.1109/HPSR64165.2025.11038869

  • 25

    BozziniA.HuoL.BrugnaroS.MorganteG.PerssonH. J.FinozziV.et al. (2025). Multispectral drone images for the early detection of bark beetle infestations: assessment over large forest areas in the Italian south-eastern Alps. Front. For. Glob. Change8:1532954. doi: 10.3389/ffgc.2025.1532954

  • 26

    BuY.QiX.Wen JunBaoXuZ. (2017). Acoustic behaviors for two species of cerambycid larvae. J. Zhejiang A F Univ.34, 50–55.

  • 27

    CappaertD.McCulloughD.PolandT.SiegertN. (2005). Emerald ash borer in North America: a research and regulatory challenge. Am. Entomol.51, 152–165. doi: 10.1093/ae/51.3.152

  • 28

    ChenX.GaoL.ShenZ. (2025). Machine learning methods for crop Pest incidence prediction: an example of greenhouse Tuta Absoluta (Meyrick), in 2025 10th International Conference on Big Data Analytics (ICBDA), (Piscataway, NJ, New Jersey, US: IEEE), 222–228.

  • 29

    ChenC.-J.HuangY.-Y.LiY.-S.ChangC.-Y.HuangY.-M. (2020). An AIoT based smart agricultural system for pests detection. IEEE Access8, 180750–180761. doi: 10.1109/ACCESS.2020.3024891

  • 30

    ChenC.-J.LiY.-S.TaiC.-Y.ChenY.-C.HuangY.-M. (2022). Pest incidence forecasting based on internet of things and long short-term memory network. Appl. Soft Comput.124:108895. doi: 10.1016/j.asoc.2022.108895

  • 31

    ChenB.WanJ.CelestiA.LiD.AbbasH.ZhangQ. (2018). Edge computing in IoT-based manufacturing. IEEE Commun. Mag.56, 103–109. doi: 10.1109/MCOM.2018.1701231

  • 32

    ChristakakisP.PapadopoulouG.MikosG.KalogiannidisN.IoannidisD.TzovarasD.et al. (2024). Smartphone-based citizen science tool for plant disease and insect Pest detection using artificial intelligence. Technologies12:101. doi: 10.3390/technologies12070101

  • 33

    ÇimenH.Çakmakİ.TourayM.HazirS. (2025). Monitoring and control of red palm weevil infestation using IoT technology. Turk. J. Zool.49, 233–240. doi: 10.55730/1300-0179.3228

  • 34

    CivolaniS.BariselliM.OstiR.BernacchiaG. (2025). Insect Pest control from chemical to biotechnological approach: constrains and challenges. Insects16:528. doi: 10.3390/insects16050528,

  • 35

    DaiM.ShenY.LiX.LiuJ.ZhangS.MiaoH. (2024). Digital twin system of Pest management driven by data and model fusion. Agriculture14:1099. doi: 10.3390/agriculture14071099

  • 36

    DingJ.NematiM.RanaweeraC.ChoiJ. (2020). IoT connectivity technologies and applications: a survey. IEEE Access8, 67646–67673. doi: 10.1109/ACCESS.2020.2985932

  • 37

    DuD. (2019). Research on Acoustic Information Characteristics and Automatic Identification of Blueberry Typical Pests. Doctoral dissertation. Guiyang, China: Guizhou University.

  • 38

    DuarteA.Acevedo-MuñozL.GonçalvesC. I.MotaL.SarmentoA.SilvaM.et al. (2020). Detection of Longhorned borer attack and assessment in Eucalyptus plantations using UAV imagery. Remote Sens.12:3153. doi: 10.3390/rs12193153

  • 39

    ElbasiE.MostafaN.AlArnaoutZ.ZreikatA. I.CinaE.VargheseG.et al. (2023). Artificial intelligence Technology in the Agricultural Sector: a systematic literature review. IEEE Access11, 171–202. doi: 10.1109/ACCESS.2022.3232485

  • 40

    Esmail KararM.ReyadO.Abdel-AtyA.-H.OwyedS.HassanM. (2021). Intelligent IoT-aided early sound detection of red palm weevils. Comp. Mat. Cont.69, 4095–4111. doi: 10.32604/cmc.2021.019059

  • 41

    EstevaA.RobicquetA.RamsundarB.KuleshovV.DePristoM.ChouK.et al. (2019). A guide to deep learning in healthcare. Nat. Med.25, 24–29. doi: 10.1038/s41591-018-0316-z

  • 42

    FarinaE.NabhenJ. J.DacoregioM. I.BataliniF.MoraesF. Y. (2022). An overview of artificial intelligence in oncology. Future Sci. OA8:FSO787. doi: 10.2144/fsoa-2021-0074,

