Next Article in Journal
Effects of Forest Surface Fuel Bed Structure on Flame Residence Time
Previous Article in Journal
Plant Roots Exert Stronger Co-Structuring Effects than Soils on the Litter Microbial Community Following the Succession of Fagus lucida Forests
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 17
Issue 4
10.3390/f17040477
forests-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

Systematic Review of Applications Using Artificial Intelligence (AI) for Wooden Materials

by
Enis Kucuk
1,* and
Urs Buehlmann
2
1
Department of Forest Faculty, Artvin Çoruh University, Artvin 08100, Türkiye
2
Department of Sustainable Biomaterials, Virginia Tech, 1650 Research Center Drive, Blacksburg, VA 24061, USA
*
Author to whom correspondence should be addressed.
Forests 2026, 17(4), 477; https://doi.org/10.3390/f17040477
Submission received: 11 March 2026 / Revised: 31 March 2026 / Accepted: 10 April 2026 / Published: 13 April 2026
(This article belongs to the Topic Artificial Intelligence, Remote Sensing and Digital Twin Driving Innovation in Sustainable Natural Resources and Ecology)
Browse Figures
Versions Notes

Abstract

This study investigates the relevant literature on applications of Artificial Intelligence (AI) for wood as a material using a systematic review and screening process. The Web of Science (WoS) database identified 50 peer-reviewed publications dealing with AI applications for wood as a material. Bibliometrix and VOSviewer software were used to evaluate publication trends, country contributions, keyword co-occurrences, and AI application areas. Based on these analyses, an annual growth rate of 23.28% between 2014 and 2025 (November) in publications published per year was measured and an average of 6.92 citations per publication was observed as of November 2025. Most notably, a considerable increase in AI-focused research after 2023 was identified. Before 2022, work done using AI tools (such as neural networks, deep learning, and others) did not necessarily use the term AI and hence were not found by our search. China, Canada, and Poland were the countries with the highest number of publications. The leading journals with publications on AI applications for wood as a material were Forests and Wood Material Science and Engineering. The most frequently occurring keywords in the publications reviewed were “AI,” “machine learning,” and “deep learning.” In general, according to the publications reviewed, AI applications for wooden materials improved productivity, material evaluation, and quality assurance. The findings highlighted the impact of AI on the sector and show that AI will change the industry.

1. Introduction

The term “artificial intelligence” (AI) was first introduced by John McCarthy [1,2], who in 1960 stated that AI is to become “a system which is to evolve intelligence of human order [3] (p. 79).” Today, the literature comprises a variety of definitions of AI. For instance, Russell and Norvig [4] synthesized various definitions and categorized AI into four general frameworks: (i) systems that think like humans, (ii) systems that think rationally, (iii) systems that act like humans, and (iv) systems that act rationally. According to Göde and Kalkan [5], AI is a combination of all four general frameworks used to facilitate human tasks to increase efficiency, solve complex problems, and enable discoveries. AI consists of several subfields, including expert systems, robotics, natural language processing, computer vision, and speech recognition and comprehension. The most common AI techniques researched today are machine learning, artificial neural networks, deep learning, expert systems, genetic algorithms, and fuzzy logic [5].
Various disciplines benefit from AI applications, and AI has attracted increased attention in recent years as an interdisciplinary field of research and innovation [6]. Artificial Intelligence (AI) today is at the cusp of significantly impacting society and industry, including but not limited to healthcare, finance, manufacturing, and entertainment, thereby profoundly changing the way people work, live, and interact with the world [7].
Artificial Intelligence (AI) is also transforming the forest products industry. The biological origin of wood results in inherent variability in the material and hence in the processes used. By adopting technological advances, including computational capacity, sensors, communication technologies, and intelligent systems [7], wood’s inherent variability can be better monitored and managed. From the relevant literature, examples of AI applications relating to wooden materials include tool condition monitoring for the automatic diagnosis of drill wear in melamine-faced particleboards [8], modeling and control of lumber drying processes [9], detection and recognition of surface defects in particleboards [10], and AI-assisted approaches in the design of wood products such as chairs [11], to name a few.
Reviews about AI for wood as a material have been published previously [7,12,13,14]. However, these reviews have a narrow focus looking at specific subfields such as AI application in wood surface inspection [12], the possibilities of AI use in the furniture/woodworking industry [7], an examination of the use of machine learning/image processing for wood log classification and grading [13], or Hong Chun et al. [14] who reviewed methods for automatic lumber surface defect detection. By design, these studies do not offer a holistic framework of AI applications in the forest products industry. However, systematically compiling and classifying current knowledge with a holistic perspective is important for both academic and industrial purposes in terms of awareness, integration, use and further development of AI technologies for the forest products industry. The aim of this study is to use the perspective of wood as a material to systematically examine AI applications in the relevant literature and to provide a holistic view of what applications exist while creating a reliable source of reference. Publications that appeared during the search query were examined, considering their AI applications domain, including but not limited to manufacturing systems, processing operations, production monitoring, quality assurance, and performance optimization. The results and implications presented are discussed, and potential directions for future research are identified below.
The rest of this publication is organized as follows: Section 2 explains the approach and scope of this literature review, Section 3 systematically classifies the available material, and Section 4 elaborates on AI applications in the context of wooden materials. Finally, Section 5 summarizes the main contributions of the study, draws conclusions, and outlines potential directions for future research.

2. Methodology

2.1. Study Design

This study conducted a systematic literature review on publications that contained the terms “Artificial Intelligence” or “AI.” Furthermore, a bibliometric analysis of relevant publications was integrated into this study [15,16]. The focus of this study are publication trends, journals, countries where the research was conducted, keywords, materials, and AI techniques. Also, this systematic review was conducted and reported in accordance with the PRISMA 2020 guidelines where applicable [17]. Literature reviews are commonly undertaken to delineate the boundaries of existing knowledge, identify research gaps, and critically assess the development of a given field [18]. Typically, the review process follows an iterative structure. It begins with defining appropriate search terms, and follows systematically searching the literature, and analyzing the material found in a consistent and methodical manner [19,20]. Beyond summarizing existing studies, systematic reviews help to reveal gaps in the literature as well as methodological inconsistencies and weaknesses, thereby contributing to the identification of priorities for future research [21,22]. Fahimnia et al. [23] proposed a multi-stages framework for conducting systematic literature reviews, consisting of defining appropriate search terms, obtaining initial search results, refining the results, generating initial data statistics, and performing detailed data analysis. The present study followed comparable methodological steps indicated above, including data acquisition and search strategy, application of inclusion criteria and selection process, data processing and evaluation, and a synthesis process. In addition, the publications found were examined in detail from a content perspective to identify their application focus on AI techniques used for the processing of wood and wood-based materials.

2.1.1. Data Acquisition and Search Strategy

A comprehensive literature search was performed using the Web of Science (WoS) database, one of the leading academic databases [24,25,26]. WoS allows a homogeneous comparison and classification process and provides compatibility with bibliometric analysis tools such as Bibliometrix (Biblioshiny version 5.2.0) [27] and VOSviewer (version 1.6.2) [28]. The categories provided by WoS, e.g., micro, meso, and other citation topics facilitate the identification of areas where research activity is concentrated within specific key domains. Citation topics are organized hierarchically. Macro topics representing broad scientific fields, meso topics covering more specific thematic areas within those fields, and micro topics focusing on narrowly defined research areas composed of closely related publications [29]. This approach minimizes the subjectivity that may arise from author-defined inputs and allows for a more objective evaluation of the literature.
Boolean and truncation operators were used to reveal potential documents in the WoS database. The search query combined the terms relating to wood materials with the AI application prospect. The query was executed as follows: (“wood” OR “softwood” OR “soft-wood” OR “hardwood” OR “hard-wood” OR “composite wood” OR “composite-wood” OR “engineered wood” OR “engineered-wood” OR “wood-based” OR “wood based” OR “medium density fiberboard” OR “medium-density fiberboard” OR “MDF” OR “particleboard” OR “particle board” OR “plywood” OR “hardboard” OR “insulation board” OR “insulation-board” OR “wood wool” OR “oriented strand board” OR “oriented-strand board” OR “OSB” OR “veneer” OR “laminated veneer” OR “laminated-veneer” OR “laminated veneer lumber” OR “LVL” OR “glue laminated timber” OR “glue-laminated timber” OR “glulam” OR “cross laminated timber” OR “cross-laminated timber” OR “CLT” OR “oriented strand lumber” OR “oriented-strand lumber” OR “OSL” OR “parallel strand lumber” OR “parallel-strand lumber” OR “PSL” OR “I joist” OR “I-joist” OR “timber” OR “lumber”) AND (“artificial intelligence” OR “AI”).