  • 43

    FarooqM. U.WaseemM.MazharS.KhairiA.KamalT. (2015). A review on internet of things (IoT). Int. J. Comp. Appl.113, 1–7. doi: 10.5120/19787-1571

  • 44

    GeorgievaM.BelilovS.DimitrovS.IlievM.TrenkinV.MirchevP.et al. (2022). Application of remote sensing data for assessment of bark beetle attacks in pine plantations in Kirkovo region, the eastern Rhodopes. Forests13:620. doi: 10.3390/f13040620

  • 45

    GhobadpourA.MonsalveG.CardenasA.MousazadehH. (2022). Off-road electric vehicles and autonomous robots in agricultural sector: trends, challenges, and opportunities. Vehicles4, 843–864. doi: 10.3390/vehicles4030047

  • 46

    GlotzbachR. (2024). Generative AI: an overview., in Society for Information Technology & Teacher Education International Conference, (Waynesville, NC USA: Association for the Advancement of Computing in Education (AACE)), 754–756

  • 47

    GokedaV.YalavarthiR. (2024). Deep hybrid model for pest detection: IoT - UAV -based smart agriculture system. J. Phytopathol.172:e13381. doi: 10.1111/jph.13381

  • 48

    GuéganJ. F.de ThoisyB.Gomez-GallegoM.JactelH. (2023). World forests, global change, and emerging pests and pathogens. Curr. Opin. Environ. Sustain.61:101266. doi: 10.1016/j.cosust.2023.101266

  • 49

    GuptaS.NagpalS. (2024) “Introduction to IoT: An Overview,’’ in IoT Evolution: Essentials & Applications. eds. M. N. Alam and S. Gupta (Dhaka: New Book Publications Ltd.), 1–25.

  • 50

    HartmannH.BattistiA.BrockerhoffE. G.BełkaM.HurlingR.JactelH.et al. (2025). European forests are under increasing pressure from global change-driven invasions and accelerating epidemics by insects and diseases. J. Kult.77, 6–24.

  • 51

    HasanM. R.RahmanM. M.ShahriarF.KhanM. S. I.Mohi UddinK. M.HasanM. M. (2024). Smart farming: leveraging IoT and deep learning for sustainable tomato cultivation and pest management. Crop Design3:100079. doi: 10.1016/j.cropd.2024.100079

  • 52

    HerrickN. J.MankinR. W. (2012). Acoustical detection of early instar Rhynchophorus ferrugineus (Coleoptera: Curculionidae) in Canary Island date palm, Phoenix canariensis (Arecales: Arecaceae). Fla. Entomol.95, 983–990. doi: 10.1653/024.095.0425

  • 53

    HessaneA.El YoussefiA.FarhaouiY.AghoutaneB.Abdellaoui AlaouiE. A.NayyarA. (2024). Artificial intelligence-driven prediction system for efficient management of Parlatoria Blanchardi in date palms. Multimed. Tools Appl.84, 15293–15329. doi: 10.1007/s11042-024-19635-5

  • 54

    HetzroniA.SorokerV.CohenY. (2016). Toward practical acoustic red palm weevil detection. Comput. Electron. Agric.124, 100–106. doi: 10.1016/j.compag.2016.03.018

  • 55

    HüttnerováT.SurovýP. (2024). Bark beetle detection method using electronic nose sensors. A possible improvement of early forest disturbance detection?Front. Forests Global Change7:1445094. doi: 10.3389/ffgc.2024.1445094

  • 56

    IllarionovaS.TregubovaP.ShukhratovI.ShadrinD.KedrovA.BurnaevE. (2024). Remote sensing data fusion approach for estimating forest degradation: a case study of boreal forests damaged by Polygraphus proximus. Front. Environ. Sci.12:1412870. doi: 10.3389/fenvs.2024.1412870

  • 57

    IordacheM.-D.MantasV.BaltazarE.PaulyK.LewyckyjN. (2020). A machine learning approach to detecting pine wilt disease using airborne spectral imagery. Remote Sens.12:2280. doi: 10.3390/rs12142280

  • 58

    JaboobA.DurrahO.ChakirA. (2024). “Artificial Intelligence: An Overview,” In Engineering Applications of Artificial Intelligence. Synthesis Lectures on Engineering, Science, and Technology. eds. A. Chakir, J. F. Andry, A. Ullah, R. Bansal and M. Ghazouani. Springer, Cham. doi: 10.1007/978-3-031-50300-9_1

  • 59

    JanieschC.ZschechP.HeinrichK. (2021). Machine learning and deep learning. Electron. Mark.31, 685–695. doi: 10.1007/s12525-021-00475-2

  • 60

    JiangW.ChenZ.ZhangH. (2024a). A time-frequency domain mixed attention-based approach for classifying wood-boring insect feeding vibration signals using a deep learning model. Insects15:1239424. doi: 10.3390/insects15040282,

  • 61

    JiangW.ChenZ.ZhangH.LiJ. (2024b). MelSPPNET—a self-explainable recognition model for emerald ash borer vibrational signals. Front. For. Glob. Change7:110066. doi: 10.3389/ffgc.2024.1239424

  • 62

    JiangQ.LiuY.RenL.SunY.LuoY. (2021). Early detection of the wood-boring insect Semanotus bifasciatus using acoustic detection technology. Research Square [Preprint]. doi: 10.21203/rs.3.rs-1069664/v1

  • 63

    JiangQ.LiuY.RenL.SunY.LuoY. (2022). Acoustic detection of the wood borer, Semanotus bifasciatus, as an early monitoring technology. Pest Manag. Sci.78, 4689–4699. doi: 10.1002/ps.7089

  • 64

    JongS. M. de.MeerF.van der (2006) Remote Sensing Image Analysis: Including the Spatial domain. Dordrecht Netherlands: Springer Netherlands.

  • 65

    KahlS.Wilhelm-SteinT.KlinckH.KowerkoD.EiblM. (2018). A baseline for large-scale bird species identification in field recordings. CLEF (Working Notes) (Germany: CEUR-WS.org.), 2155.

  • 66

    KapetasD.ChristakakisP.FaliagkaS.KatsoulasN.PechlivaniE. M. (2025a). AI-driven insect detection, real-time monitoring, and population forecasting in greenhouses. AgriEngineering7:29. doi: 10.3390/agriengineering7020029

  • 67

    KapetasD.ChristakakisP.GogliaV.PechlivaniE. M. (2025b). AI-based robotic trap for real-time insect detection, monitoring and population prediction., in 2025 IEEE 6th International Conference on Image Processing, Applications and Systems (IPAS), (Thessaloniki, Greece: IEEE), 1–7.

  • 68

    KatholeA. B.VhatkarK. N.PatilS. D. (2024). IoT-enabled pest identification and classification with new meta-heuristic-based deep learning framework. Cybern. Syst.55, 380–408. doi: 10.1080/01969722.2022.2122001

  • 69

    KhakimovA.SalakhutdinovI.OmolikovA.UtaganovS. (2022). Traditional and current-prospective methods of agricultural plant diseases detection: a review. IOP Conf. Ser. Earth Environ. Sci.951:012002. doi: 10.1088/1755-1315/951/1/012002

  • 70

    KhanA.MalebaryS. J.DangL. M.BinzagrF.SongH.-K.MoonH. (2024). AI-enabled crop management framework for pest detection using visual sensor data. Plants13:653. doi: 10.3390/plants13050653,

  • 71

    KoubaaA.AldawoodA.SaeedB.HadidA.AhmedM.SaadA.et al. (2020). Smart palm: an IoT framework for red palm weevil early detection. Agronomy10:987. doi: 10.3390/agronomy10070987

  • 72

    KumbulaS. T.MafongoyaP.PeerbhayK. Y.LotteringR. T.IsmailR. (2019). Using Sentinel-2 multispectral images to map the occurrence of the cossid moth (Coryphodema tristis) in Eucalyptus Nitens plantations of Mpumalanga, South Africa. Remote Sens.11:278. doi: 10.3390/rs11030278

  • 73

    LatifM. S.KazmiR.KhanN.MajeedR.IkramS.Ali-ShahidM. M. (2022). Pest prediction in rice using IoT and feed forward neural network. KSII Trans. Internet Inf. Syst.16, 17–37. doi: 10.3837/tiis.2022.01.008

  • 74

    LiJ.LiX.JuM.ZhaoX.WangY.YangF. (2024a). Lightweight implementation of the signal enhancement model for early wood-boring pest monitoring. Forests15:1903. doi: 10.3390/f15111903

  • 75

    LiJ.WuZ.XuY.YangF. (2025). An emerald ash borer early monitoring method for low-noise environments based on spectral features extraction of boring vibration signals. Comput. Electron. Agric.238:110739. doi: 10.1016/j.compag.2025.110739

  • 76

    LiJ.ZhaoX.LiX.JuM.YangF. (2024b). A method for classifying wood-boring insects for pest control based on deep learning using boring vibration signals with environment noise. Forests15:1875. doi: 10.3390/f15111875

  • 77

    LillesandT.KieferR. W.ChipmanJ. (2015). Remote Sensing and Image Interpretation. Hoboken, New Jersey: John Wiley & Sons.