2.1.2. Inclusion Criteria and Selection Process

Eligibility criteria were specified to determine suitable publications. First, the studies considered for review were required to directly address wood as a material and indicate a clear connection with AI applications. In terms of accessibility and comprehensibility, only documents published in English were evaluated. In terms of providing a broad overview of temporal trends in the literature, no publication date restrictions were applied. Additionally, to maintain academic rigor, only peer-reviewed journal publications were included in the scope of this study.
Given these exclusion criteria, publications that contained the search terms but were contextually irrelevant (e.g., “AI” referred to Alaskan Indian or active ingredient, or misleading results related to cognitive load theory (CLT)) were excluded from the analysis. In addition, studies that examined wood at the cellular, chemical, or anatomical levels were also excluded from this review. These eligibility criteria were used to identify studies that provide concrete and noteworthy evidence about how to effectively utilize AI in fields related to the use of wood as a material to focus on research offering high practical value and actionable insights.

2.1.3. Data Processing and Evaluation

First, a preliminary search was executed on WoS using the defined keywords shown above. Irrelevant studies were excluded after careful examination of the titles and abstracts and full texts when necessary. Mendeley (v2.140.0) [30] was used for reference management (but not for the data extraction process). Subsequently, the defined eligibility criteria were considered to examine the full texts of the accessible publications.
Besides benefiting from Mendeley [30], Bibliometrix [27] and VOSviewer (version 1.6.2) [28] were employed for the bibliometric analysis; however, the software programs did not affect the data extraction process as they were used for organizing and managing the references and for bibliometric mapping and visualization. The data extracted from the publications identified included references, year, application focus, wood material involved, AI methods, and other relevant details.
The screening and selection of publications were performed based on the predefined eligibility criteria. Before the final examination, titles and abstracts were examined and, if necessary, a full text assessment was executed. In cases of uncertainty regarding eligibility, the final inclusion decision was made jointly by the two authors.

2.1.4. Synthesis Process

A qualitative content analysis approach was used to evaluate the data obtained within the scope of the systematic review [31]. Each publication was examined in terms of the research objective, the AI methodology applied, the application focus, the type of wood material used, and the findings obtained. During this process, similarities and differences among the studies were systematically compared, and the data were grouped by the application focus of the study described.
In the synthesis process, the bibliographic characteristics of the publications (e.g., publication year, journal, number of authors, country, and number of citations) were evaluated, and the data was integrated with the bibliometric analysis findings, including publication trends, most prolific countries, and author keyword co-occurrence patterns. Subsequently, the AI methods (e.g., deep learning, machine learning, heuristic algorithms, neural nets, adaptive systems, and others) and application areas (e.g., defect detection, quality analysis, process optimization, and others) used in these publications were identified. Studies with similar themes were brought together to form conceptual clusters, and the findings were synthesized thematically. Furthermore, a frequency-based examination of the characteristics and contributions of the publications, including their research aim and relevance to wood materials were considered. The findings were introduced through an Application Focus, Wood Material Types, and AI Tools and Methods analysis, which aimed to visualize the general trends and research gaps in the field.

2.2. Limitations

This study only reviewed publications in English. No non-peer-reviewed publications, conference proceedings (peer-reviewed or non-peer-reviewed), or industrial reports were included in this study. The search was done using the Web of Science (WoS) database [24,25,26]. Also, instead of focusing on a specific domain (for example, Yi et al. [12] who reviewed the prospects of AI-based wood surface inspection), this study aimed at providing a general framework for the use of AI applications used for wooden materials. The aim of the study was not to provide in-depth methodological or application-focused detail, but rather to outline the main trends, key application areas, and emerging research directions. Also, studies examining wood itself, such as on the cellular, anatomical, chemical, or microstructural level, were excluded from the scope of this review. Moreover, the keyword selection used for this review may limit access to some possible publications. For instance, even though a publication may describe building blocks for AI, the term “artificial intelligence” or “AI” may not appear in the text, but “neural networks,” “deep learning,” or “machine learning” may be used, like, for example, in publications by Singer et al. [32], Van Nguyen et al. [33], or Thomas [34]. Hence, such publications were not reviewed, similar to other systematic review publications in existence [35,36,37,38,39].
Due to the high heterogeneity of the publications identified in terms of research objectives, materials, and AI techniques, a risk of bias assessment was not performed as this review aimed to provide a qualitative comparison.
GenAI tools (ChatGPT, GPT-5.2 Instant, OpenAI; Grammarly, v1.2.248.1873) were used for searching and understanding words and terms encountered in preparing the manuscript and to translate words and ideas formed in other languages into English. GenAI was not used in the preparation of this manuscript or to form parts of the work presented and the text used.

3. Results

On 5 November 2025, the keywords provided in Section 2.1.1 were searched in the Web of Science (WoS) database (record identification), yielding 7349 publications (Figure 1). After screening the titles and abstracts, followed by a full-text review, 51 publications that met all the inclusion criteria (among them the use of “Artificial Intelligence” and/or “AI” in the text) were selected for the systematic review (record exclusion). However, one publication could not be retrieved: the publication by Pham and Alcock [40] was inaccessible due to pay wall restrictions, and interlibrary loan requests were unsuccessful. Hence, 50 publications were left to be reviewed (final inclusion).

3.1. Bibliometric Analysis

3.1.1. Data Summary of Bibliometrics

Table 1 shows the main bibliometric information obtained with Bibliometrix [27] of the publications reviewed from 35 different journals. Publications identified in our search were published between 2014 and November 2025. The average number of citations per publication was 6.92, and 229 keywords were used, with “artificial intelligence” (with 15 occurrences) being the most frequently used. Each publication was authored by an average of 4.48 researchers, with 12% of the collaborations international.

3.1.2. Annual Publication Trends

Academic interest in AI applications on wood as a material increased after 2022, with 10, 14, and 10 publications in 2023, 2024, and in 2025 (until November), respectively (Figure 2). Considering all 50 publications found, the annual average growth rate (which describes how quickly the number of publications increases each year over the study period and helps indicate whether a research field is growing, remaining stable, or declining) during the 2014–2025 period was 23.28%, and the average age of the publications was 2.46 years.
Table 1. Data summary of the bibliometric dataset.
Table 1. Data summary of the bibliometric dataset.
SummaryResults
Data
Timespan2014–2025
Sources (Journals, Books, etc.)35
Documents50
Annual Growth Rate of Publications (%) (2014–2025)23.28
Document Average Age (years)2.46
Average citations per doc (as of November. 2025)6.92
Total References cited1806
Document Contents
Keywords Plus *84
Author’s Keywords229
Authors
Total Authors204
Single-authored documents0
Average Co-Authors per document4.48
International co-authorships (%)12
* Keywords Plus “Are words or phrases that frequently appear in the titles of an article’s references, but do not appear in the title of the article itself [41]”.