  • 78

    LiuX.ChenZ.ZhangH.LiJ.JiangQ.RenL.et al. (2023). The interpretability of the activity signal detection model for wood-boring pests Semanotus bifasciatus in the larval stage. Pest Manag. Sci.79, 3830–3842. doi: 10.1002/ps.7566,

  • 79

    LiuM.-Y.LiG.-Y.ShiL.LiY.-Y.LiuH. (2025). Detection of the stem-boring damage by pine shoot beetle (Tomicus spp.) to Yunan pine (Pinus yunnanensis Franch.) using UAV hyperspectral data. Front. Plant Sci.16:1514580. doi: 10.3389/fpls.2025.1514580

  • 80

    LiuX.ZhangH.JiangQ.RenL.ChenZ.LuoY.et al. (2022). Acoustic denoising using artificial intelligence for wood-boring pests Semanotus bifasciatus larvae early monitoring. Sensors22:3861. doi: 10.3390/s22103861,

  • 81

    LuY.LiuP.TanC. (2025). MA-YOLO: a pest target detection algorithm with multi-scale fusion and attention mechanism. Agronomy15:1549. doi: 10.3390/agronomy15071549

  • 82

    LuoY.HuangH.RoquesA. (2023). Early monitoring of forest wood-boring pests with remote sensing. Annu. Rev. Entomol.68, 277–298. doi: 10.1146/annurev-ento-120220-125410,

  • 83

    LuoQ.WangH.ZhenZ.XiangboK. (2011). Automatic stridulation identification of bark beetles based on MFCC and BP network. J. Beijing For. Univ33, 81–85.

  • 84

    MabizelaX. M.KgopaA. T.MonchusiB. B. (2025) A systematic literature review on deep learning applications in agricultural Pest detection and control. in Proceedings of the 2025 International Conference on Artificial Intelligence and its ApplicationsNew York, NY, USAACM1––8

  • 85

    MahantaD. K.KrishnappaC.BhoiT. K. (2025). Microbial control of forest insect pests over 60 years (1964–2024): network analysis and bibliometric mapping. J. Natural Pesticide Res.12:100132. doi: 10.1016/j.napere.2025.100132

  • 86

    MankinR. W.BurmanH.MenocalO.CarrilloD. (2018). Acoustic detection of Mallodon dasystomus (Coleoptera: Cerambycidae) in Persea americana (Laurales: Lauraceae) branch stumps. Fla. Entomol.101, 321–323. doi: 10.1653/024.101.0226

  • 87

    MansourM.GamalA.AhmedA. I.SaidL. A.ElbazA.HerencsarN.et al. (2023). Internet of things: a comprehensive overview on protocols, architectures, technologies, simulation tools, and future directions. Energies16:3465. doi: 10.3390/en16083465

  • 88

    MarkovićD.VujičićD.TanaskovićS.ĐorđevićB.RanđićS.StamenkovićZ. (2021). Prediction of pest insect appearance using sensors and machine learning. Sensors21:4846. doi: 10.3390/s21144846,

  • 89

    MartínezR. D.IzquierdoA.VillacortaJ. J.del ValL.BasterraL.-A. (2023). Acoustic detection and localisation system for Hylotrupes bajulus L. larvae using a MEMS microphone array. Appl. Acoust.213:109618. doi: 10.1016/j.apacoust.2023.109618

  • 90

    MaruthaduraiR.VeerakumarT.VeershettyC.Sathis ChakaravarthiA. N. (2022). Acoustic detection of stem and root borer Neoplocaederus ferrugineus (Coleoptera: Cerambycidae) in cashew. J. Asia Pac. Entomol.25:101968. doi: 10.1016/j.aspen.2022.101968

  • 91

    MendelZ.VoetH.ModanN.NaorR.MentD. (2024). Seismic sensor-based management of the red palm weevil Rhynchophorus ferrugineus in date palm plantations. Pest Manag. Sci.80, 1053–1064. doi: 10.1002/ps.7836,

  • 92

    MillerT.MikiciukG.DurlikI.MikiciukM.ŁobodzińskaA.ŚniegM. (2025). The IoT and AI in agriculture: the time is now—a systematic review of smart sensing technologies. Sensors25:3583. doi: 10.3390/s25123583