3.1.3. Journal Distribution

An examination of the distribution of publications by journals reveals that the highest number of publications was published in Forests (7 publications, total of 51 citations as of November 2025), followed by Applied Sciences (5 publications, 13 citations), Wood Material Science and Engineering (4 publications, 7 citations), and Sensors (3 publications, 24 citations) as shown in Table 2. All the remaining publications (31) were published in journals that published only one publication that fit our search criteria (31 publications, Table 2).

3.1.4. Most Cited Publications

Table 3 shows the leading publications in terms of citation counts, based on WoS data as of November 2025. Studies with an annual average of five or more citations were included in the table. The most cited publication was by Nasir et al. [42], which focused on monitoring acoustic emissions during the sawing process using AI (54 citations), followed by Gao et al. [43], which applied deep learning for knot detection and had 32 citations by November 2025. Gao et al. [43] was followed by the publication by Chen et al. [44], which addressed defect detection in edge-glued panels using deep learning with 28 citations. This was followed by Przybyś-Małaczek et al. [45] on tool condition monitoring using machine learning (18 citations), the study by Szwajka et al. [46], focused on predicting surface roughness in MDF milling processes using artificial neural networks and fuzzy modeling (17 citations), and Alsakka et al. [47], who examined a computer vision-based process data-collection approach (17 citations).

3.1.5. Country-Level Scientific Output

Table 4 presents the scientific output by country. In this study, the “Country Scientific Production” indicator from the Bibliometrix package [27] was used to evaluate each country’s contribution. This indicator does not directly measure the number of publications by country by lead author; instead, it counts the number of author affiliations in each country (author appearances). Based on this calculation approach, the highest number of author appearances was recorded for China (18), followed by Canada (16) and Poland (12). These countries are followed by India (9), and then by Brazil, Malaysia, and Slovakia (5 each). Turkey appears in the list with four author appearances, while other countries are represented by a single appearance.

3.1.6. Country-Level Analysis of Corresponding Authors

While Table 4 focuses on the participating countries (by author appearances), Table 5 displays the number of publications by country, based only on the corresponding author’s data. The results were obtained from a Bibliometrix analysis [27] using WoS data, updated as of November 2025. Twenty-five countries were identified, with China having the highest number of publications (nine publications), followed by Canada and Poland (five each), India and Romania (three each), and Brazil, Italy, Malaysia, Switzerland, and Turkey (two each). The remaining five countries are represented by a single publication.

3.1.7. Keyword Analysis

Figure 3 presents a word cloud generated using the Bibliometrix software [27] based on author keywords. The publications were also imported into VOSviewer [28] for further bibliometric analysis. As a result of the analysis, a total of 229 keywords were identified and grouped into 27 clusters, as shown in Figure 4. Keyword frequency provides useful clues about the intellectual landscape and conceptual orientation of the research area [16]. The co-occurrence network of author keywords reveals that the research domain is distinctly centered around AI. The concept of AI serves as both the positional and conceptual core of the network, around which all other themes are clustered. Techniques including machine learning, deep learning, image processing, and defect detection appear as the most central components of the network. The joint presence of wood and machine vision highlights the growing role of intelligent algorithms in wood processing and material evaluation.
The network shown in Figure 4 suggests a mutual interaction between data-oriented research methods and their real world applications within the wood industry. The most frequent keyword is “artificial intelligence” (15 occurrences), followed by “deep learning” (5 occurrences), “machine learning” (5 occurrences), “defect detection” (4 occurrences), and “wood” (4 occurrences).

3.1.8. Application Focus of Publications

Table 6 shows the focal points of each publication using the terminology employed. At the present stage of the evolution of AI techniques, AI-based tools were most frequently applied for defect detection in wooden materials. To that end, Verly Lopes et al. [48], Gao et al. [43], and Tran et al. [49] focused on the rapid detection of knot defects in wood using AI instead of human inspection. Fan et al. [50] aimed to perform intelligent cutting of timber by automatically detecting knot cracks. Similarly, in terms of surface defect detection, Ji et al. [51] conducted a study on creating intelligent cutting plans of timber and automatic wood species recognition, while Chen et al. [44] focused on detecting edge defects in edge-banded panels. Li et al. [52] aimed to identify defects such as edge breakage, scratches, and surface damage on melamine-coated decorative particleboard surfaces, whereas Zhou et al. [53] and Zhang et al. [10] worked on detecting defects on particleboard surfaces. Chun et al. [54] focused on automatically detecting surface defects in lumber (Meranti, Merbau, Kembang Semangkuk (KSK), and Rubberwood), including brown stain, blue stain, knots, borer holes, rot, bark pockets, wane, and splits, while Achatz and Schubert [55] aimed to detect log quality and possible insect damage. Studies that addressed internal defects in wooden materials include Vijayalakshmi et al. [56], who aimed for the classification of both surface and internal defects in lumber non-destructively; Yu et al. [57] focused on detecting internal damage in wooden utility poles; Kato et al. [58] by assessing internal crack severity using cross-sectional images; Xu et al. [59] detected decay, voids, or cracks inside logs; and Vacek et al. [60] used AI to identify wood species and the pith.
Studies on the machining process of wood include Murugesan et al. [61], who analyzed delamination during the drilling of wood–plastic composite (WPC) panels. Abramov et al. [62] aimed to develop fault prediction for motors and bearings in woodworking machines. Nasir et al. [42] predicted power and surface waviness for circular saw cutting. Chaiprabha and Chancharoen [63] developed an intelligent drilling machine capable of detecting the drill’s first contact with the material, the moment it penetrates through, and the type of material (MDF, acrylic, or glass) using only position and force data. Jegorowa et al. [64] and Antoniuk et al. [8] analyzed drill wear from images of holes drilled in melamine-faced particleboard, while Przybyś-Małaczek et al. [45] developed a method to classify the wear condition of cutting tools. Sieradzki et al. [65] aimed to evaluate the condition of holes drilled in melamine-coated particleboard.
Another major application focus centered around wood’s mechanical and physical properties. Bardak and Bardak [66] proposed a fast, low-cost, and effective method for estimating oak wood density. Similarly, Jovic et al. [67] aimed to predict both the factors affecting wood bonding strength and the power consumption during wood processing, while Szwajka et al. [46] aimed to estimate surface roughness during MDF milling using measuring vibrations data. Josué et al. [68] aimed to predict the tensile strength of bevel glue joints, and Kılıç et al. [69] aimed to predict veneer bonding strength for different combinations of board type, coating type, adhesive type, and cutting direction. Miguel et al. [70] aimed to predict the mechanical strength properties of wood with high accuracy using non-destructive data such as density and ultrasonic measurements, and Li et al. [71] aimed to predict the compressive strength of softwoods non-destructively and with high accuracy.
Some studies addressed operational aspects, e.g., operation and production monitoring. Shahi et al. [72] developed a decision-support model to assess sawmill efficiency, predict future productivity, and identify optimal input–output combinations. Alsakka et al. [47] developed a computer vision system called TiDA, which automatically monitors process times in lightweight wood panelized construction based on AI and not requiring an underlining dataset. Carratù et al. [73] provided a system that measures log diameters on trucks using security camera images. Borz and Păun [74] recognized saw machine operating conditions from videos and presented a low-cost production monitoring system for small-scale sawmills, while Cheţa et al. [75] showed that operational states of a handheld bandsaw (cutting, movement, pauses) can be automatically detected and monitored using low-cost sensors and Artificial Intelligence methods. Borz et al. [76] classified operating states of a vertical multi-saw machine (off, transition, running) using vibration data from an inexpensive accelerometer attached to the machine.
Other studies, such as Sexton et al. [77], focused on optimizing processes aimed to provide a system that predicts wood thickness (based on historical data) and automatically adjusts wood planers. Haddouche et al. [78] aimed to achieve reduced energy consumption and operational costs by managing heating and environmental conditions in a facility. Sundarapandiyan et al. [79] identified the best combination of drill speed, feed rate, and diameter to minimize thrust force and torque during the drilling of wood–plastic composite (WPC) panels, while Raamets et al. [80] aimed to increase production efficiency by creating a digital twin of a factory analyzed with AI. Yeshiwas et al. [81] determined the optimal fiber length, fiber ratio, and plastic ratio to improve the mechanical properties of wood–plastic composites using AI methods.
Other studies focused on a quality perspective. Melo and Miguel [82] aimed to predict quality parameters in particleboard production, while Ericsson et al. [83] performed wood surface quality control using robotics and AI within the framework of Industry 4.0. Ricardo et al. [84] developed a system that automatically analyzes oak staves used for wine barrels, and Ruttico et al. [85] presented Woodot, a mobile robot that uses an AI-powered vision system and a robotic arm to automate knot repair on curved glulam beams. Ramos-Maldonado et al. [86] developed a model that predicts coating quality using large-scale real factory data, while Settimi et al. [87] aimed to determine the position and orientation of tools such as drills and saws used in carpentry with the help of cameras and AI, and to provide users with correct angles, depth, and direction during drilling and cutting via augmented reality (AR).
Of the 50 publications identified by our search in WoS, two review articles were found. In this case, Kitek-Kuzman et al. [7] presented the potential uses of AI in the furniture and woodworking industry. Yi et al. [12] reviewed the prospects of AI-based wood surface inspection.