  • 93

    MinW.ZhaiM.ChenS.HuangL.WangF.ZhuT. (2025). Mobile acoustic net: a novel early detection model for wood-boring pests. Comput. Electron. Agric.229:109699. doi: 10.1016/j.compag.2024.109699

  • 94

    MoZ.BaoX.LiY.WangY.WangC.LiF. (2025). Pest recognition and classification using hybrid quantum convolution and diverse branch block. J. Appl. Entomol.149, 1335–1350. doi: 10.1111/jen.13457

  • 95

    MohanrajB.MurugesanJ.AyyappanR.RubighaM.KalaiselviS.PriyaM. (2025). Blockchain-enabled AI for Pest prediction in bamboo Rice farming., in 2025 6th International Conference on Data Intelligence and Cognitive Informatics (ICDICI), (Piscataway, New Jersey, US: IEEE), 63–69.

  • 96

    Muhammad NadeemR.JaffarA.Muhammad SaleemR. (2022). IoT and machine learning based stem borer pest prediction. Intell. Autom. Soft Comput.31, 1377–1392. doi: 10.32604/iasc.2022.020680

  • 97

    NahiyoonS. A.RenZ.WeiP.LiX.LiX.XuJ.et al. (2024). Recent development trends in plant protection UAVs: a journey from conventional practices to cutting-edge technologies—a comprehensive review. Drones8:457. doi: 10.3390/drones8090457

  • 98

    NardiD.BozziniA.MorganteG.GaccioneA.FinozziV.BattistiA. (2023). Participatory ground data are complementary to satellite bark beetle detection. Ann. For. Sci.80:46. doi: 10.1186/s13595-023-01216-5

  • 99

    NäsiR.HonkavaaraE.Lyytikäinen-SaarenmaaP.BlomqvistM.LitkeyP.HakalaT.et al. (2015). Using UAV-based photogrammetry and hyperspectral imaging for mapping bark beetle damage at tree-level. Remote Sens.7, 15467–15493. doi: 10.3390/rs71115467

  • 100

    NicolettiR.De MasiL.MigliozziA.CalandrelliM. M. (2024). Analysis of dieback in a coastal pinewood in Campania, southern Italy, through high-resolution remote sensing. Plants13:182. doi: 10.3390/plants13020182,

  • 101

    NjorogeA. W.AffognonH.MutungiC.RohdeB.RichterU.HenselO.et al. (2016). Frequency and time pattern differences in acoustic signals produced by Prostephanus truncatus (horn) (Coleoptera: Bostrichidae) and Sitophilus zeamais (Motschulsky) (Coleoptera: Curculionidae) in stored maize. J. Stored Prod. Res.69, 31–40. doi: 10.1016/j.jspr.2016.06.005

  • 102

    NyakuriJ. P.NkundinezaC.GateraO.NkurikiyeyezuK.MwitendeG. (2025). AI-IoT driven system for agricultural pest outbreak risk prediction. Research Square [Preprint]. doi: 10.21203/rs.3.rs-6881038/v1

  • 103

    OmerA. O.AlharbiH. A.HusainM.RasoolK. G.AlwaneenW. S.AldawoodA. S. (2024). Acoustic sensor-based field efficacy evaluation of three different insecticides—trunk injections against the red palm weevil, Rhynchophorus ferrugineus. Sound & Vibration59:1787. doi: 10.59400/sv1787

  • 104

    OnyejegbuL. N. (2023). AI and IoT-based model for the detection and treatment of sweet potato pests and diseases in precision farming., in Deep Learning Indaba 2023. Accra, Ghana: OpenReview.net.

  • 105

    PalaniH. K.IlangovanS.SenthilvelP. G.ThirupurasundariD. R.KumarR. K. (2023). AI-powered predictive analysis for Pest and disease forecasting in crops, in 2023 International Conference on Communication, Security and Artificial Intelligence (ICCSAI)(Piscataway, New Jersey, US: IEEE), 950–954.

  • 106

    PanopoulouC.AntonopoulosA.ArapostathiE.StamouliM.KatsilerosA.TsagkarakisA. (2024). Using multispectral data from UAS in machine learning to detect infestation by Xylotrechus chinensis (Chevrolat) (Coleoptera: Cerambycidae) in mulberries. Agronomy14:2061. doi: 10.3390/agronomy14092061

  • 107

    Pierre NyakuriJ.NkundinezaC.GateraO.NkurikiyeyezuK. (2024). State-of-the-art deep learning algorithms for internet of things-based detection of crop pests and diseases: a comprehensive review. IEEE Access12, 169824–169849. doi: 10.1109/ACCESS.2024.3455244