3.1.9. Wood Material Types Used

Table 6 also presents the wooden materials discussed in the publications reviewed. Solid wood (spruce, fir, pine, larch, hemlock, cedar, oak, beech, ash, birch, eucalyptus, acacia, black cherry, red oak, white oak, and poplar), and wood components produced from these species (e.g., lumber samples, wood utility poles, wood staves, and wooden window frames and doors) were discussed. Panel products discussed included plywood, oriented strand board (OSB), particleboard, and medium-density fiberboard (MDF). Other publications focused on structural or engineered timber or composite elements, glued-laminated timber (glulam), and cross-laminated timber (CLT). Additionally, wood–non-wood composite materials, such as wood–plastic composites (WPC), were also used in some of these studies.
Table 6. Application focus and materials used in the reviewed studies.
Table 6. Application focus and materials used in the reviewed studies.
PublicationApplication FocusMaterial
[62]Motor-Bearing DiagnosticsParticleboard
[82]Predicting Particleboard QualityParticleboards (Eucalyptus grandis)
[66]Prediction of the Wood DensityWood Samples—Pedunculate oak (Quercus robur)
[67]The Wood Bonding Strength AnalyzingWood Samples—Beech and Spruce
[70]Predicting Wood ResistanceWood Samples—Eucalyptus grandis
[42]Monitoring Sawing ProcessWood Samples—Douglas-fir wood
[48]Wood Knot DetectionWooden Board—Southern yellow pine
[57]Defect DetectionWood Utility Poles
[74]Sawmill Process MonitoringWood Logs—Norway spruce
[75]Production MonitoringWood Logs
[43]Wood Knot DetectionWood Logs—SpruceWood
[83]Automated Inspection System for Wood SurfacesCross Laminated Timber (CLT)
[58]Defect DetectionTimber—Cedar
[44]Defect DetectionEdge-Glued Wooden Panels
[72]Evaluating Operational Performance in SawmillsNot Applicable
[76]Monitoring The Operational Performance of Sawmilling FacilitiesWood Logs—Coniferous Logs
[71]Determination of Mechanical PropertiesWood Samples—Larch, Hemlock, and Mongolica
[51]Defect Detection and GradingTimber—Betula
[59]Defect DetectionWood Logs—Black Cherry, Red Oak, White Oak, Cottonwood
[63]Monitoring Drilling OperationsMedium-Density Fiberboard (MDF)
[64]Tool Condition MonitoringParticleboard—(Melamine Faced Chipboard)
[45]Detecting Tool WearParticleboard—(Chipboard)
[8]Drill Wear RecognitionParticleboard—(Melamine Faced Chipboard)
[47]Manufacturing Efficiency MeasurementLightweight Wood Panelized Construction
[73]Log Diameter MeasurementWood Logs
[46]Surface Roughness PredictionMedium-Density Fiberboard (MDF)
[52]Defect DetectionParticleboard—(melamine-impregnated decorative particleboard)
[54]Defect DetectionTimber—Meranti, Merbau, KSK, and Rubberwood
[60]Species Classification and Pith DetectionWood Logs—Oak, Aspen, Beech, Linden, Ash, Spruce, Pine
[53]Defect DetectionParticleboard
[50]Defect DetectionTimber—(Cerasus pseudocerasus, Fagus longipetiolata Seem, Pinus yunnanensis Franch, Fraxinus mandschurica Rupr, and Betula platyphylla Suk.)
[56]Defect DetectionWood Samples—Jungle Cork (Holoptelea integrifolia or the Indian elm), Babul (Vachellia nilotica, commonly known as Acacia nilotica), and Pala Indigo (Wrightia tinctoria or dyer’s oleander)
[65]Drill Wear RecognitionParticleboard—(melamine-faced chipboard)
[7]AI Applications ReviewNot Applicable
[87]Tool Pose DetectionNot Applicable
[77]Model-Based Control of Wood PlanersLumber samples
[68]Tensile Strength PredictionTimber—Terminalia superba (Fraké)
[69]Bonding Strength PredictionFlakeboard, MDF, Oriented Strand board (OSB), and Veneer (beech, oak, pine)
[84]Wood Stave ClassificationStaves—Oak Staves
[12]Surface Defect InspectionNot Applicable
[49]Wood Knot Detection and ClassificationWood Samples—Acacia
[55]Quality ClassificationWood Logs—(Spruce, Fir, Douglas Fir, Larch)
[10]Particleboard Surface Defect DetectionParticleboard
[61]Delamination Of Wood Composite MaterialWood Plastic Composite—(WPC)
[78]Energy Efficiency ManagementMDF and Melamine Panels
[79]Drilling Process OptimizationWood Plastic Composite—(WPC)
[80]Virtual Factory ModelingWooden Window Frames and Doors
[81]Mechanical Properties OptimizationWood Plastic Composite—(WPC)
[85]Defect DetectionGlue-Laminated Timber—(Glulam)
[86]Veneer Quality PredictionPlywood—Pinus radiata veneer