  • 108

    PotamitisI.EliopoulosP.RigakisI. (2017). Automated remote insect surveillance at a global scale and the internet of things. Robotics6:19. doi: 10.3390/robotics6030019

  • 109

    PotamitisI.RigakisI. (2024). Automated vibroacoustic monitoring of trees for borer infestation. Sensors24:3074. doi: 10.3390/s24103074,

  • 110

    PotamitisI.RigakisI.TatlasN.-A.PotirakisS. (2019). In-vivo vibroacoustic surveillance of trees in the context of the IoT. Sensors19:1366. doi: 10.3390/s19061366,

  • 111

    PrasathB.AkilaM. (2023). IoT-based pest detection and classification using deep features with enhanced deep learning strategies. Eng. Appl. Artif. Intell.121:105985. doi: 10.1016/j.engappai.2023.105985

  • 112

    QaziS.KhawajaB. A.FarooqQ. U. (2022). IoT-equipped and AI-enabled next generation smart agriculture: a critical review, current challenges and future trends. IEEE Access10, 21219–21235. doi: 10.1109/ACCESS.2022.3152544

  • 113

    RahmanW.HossainM. M.HasanM. M.IqbalM. S.RahmanM. M.HasanK. F.et al. (2024). Automated detection of harmful insects in agriculture: a smart framework leveraging IoT, machine learning, and blockchain. IEEE Trans. Artif. Intell.5, 4787–4798. doi: 10.1109/TAI.2024.3394799

  • 114

    RashidA. B.KausikM. A. K. (2024). AI revolutionizing industries worldwide: a comprehensive overview of its diverse applications. Hybrid Adv.7:100277. doi: 10.1016/j.hybadv.2024.100277

  • 115

    ReaganT. E.MulcahyM. M. (2019). Interaction of cultural, biological, and varietal controls for management of stalk borers in Louisiana sugarcane. Insects10:305. doi: 10.3390/insects10090305,

  • 116

    RigakisI.PotamitisI.TatlasN.-A.PotirakisS. M.NtalampirasS. (2021). TreeVibes: modern tools for global monitoring of trees for borers. Smart Cities4, 271–285. doi: 10.3390/smartcities4010017

  • 117

    RustiaD. J. A.ChiuL.LuC.WuY.ChenS.ChungJ.et al. (2022). Towards intelligent and integrated pest management through an AIoT -based monitoring system. Pest Manag. Sci.78, 4288–4302. doi: 10.1002/ps.7048

  • 118

    RyallsJ. M. W.GarrattM. P. D.SpadaroD.MauchlineA. L. (2024). The benefits of integrated pest management for apple depend on pest type and production metrics. Front. Sustain. Food Syst.8:1321067. doi: 10.3389/fsufs.2024.1321067

  • 119

    SahuS. K.MokhadeA.BokdeN. D. (2023). An overview of machine learning, deep learning, and reinforcement learning-based techniques in quantitative finance: recent progress and challenges. Appl. Sci.13:1956. doi: 10.3390/app13031956

  • 120

    SalamaR.Al-TurjmanF.ChaudharyP.YadavS. P. (2023). (Benefits of internet of things (IoT) applications in health care - an overview), in 2023 International Conference on Computational Intelligence, Communication Technology and Networking (CICTN), (Piscataway, New Jersey, US: IEEE), 778–784.

  • 121

    SaleemR. M.BashirR. N.FaheemM.HaqM. A.AlhussenA.AlzamilZ. S.et al. (2023). Internet of things based weekly crop pest prediction by using deep neural network. IEEE Access11, 85900–85913. doi: 10.1109/ACCESS.2023.3301504

  • 122

    SaleemM. H.PotgieterJ.ArifK. M. (2021). Automation in agriculture by machine and deep learning techniques: a review of recent developments. Precis. Agric.22, 2053–2091. doi: 10.1007/s11119-021-09806-x

  • 123

    SankarR.AnubalaM.HemalathaM.LavanyaU.NarmadhaS. (2025). Twin-layer deep learning model with IoT integration for intelligent Pest detection in smart farming, in 2025 Fourth International Conference on Smart Technologies, Communication and Robotics (STCR), (Piscataway, New Jersey, US: IEEE), 1–6.