3.1.10. AI Techniques Employed

Among 50 reviewed publications, two review articles [7,12] were not considered for the review of the AI techniques employed. Instead, the remaining 48 publications were the subject of this section. In this case, AI techniques were used for modeling, prediction, and decision-making processes for wooden materials (Table 7). In this case, Deep Learning tools (convolutional neural networks—CNNs), including ResNet, YOLO, and EfficientNet, demonstrated high accuracy in image-based defect detection and classification studies [43,50,54,56]. Also, classical machine learning algorithms (e.g., support vector machines—SVM, random forests—RF, and k-nearest neighbors—KNN) were used to improve predictive accuracy and robustness in the context of wooden material [45,64,86].
Table 7. AI methods reported in the reviewed studies.
Table 7. AI methods reported in the reviewed studies.
ReferenceAI-Based Techniques *
[62]Fuzzy Logic
[55]ResNet50, EfficientNetB0, Grad-CAM
[74]ANN
[47]YOLOv4
[8]ResNet-18, KNN, RF, XGBoost, LGBM, SVM
[66]FIS
[76]NN, RF
[73]YOLOv4
[63]SVM
[44]Inception ResNetV2, K-means
[75]ANN
[54]ResNet50
[83]CNN
[50]YOLOv5, YOLOX, YOLOv7
[43]ResNet-18
[78]ANN
[64]SVM, KNN, RF, RBF, MLP
[51]Machine Vision, NN
[68]ANFIS, LSTM
[67]ANN
[58]Shallow CNN
[69]ANN, ANFIS
[71]SE-DenseNet
[52]YOLOv8
[82]ANN
[70]ANN
[61]Fuzzy Logic, PSO, HOA
[42]ANFIS, PSO
[45]Decision Trees, RF, XGBoost, Gradient Boosting, KNN, SVM, LGBM
[80]K-means clustering
[86]RF, XGBoost, KNN, SVM, Lasso, Logistic Regression
[84]EfficientNet-B1, ResNet152V2, Saliency Maps, Grad-CAM
[85]SAM
[87]ResNet18, EfficientNetV2
[77]Gaussian Processes
[72]ANN
[65]VGG16, VGG19, ResNet101, LIME, Grad-CAM
[79]BWOA
[46]RBF-NN, TSK Fuzzy Logic
[49]DenseNet201, YOLO
[60]AlexNet, GoogleNet, NASNet-Large, Inception-v3, ViT, HRNet-W32
[48]YOLOv3
[56]U-Net, YOLO
[59]improved PSO
[81]ANN, Genetic Algorithm
[57]CART, KNN, SVM
[10]Ganomaly, ResNet
[53]Transformer, YOLOv8
* AI-based techniques include Artificial Intelligence algorithms, machine learning models, deep learning architectures, optimization algorithms integrated with AI systems, and explainable AI methods.
In some of the studies reviewed, fuzzy logic approaches were employed (e.g., adaptive neuro-fuzzy inference systems—ANFIS), and genetic algorithms or some other algorithms were used (e.g., particle swarm optimization—PSO [61,66,81]). Only Achatz and Schubert [55], Ricardo et al. [84], and Sieradzki et al. [65] applied explainable AI (XAI). XAI is defined by the Defense Advanced Research Projects Agency (DARPA) as to “Produce more explainable models, while maintaining a high level of learning performance (prediction accuracy); and enable human users to understand, appropriately trust, and effectively manage the emerging generation of artificially intelligent partners [88].”

4. Discussion

The earliest publication in which the term “Artificial Intelligence” occurs found by the Web of Science (WoS) database is Tikhomirov [89] in 1975. Tikhomirov discussed philosophical and the psychological aspect of Artificial Intelligence. The first publication about the application of AI for wooden materials dates to 2014 [62] and described the efforts for fault detection in a woodworking machine. Since then (to November 2025), 49 applications of AI for wooden materials were published in peer-reviewed journals. These 50 publications were published across 35 different journals, which also shows that efforts of using AI for wooden materials are relatively dispersed and no leading journals have been established, underscoring the interdisciplinary nature of the field. Among the journals that directly address wood-related topics, Forests (7 publications) and Wood Material Science and Engineering (4 publications) stand out. However, 31 journals contributed only a single publication.
Of the 50 publications identified over the 12-year period, 34 (64%) were published within the final three years, suggesting that research interest in this area has recently intensified. The relatively high annual growth rate (23.28%) of publications per year and the low average document age (2.46 years) underline that most of the works have been produced in recent years, suggesting that interest in AI applications for the wooden materials sector has increased recently. A total of 231 researchers contributed to the publications reviewed, and none of the publications were single-authored, pointing to a strong tendency toward collaborative research in the field.
The Countries’ Scientific Production analysis revealed that China, Canada, and Poland were the countries with the highest number of authors in the publications found. China’s prominent position may reflect its comprehensive national AI strategies aligned with the goal of becoming “the world’s leading AI power by 2030” [90]. To that end, China reportedly spent a combined public and private investment of nearly US $100 billion on AI in 2025 alone, showing the country’s strong national commitment to this field [90]. The visibility of Canada and Poland seems to be largely influenced by studies with multiple co-authors and interdisciplinary collaboration inside each country, where researchers from different academic units contribute to a higher number of author appearances per publication. Besides collaborative research practices, China’s strong position reflects a high volume of publications. Other countries display lower production levels, which likely result from fewer publications and more limited patterns of co-authorship.
The most highly cited study is the publication by Nasir et al. [42] titled “Acoustic emission monitoring of sawing process: AI approach for optimal sensory feature selection,” which has received 54 citations. According to the WoS Citations Topics (micro) classification, many citations to this work originate from publications related to wood properties (32 citations) and advanced machining (9 citations). Nasir et al.’s [42] publication is followed by the study of Gao et al. [43], which has accumulated 32 citations and focuses on deep convolutional neural networks for wood knot defect detection. Citations for this study are mainly concentrated in the areas of deep visual recognition (eight citations), rotating machinery diagnostics (two citations), and wood properties (two citations). Considering that both studies are relatively recent and attract citations from multiple research disciplines suggests a growing scholarly interest at the intersection of AI and wooden materials, while also indicating an increasingly interdisciplinary direction in the literature.
As expected, the keyword co-occurrence analysis indicates that the research field is clearly centered on AI (Figure 4). In this case, the “artificial intelligence” keyword, as the most frequent term to appear, indicates that it is the core conceptual focus of the literature. Other frequent keywords like “deep learning,” “machine learning,” “defect detection,” and “wood” reflect a close linkage between methodological AI approaches and their practical applications in wooden materials.
The studies reviewed indicate that AI methods can achieve high accuracy and efficiency in detecting knots, cracks, surface deformations, and internal defects in wooden materials. In this case, image-processing-based approaches show a strong capability in detecting both surface and internal defects and, in many cases, performed better than traditional inspection methods [43,50,56]. This body of research indicates that AI is increasingly being adopted as a practical and reliable approach for determining wooden material quality, with clear benefits in terms of speed and accuracy.
In some other studies, tool wear, delamination, and surface defects were successfully predicted with the help of AI applications and led to more stable manufacturing processes [8,45,61,64]. Quality control and operational efficiency appear to benefit greatly from the use of AI applications and further research will increase the benefit of such tools.
Artificial Intelligence (AI) methods appear to be practical tools for predicting mechanical and physical properties of wooden materials, including density, strength, bonding performance, and surface roughness [46,66,68,69]. Such data-driven and predictive approaches are increasingly being adopted for material characterization, offering an alternative to traditional experimental methods.
Monitoring shop-floor operations (including machine conditions, processing times, and cutting and drilling stages) using AI can be regarded as an effective approach for improving production efficiency and enhancing process visibility [47,73,75,76]. It appears that AI-driven monitoring and decision-support systems appear particularly promising for improving automation in small- and medium-sized manufacturing environments.
Studies relating to parameter optimization demonstrate that production variables such as drilling speed, feed rate, and energy consumption can be effectively optimized using AI [77,78,81]. These studies pointed out the potential of AI-supported modeling and optimization techniques to contribute to improved energy efficiency and more stable manufacturing operations.
Studies, including quality classification, surface inspection, and robot-assisted repair, indicate that AI contributes to faster and more standardized quality assurance processes for products made using wooden materials [82,83,84,85,86]. Their work demonstrates that AI visual inspection and automation systems offer a powerful alternative to traditional quality control mechanisms.
A wide variety of wooden materials were involved in the studies reviewed, indicating that AI applications extend across both natural and engineered wood products. Solid wood species (spruce, beech, oak) were commonly used materials for the defect detection and classification studies [51,55], while panel-based materials (MDF, OSB, particleboard) were materials for studies relating to surface quality, bonding strength, and roughness estimation [46,52,82].
Hybrid approaches that integrate methods such as deep learning, optimization algorithms, and fuzzy logic enhance both process efficiency and the accuracy of decision-making mechanisms. These AI systems in areas such as quality control, surface inspection, and process optimization contribute to the digital transformation of the industry using wooden materials in line with the Industry 4.0 vision [91].

5. Conclusions

This study reviewed the relevant literature on applications of Artificial Intelligence (AI) for wood as a material using a systematic review process, uncovering 50 publications, published between 2014 and 2025. Only publications in English and publications that used the term “Artificial Intelligence” or “AI” in its body were considered, leaving out publications that discuss the building blocks of AI like deep learning, machine learning, neural networks, and others. Also, this systematic review only covers work published in peer-reviewed journals, leaving out conference papers and research by companies and institutions who do not share their findings publicly.
Publications published on applications of AI for wood as a material have increased notably over the past few years, with 14 publications published in 2024 and 10 publications in 2025 through November, respectively. Thirty-five different journals published the 50 AI-related studies, with China, Canada, and Poland being the most prolific contributors to such studies. The word cloud of author keywords around Artificial Intelligence shows particle swarm optimization, convolutional and artificial neural networks, computer vision, defect detection, deep learning, machine learning, machine vision and Industry 4.0 as the most used keywords in the publications reviewed. Applications for defect detection, machine learning-based quality assessment, production process monitoring, and parameter optimization were reported using applied AI methods such as deep learning techniques (particularly CNN-based architectures) as well as classical machine learning approaches (e.g., SVM, RF, ANN).
This study confirms the growing use of AI in the field of wood and forest products research. Despite this growth, the use of explainable AI (XAI) remains limited, with only three studies explicitly benefiting from this approach. Accordingly, future research could place greater emphasis on XAI-based methods, which have been increasingly highlighted in the literature for their role in improving the interpretability and reliability of complex AI models. As this review shows, the wood and forest-based industry has started to use AI for their competitive advantage. Time will tell how the industry will fare in the competition with substitutes and with consumers’ preferences.

Author Contributions

Conceptualization, E.K. and U.B.; methodology, E.K. and U.B.; software, E.K.; validation, E.K. and U.B.; formal analysis, E.K.; investigation, E.K.; resources, E.K. and U.B.; data curation, E.K.; writing—original draft preparation, E.K.; writing—review and editing, U.B.; visualization, E.K. and U.B.; supervision, U.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

Data are available upon request.

Acknowledgments

The authors thank R. E. Thomas, USDA Forest Service and M. Bumgardner, USDA Forest Service for their thorough review and for their beneficial inputs. We also appreciate the thoughtful input from three anonymous reviewers. In this work, GenAI chatbots from OpenAI’s ChatGPT (GPT-5.2 Instant), and Grammarly (v1.2.248.1873) were used for explaining terms and translating information from other languages into English. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
AIArtificial Intelligence
ANNArtificial Neural Network
ANFISAdaptive Neuro-Fuzzy Inference System
BWOABinary Whale Optimization Algorithm
CNNConvolutional Neural Network
DenseNetDensely Connected Convolutional Network
EfficientNetEfficient Convolutional Neural Network Architecture
FISFuzzy Inference System
Grad-CAMGradient-weighted Class Activation Mapping
HOAHippopotamus Optimization Algorithm
HRNetHigh-Resolution Network
KNNk-Nearest Neighbors
LGBMLight Gradient Boosting Machine
LIMELocal Interpretable Model-Agnostic Explanations
LSTMLong Short-Term Memory
MLPMulti-layer Perception
NASNetNeural Architecture Search Network
NNNeural Network
PSOParticle Swarm Optimization
RBFRadial Basis Function
ResNetResidual Network
RFRandom Forest
SAMSegment Anything Model
SVMSupport Vector Machine
TSKTakagi–Sugeno–Kang (Fuzzy Model)
XAIExplainable Artificial Intelligence
ViTVision Transformer
XGBoostExtreme Gradient Boosting
YOLOYou Only Look Once