  • 124

    SaranyaT.DeisyC.SrideviS.AnbananthenK. S. M. (2023). A comparative study of deep learning and internet of things for precision agriculture. Eng. Appl. Artif. Intell.122:106034. doi: 10.1016/j.engappai.2023.106034

  • 125

    SevinçV. (2023). Assessment of the effects of the biotic and abiotic harmful factors on the amount of industrial wood production with deep learning. Environ. Sci. Pollut. Res.30, 41999–42015. doi: 10.1007/s11356-023-25161-y,

  • 126

    SevinçV. (2025). Evaluating the effectiveness of the forest pests and diseases control methods on the industrial wood production using deep learning. Sci. Rep.15:34679. doi: 10.1038/s41598-025-06932-8,

  • 127

    ShafiqueS.RashidH.SarwarM. A.AkhtarS.RazaU.FarhanM.et al. (2025). Integrated pest management and controlling strategies in Pakistan. Indus J. Biosci. Res.3, 803–808. doi: 10.70749/ijbr.v3i6.1605

  • 128

    ShaoH.ZhangH.HuP.ChenY.LiW.SunL.et al. (2024). Feasibility study of forest wood-boring pest detection base on hyperspectral LiDAR. Int. J. Remote Sens.45, 1926–1948. doi: 10.1080/01431161.2024.2320180

  • 129

    SharadaK.ChoudharyS. L.HarikrishnaT.DixitR. S.SumanS. K.Ayyappa ChakravarthiM.et al. (2025). GeoAgriGuard: AI-driven pest and disease management with remote sensing for global food security. Remote Sens. Earth Syst. Sci.8, 409–422. doi: 10.1007/s41976-025-00192-w

  • 130

    ShiH.ChenZ.ZhangH.LiJ.LiuX.RenL.et al. (2022). A waveform mapping-based approach for enhancement of trunk borers’ vibration signals using deep learning model. Insects13:596. doi: 10.3390/insects13070596,

  • 131

    ShiH.ChenZ.ZhangH.LiJ.LiuX.RenL.et al. (2023). Enhancement of boring vibrations based on cascaded dual-domain features extraction for insect pest Agrilus planipennis monitoring. Forests14:902. doi: 10.3390/f14050902

  • 132

    ShresthaA.HickeJ. A.MeddensA. J. H.KarlJ. W.StahlA. T. (2024). Evaluating a novel approach to detect the vertical structure of insect damage in trees using multispectral and three-dimensional data from drone imagery in the northern Rocky Mountains, USA. Remote Sens.16:1365. doi: 10.3390/rs16081365

  • 133

    StevanoskaS.DavcevD.JovanovskaE. M.MitreskiK. (2020). IoT--based system for real-time monitoring and insect detection in vineyards, in Proceedings of the 18th ACM Symposium on Mobility Management and Wireless Access, (New York, NY, USA: ACM), 133–136.

  • 134

    SutantoK.Al-ShahwanI.HusainM.RasoolK.MankinR.AldawoodA. (2023). Field evaluation of promising indigenous entomopathogenic fungal isolates against red palm weevil, Rhynchophorus ferrugineus (Coleoptera: Dryophthoridae). J. Fungi9:68. doi: 10.3390/jof9010068,

  • 135

    Vidya MadhuriE.RupaliJ. S.SharanS. P.Sai PoojaN.SujathaG. S.SinghD. P.et al. (2025). Transforming pest management with artificial intelligence technologies: the future of crop protection. J. Crop Health77:48. doi: 10.1007/s10343-025-01109-9

  • 136

    WangL.GaoY.LiuY.ZhongL.WangS.MaY.et al. (2025). Monitoring pine shoot beetle damage using UAV imagery and deep learning semantic segmentation under different forest backgrounds. Forests16:668. doi: 10.3390/f16040668

  • 137

    WangL.LiC.LuoY.WangG.DouZ.HaqI. U.et al. (2023). Current and future control of the wood-boring pest Anoplophora glabripennis. Insect Sci.30, 1534–1551. doi: 10.1111/1744-7917.13187,

  • 138

    WangJ.LinQ.MengS.HuangH.LiuY. (2024). Individual tree-level monitoring of pest infestation combining airborne thermal imagery and light detection and ranging. Forests15:112. doi: 10.3390/f15010112

  • 139

    WangB.MaoY.AshryI.Al-FehaidY.Al-ShawafA.NgT. K.et al. (2021). Towards detecting red palm weevil using machine learning and fiber optic distributed acoustic sensing. Sensors21:1592. doi: 10.3390/s21051592,

  • 140

    WangJ.MengS.LinQ.LiuY.HuangH. (2022). Detection of Yunnan pine shoot beetle stress using UAV-based thermal imagery and LiDAR. Appl. Sci.12:4372. doi: 10.3390/app12094372

  • 141

    XuX.LiJ.ZhangH. (2023). Multi-channel time-domain boring-vibration-enhancement method using RNN networks. Insects14:817. doi: 10.3390/insects14100817,

  • 142

    YanJ.WangX. (2022). Unsupervised and semi-supervised learning: the next frontier in machine learning for plant systems biology. Plant J.111, 1527–1538. doi: 10.1111/tpj.15905,

  • 143

    YazgacB. G.KirciM.KivanM. (2016). Detection of sunn pests using sound signal processing methods, in 2016 Fifth International Conference on Agro-Geoinformatics (Agro-Geoinformatics), (Piscataway, New Jersey, US: IEEE), 1–6.