References

  1. Wieneke, H.; Voigt, I. Principles of Artificial Intelligence and Its Application in Cardiovascular Medicine. Clin. Cardiol. 2024, 47, e24148. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  2. Rajaraman, V. JohnMcCarthy—Father of Artificial Intelligence. Resonance 2014, 19, 198–207. [Google Scholar] [CrossRef] [Scilit]
  3. McCarthy, J. Programs with Common Sense; Her Majesty’s Stationery Office (HMSO): Teddington, UK, 1958; pp. 75–84.
  4. Russell, S.; Norvig, P. A Modern, Agent-Oriented Approach to Introductory Artificial Intelligence. ACM SIGART Bull. 1995, 6, 24–26. [Google Scholar] [CrossRef] [Scilit]
  5. Göde, A.; Kalkan, A. What is Artificial Intelligence? In Matematik ve Fen Bilimleri Üzerine Araştırmalar-IV; Akpınar, A., Ed.; Özgür Yayınları: Gaziantep, Turkey, 2023; ISBN 978-975-447-841-9. [Google Scholar]
  6. Liu, N.; Shapira, P.; Yue, X. Tracking Developments in Artificial Intelligence Research: Constructing and Applying a New Search Strategy. Scientometrics 2021, 126, 3153–3192. [Google Scholar] [CrossRef] [Scilit]
  7. Kitek-Kuzman, M.; Kropivšek, J.; Kariž, M. Possibilities of Using Artificial Intelligence in Furniture/Woodworking Industry. Drv. Ind. 2024, 75, 371–380. [Google Scholar] [CrossRef] [Scilit]
  8. Antoniuk, I.; Kurek, J.; Krupa, A.; Wieczorek, G.; Bukowski, M.; Kruk, M.; Jegorowa, A. Advanced Feature Extraction Methods from Images of Drillings in Melamine Faced Chipboard for Automatic Diagnosis of Drill Wear. Sensors 2023, 23, 1109. [Google Scholar] [CrossRef] [Scilit]
  9. Elustondo, D.; Matan, N.; Langrish, T.; Pang, S. Advances in Wood Drying Research and Development. Dry. Technol. 2023, 41, 890–914. [Google Scholar] [CrossRef] [Scilit]
  10. Zhang, C.; Wang, C.; Zhao, L.; Qu, X.; Gao, X. A Method of Particleboard Surface Defect Detection and Recognition Based on Deep Learning. Wood Mater. Sci. Eng. 2025, 20, 50–61. [Google Scholar] [CrossRef] [Scilit]
  11. Liu, J.; Gao, H.; Yezhova, O. Generative Design of Bamboo Furniture Combining Game Theory and AI-Generated Content. BioRes 2025, 20, 8611–8631. [Google Scholar] [CrossRef] [Scilit]
  12. Yi, L.P.; Akbar, M.F.; Wahab, M.N.A.; Rosdi, B.A.; Fauthan, M.A.; Shrifan, N.H.M.M. The Prospect of Artificial Intelligence-Based Wood Surface Inspection: A Review. IEEE Access 2024, 12, 84706–84725. [Google Scholar] [CrossRef] [Scilit]
  13. Sandvik, Y.J.; Futsæther, C.M.; Liland, K.H.; Tomic, O. A Comparative Literature Review of Machine Learning and Image Processing Techniques Used for Scaling and Grading of Wood Logs. Forests 2024, 15, 1243. [Google Scholar] [CrossRef] [Scilit]
  14. Hong Chun, T.; Hashim, U.R.; Ahmad, S.; Salahuddin, L.; Choon, N.H.; Kanchymalay, K. A Review of the Automated Timber Defect Identification Approach. Int. J. Electr. Comput. Eng. IJECE 2023, 13, 2156. [Google Scholar] [CrossRef] [Scilit]
  15. Xiao, Y.; Watson, M. Guidance on Conducting a Systematic Literature Review. J. Plan. Educ. Res. 2019, 39, 93–112. [Google Scholar] [CrossRef] [Scilit]
  16. Donthu, N.; Kumar, S.; Mukherjee, D.; Pandey, N.; Lim, W.M. How to Conduct a Bibliometric Analysis: An Overview and Guidelines. J. Bus. Res. 2021, 133, 285–296. [Google Scholar] [CrossRef] [Scilit]
  17. Page, M.J.; McKenzie, J.E.; Bossuyt, P.M.; Boutron, I.; Hoffmann, T.C.; Mulrow, C.D.; Shamseer, L.; Tetzlaff, J.M.; Akl, E.A.; Brennan, S.E.; et al. The PRISMA 2020 Statement: An Updated Guideline for Reporting Systematic Reviews. BMJ 2021, 372, n71. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  18. Tranfield, D.; Denyer, D.; Smart, P. Towards a Methodology for Developing Evidence-Informed Management Knowledge by Means of Systematic Review. Br. J. Manag. 2003, 14, 207–222. [Google Scholar] [CrossRef] [Scilit]
  19. Saunders, M.; Lewis, P.; Thornhill, A. Research Methods for Business Students, 5th ed.; Person Education Limited: Harlow, UK, 2009. [Google Scholar]
  20. Dodgson, J.E. Critical Analysis: The Often-Missing Step in Conducting Literature Review Research. J. Hum. Lact. 2021, 37, 27–32. [Google Scholar] [CrossRef] [Scilit]
  21. Paul, J.; Criado, A.R. The Art of Writing Literature Review: What Do We Know and What Do We Need to Know? Int. Bus. Rev. 2020, 29, 101717. [Google Scholar] [CrossRef] [Scilit]
  22. Rowley, J.; Slack, F. Conducting a Literature Review. Manag. Res. News 2004, 27, 31–39. [Google Scholar] [CrossRef] [Scilit]
  23. Fahimnia, B.; Sarkis, J.; Davarzani, H. Green Supply Chain Management: A Review and Bibliometric Analysis. Int. J. Prod. Econ. 2015, 162, 101–114. [Google Scholar] [CrossRef] [Scilit]
  24. Birkle, C.; Pendlebury, D.A.; Schnell, J.; Adams, J. Web of Science as a Data Source for Research on Scientific and Scholarly Activity. Quant. Sci. Stud. 2020, 1, 363–376. [Google Scholar] [CrossRef] [Scilit]
  25. Zhu, J.; Liu, W. A Tale of Two Databases: The Use of Web of Science and Scopus in Academic Papers. Scientometrics 2020, 123, 321–335. [Google Scholar] [CrossRef] [Scilit]
  26. Chadegani, A.A.; Salehi, H.; Yunus, M.M.; Farhadi, H.; Fooladi, M.; Farhadi, M.; Ebrahim, N.A. A Comparison between Two Main Academic Literature Collections: Web of Science and Scopus Databases. arXiv 2013, arXiv:1305.0377. [Google Scholar] [CrossRef] [Scilit]
  27. Aria, M.; Cuccurullo, C. Bibliometrix: An R-Tool for Comprehensive Science Mapping Analysis. J. Informetr. 2017, 11, 959–975. [Google Scholar] [CrossRef] [Scilit]
  28. VOSviewer—Visualizing Scientific Landscapes. Available online: https://www.vosviewer.com// (accessed on 27 March 2026).
  29. Citation Topics. Available online: https://incites.zendesk.com/hc/en-gb/articles/22514077746961-Citation-Topics (accessed on 27 March 2026).
  30. Mendeley|Homepage. Available online: https://www.mendeley.com/ (accessed on 27 March 2026).
  31. Gazquez-Garcia, J.; Sánchez-Bocanegra, C.L.; Sevillano, J.L. AI in the Health Sector: Systematic Review of Key Skills for Future Health Professionals. JMIR Med. Educ. 2025, 11, e58161. [Google Scholar] [CrossRef] [Scilit]
  32. Singer, H.; İlçe, A.C.; Şenel, Y.E.; Burdurlu, E. Artificial Neural Network–Based Prediction Model to Minimize Dust Emission in the Machining Process. Saf. Health Work 2024, 15, 317–326. [Google Scholar] [CrossRef] [Scilit]
  33. Nguyen, T.; Nguyen, T.; Ji, X.; Guo, M. Predicting Color Change in Wood During Heat Treatment Using an Artificial Neural Network Model. BioResources 2018, 13, 6250–6264. [Google Scholar] [CrossRef] [Scilit]
  34. Thomas, E. An Artificial Neural Network for Real-Time Hardwood Lumber Grading. Comput. Electron. Agric. 2017, 132, 71–75. [Google Scholar] [CrossRef] [Scilit]
  35. Abyaneh, A.G.; Ghanbari, H.; Mohammadi, E.; Amirsahami, A.; Khakbazan, M. An Analytical Review of Artificial Intelligence Applications in Sustainable Supply Chains. Supply Chain Anal. 2025, 12, 100173. [Google Scholar] [CrossRef] [Scilit]
  36. Chintalapati, S.; Pandey, S.K. Artificial Intelligence in Marketing: A Systematic Literature Review. Int. J. Mark. Res. 2022, 64, 38–68. [Google Scholar] [CrossRef] [Scilit]
  37. Vishwakarma, L.P.; Singh, R.K.; Mishra, R.; Kumari, A. Application of Artificial Intelligence for Resilient and Sustainable Healthcare System: Systematic Literature Review and Future Research Directions. Int. J. Prod. Res. 2025, 63, 822–844. [Google Scholar] [CrossRef] [Scilit]
  38. Vlačić, B.; Corbo, L.; Costa E Silva, S.; Dabić, M. The Evolving Role of Artificial Intelligence in Marketing: A Review and Research Agenda. J. Bus. Res. 2021, 128, 187–203. [Google Scholar] [CrossRef] [Scilit]