  • 144

    YinJ.ZhangH.ChenZ.LiJ. (2025). Detecting emerald ash borer boring vibrations using an encoder-decoder and improved DenseNet model. Pest Manag. Sci.81, 384–401. doi: 10.1002/ps.8442,

  • 145

    ZarboubiM.BelloutA.ChabaaS.DliouA. (2024a). Revolutionizing tomato pest management: synergy of deep learning, IoT, and precision agriculture. Math. Model. Comput.11, 1167–1178. doi: 10.23939/mmc2024.04.1167

  • 146

    ZarboubiM.BelloutA.ChabaaS.DliouA. (2025). Custombottleneck-VGGNet: advanced tomato leaf disease identification for sustainable agriculture. Comput. Electron. Agric.232:110066. doi: 10.1016/j.compag.2025.110066

  • 147

    ZarboubiM.BelloutA.ChabaaS.DliouA. (2026). Enhancing integrated pest management with IoT and YOLO-Evo: a smart, low-cost monitoring system for sustainable apple farming. Results Eng.29:108850. doi: 10.1016/j.rineng.2025.108850

  • 148

    ZarboubiM.BelloutA.ChabaaS.DliouA.ZeroualA. (2024b). IoT-based Pest detection in agriculture using raspberry pi and YOLOv10m for precision farming. ITM Web of Conferences69:04017. doi: 10.1051/itmconf/20246904017

  • 149

    ZarboubiM.ChabaaS.DliouA.ZeroualA. (2024c). Smart Pest control in grain warehouses: YOLOv8-powered IoT robot Car for precision agriculture., in 2024 International Conference on Global Aeronautical Engineering and Satellite Technology (GAST), (New York, NY, USA: IEEE), 1–6.

  • 150

    ZengM.ChenS.LiuH.WangW.XieJ. (2024). HCFormer: a lightweight pest detection model combining CNN and ViT. Agronomy14:1940. doi: 10.3390/agronomy14091940

  • 151

    ZhangC.ChenZ.ZhangH.LiJ. (2024). A dual branch time-frequency multi-dilated dense network for wood-boring pest activity signal enhancement in the larval stage. Forests16:20. doi: 10.3390/f16010020

  • 152

    ZhangH.LiJ.CaiG.ChenZ.ZhangH. (2023). A CNN-based method for enhancing boring vibration with time-domain convolution-augmented transformer. Insects14:631. doi: 10.3390/insects14070631,

  • 153

    ZhouQ.ZhangX.YuL.RenL.LuoY. (2021). Combining WV-2 images and tree physiological factors to detect damage stages of Populus gansuensis by Asian longhorned beetle (Anoplophora glabripennis) at the tree level. For. Ecosyst.8:35. doi: 10.1186/s40663-021-00314-y

  • 154

    ZhuD.XieL.ChenB.TanJ.DengR.ZhengY.et al. (2023). Knowledge graph and deep learning based pest detection and identification system for fruit quality. Internet of Things21:100649. doi: 10.1016/j.iot.2022.100649

  • 155

    Zincir-HeywoodN.MelliaM.DiaoY. (2021). “Overview of artificial intelligence and machine learning,” in Communication Networks and Service Management in the Era of Artificial Intelligence and Machine Learning, (Hoboken, New Jersey, USA: Wiley), 19–32.

Summary

Keywords

acoustic sensors, artificial intelligence of things, borer-pests, RS, smart pest management

Citation

Bibi M, Skarmeta AF and Khan SZ (2026) Revolutionizing pest control: harnessing cutting-edge technologies in controlling forest pests. Front. For. Glob. Change 9:1777311. doi: 10.3389/ffgc.2026.1777311

Received

29 December 2025

Revised

26 February 2026

Accepted

06 March 2026

Published

08 April 2026

Volume

9 - 2026

Edited by

Shulin He, Chongqing Normal University, China

Reviewed by

Mohamed Zarboubi, Universite Ibn Zohr Ecole Nationale des Sciences Appliquees, Morocco

Volkan Sevinç, Mugla University, Türkiye

Updates

Copyright

*Correspondence: Maria Bibi, ; Antonio F. Skarmeta,

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Article metrics