  39. Wang, S.; Wang, F.; Zhu, Z.; Wang, J.; Tran, T.; Du, Z. Artificial Intelligence in Education: A Systematic Literature Review. Expert Syst. Appl. 2024, 252, 124167. [Google Scholar] [CrossRef] [Scilit]
  40. Pham, D.T.; Alcock, R.J. Artificial Intelligence Based Techniques for Processing Segmented Images of Wood Boards. Proc. Inst. Mech. Eng. Part E J. Process Mech. Eng. 1998, 212, 119–129. [Google Scholar] [CrossRef] [Scilit]
  41. KeyWords Plus Generation, Creation, and Changes. Available online: https://support.clarivate.com/ScientificandAcademicResearch/s/article/KeyWords-Plus-generation-creation-and-changes?language=en_US (accessed on 27 March 2026).
  42. Nasir, V.; Cool, J.; Sassani, F. Acoustic Emission Monitoring of Sawing Process: Artificial Intelligence Approach for Optimal Sensory Feature Selection. Int. J. Adv. Manuf. Technol. 2019, 102, 4179–4197. [Google Scholar] [CrossRef] [Scilit]
  43. Gao, M.; Song, P.; Wang, F.; Liu, J.; Mandelis, A.; Qi, D. A Novel Deep Convolutional Neural Network Based on ResNet-18 and Transfer Learning for Detection of Wood Knot Defects. J. Sens. 2021, 2021, 4428964. [Google Scholar] [CrossRef] [Scilit]
  44. Chen, L.-C.; Pardeshi, M.S.; Lo, W.-T.; Sheu, R.-K.; Pai, K.-C.; Chen, C.-Y.; Tsai, P.-Y.; Tsai, Y.-T. Edge-Glued Wooden Panel Defect Detection Using Deep Learning. Wood Sci. Technol. 2022, 56, 477–507. [Google Scholar] [CrossRef] [Scilit]
  45. Przybyś-Małaczek, A.; Antoniuk, I.; Szymanowski, K.; Kruk, M.; Kurek, J. Application of Machine Learning Algorithms for Tool Condition Monitoring in Milling Chipboard Process. Sensors 2023, 23, 5850. [Google Scholar] [CrossRef] [Scilit]
  46. Szwajka, K.; Zielińska-Szwajka, J.; Trzepieciński, T. The Use of a Radial Basis Function Neural Network and Fuzzy Modelling in the Assessment of Surface Roughness in the MDF Milling Process. Materials 2023, 16, 5292. [Google Scholar] [CrossRef] [Scilit]
  47. Alsakka, F.; El-Chami, I.; Yu, H.; Al-Hussein, M. Computer Vision-Based Process Time Data Acquisition for Offsite Construction. Autom. Constr. 2023, 149, 104803. [Google Scholar] [CrossRef] [Scilit]
  48. Verly Lopes, D.J.; Bobadilha, G.D.S.; Grebner, K.M. A Fast and Robust Artificial Intelligence Technique for Wood Knot Detection. BioResources 2020, 15, 9351–9361. [Google Scholar] [CrossRef] [Scilit]
  49. Tran, V.; Lam, D.; Le, T. VNWoodKnot: A Benchmark Image Dataset for Wood Knot Detection and Classification. Data Brief 2025, 62, 112039. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  50. Fan, C.; Zhuang, Z.; Liu, Y.; Yang, Y.; Zhou, H.; Wang, X. Bilateral Defect Cutting Strategy for Sawn Timber Based on Artificial Intelligence Defect Detection Model. Sensors 2024, 24, 6697. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  51. Ji, M.; Zhang, W.; Diao, X.; Wang, G.; Miao, H. Intelligent Automation Manufacturing for Betula Solid Timber Based on Machine Vision Detection and Optimization Grading System Applied to Building Materials. Forests 2023, 14, 1510. [Google Scholar] [CrossRef] [Scilit]
  52. Li, R.; Xu, Z.; Yang, F.; Yang, B. Defect Detection for Melamine-Impregnated Paper Decorative Particleboard Surface Based on Deep Learning. Wood Mater. Sci. Eng. 2026, 21, 379–392. [Google Scholar] [CrossRef] [Scilit]
  53. Zhou, H.; Xia, H.; Fan, C.; Lan, T.; Liu, Y.; Yang, Y.; Shen, Y.; Yu, W. Intelligent Detection Method for Surface Defects of Particleboard Based on Super-Resolution Reconstruction. Forests 2024, 15, 2196. [Google Scholar] [CrossRef] [Scilit]
  54. Chun, T.H.; Hashim, U.R.; Ahmad, S. Timber Defect Identification: Enhanced Classification with Residual Networks. Int. J. Adv. Comput. Sci. Appl. 2024, 15, 665–671. [Google Scholar] [CrossRef] [Scilit]
  55. Achatz, J.; Schubert, M. Optical Detection of Beetle-Related Indicators and Stem Quality in Roundwood Using Convolutional Neural Networks. J. Wood Sci. 2025, 71, 27. [Google Scholar] [CrossRef] [Scilit]
  56. Vijayalakshmi, S.; Mrudhula, S.; Ashok Kumar, V.; Agastin; Varun; Latha, A.M. Artificial Intelligence-Driven Timber Wood Defect Characterization from Terahertz Images. J. Nondestruct. Eval. 2024, 43, 116. [Google Scholar] [CrossRef] [Scilit]
  57. Yu, Y.; Subhani, M.; Hoshyar, A.N.; Li, J.; Li, H. Automated Health Condition Diagnosis of in Situ Wood Utility Poles Using an Intelligent Non-Destructive Evaluation (NDE) Framework. Int. J. Struct. Stab. Dyn. 2020, 20, 2042002. [Google Scholar] [CrossRef] [Scilit]
  58. Kato, S.; Wada, N.; Shiogai, K.; Tamaki, T.; Kagawa, T.; Toyosaki, R.; Nobuhara, H. Automatic Classification of Crack Severity from Cross-Section Image of Timber Using Simple Convolutional Neural Network. Appl. Sci. 2022, 12, 8250. [Google Scholar] [CrossRef] [Scilit]
  59. Xu, F.; Wu, Y.; Lin, H.; Liu, Y.; Wang, X.; Ross, R.J.; Tian, G. IPSO-VMD Based Signal Feature Extraction and Internal Defect Detection of Hardwood Logs through Acoustic Impact Test. NDT E Int. 2023, 139, 102942. [Google Scholar] [CrossRef] [Scilit]
  60. Vacek, O.; Gergeľ, T.; Bucha, T.; Gracovský, R.; Gejdoš, M. Automatic Wood Species Classification and Pith Detection in Log CT Images. Forests 2024, 15, 2207. [Google Scholar] [CrossRef] [Scilit]
  61. Murugesan, S.; Kayaroganam, P.; Natarajan, S.; Balasubramanian, L. Analysis and Optimization of the Delamination Factor in the Drilling of Wood-Thermo Plastic Composites: An Approach Based on Artificial Intelligence. J. Thermoplast. Compos. Mater. 2025, 08927057251391670. [Google Scholar] [CrossRef] [Scilit]
  62. Abramov, I.; Nikitin, Y.; Zorina, O.; Bozek, P.; Stepanov, P.; Stollmann, V. Monitoring of technical condition of motors and bearigs of woodworking equipment. Acta Fac. Xylologiae Zvolen Publica Slovaca 2014, 56, 97–104. [Google Scholar]
  63. Chaiprabha, K.; Chancharoen, R. Innovative Smart Drilling with Critical Event Detection and Material Classification. J. Manuf. Mater. Process. 2023, 7, 155. [Google Scholar] [CrossRef] [Scilit]
  64. Jegorowa, A.; Kurek, J.; Antoniuk, I.; Krupa, A.; Wieczorek, G.; Świderski, B.; Bukowski, M.; Kruk, M. Automatic Estimation of Drill Wear Based on Images of Holes Drilled in Melamine Faced Chipboard with Machine Learning Algorithms. Forests 2023, 14, 205. [Google Scholar] [CrossRef] [Scilit]
  65. Sieradzki, A.; Bednarek, J.; Jegorowa, A.; Kurek, J. Explainable AI (XAI) Techniques for Convolutional Neural Network-Based Classification of Drilled Holes in Melamine Faced Chipboard. Appl. Sci. 2024, 14, 7462. [Google Scholar] [CrossRef] [Scilit]
  66. Bardak, T.; Bardak, S. Prediction of Wood Density by Using Red-Green-Blue (RGB) Color and Fuzzy Logic Techniques. J. Polytech. 2017, 20, 979–984. [Google Scholar] [CrossRef] [Scilit]
  67. Jovic, S.; Golubovic, Z.; Stojanovic, J. Wood Bonding Strength Sensitivity Estimation and Power Consumption Prediction in Wood Machining Process by Artificial Intelligence Methods. Sens. Rev. 2017, 37, 444–447. [Google Scholar] [CrossRef] [Scilit]
  68. Josué, D.F.; Agnès Virginie, T.; Guy Edgar, N.; Bienvenu, K. Prediction of the Tensile Strength of a Tropical Wood Species Terminalia superba Assembled by Gluing: A Comparative Intelligent Study. Geomech. Geoengin. 2024, 19, 1113–1125. [Google Scholar] [CrossRef] [Scilit]
  69. Kiliç, K.; Karaman, İ.; Kiliç, İ.; Özcan, U. Prediction of Veneer Bonding Strength of Wood-Based Composites Through Soft-Computing Models. Drew. Pr. Nauk. Doniesienia Komun. 2024, 67, 214. [Google Scholar] [CrossRef] [Scilit]
  70. Miguel, E.P.; Melo, R.R.D.; Serenini Junior, L.; Menezzi, C.H.S.D. Using Artificial Neural Networks in Estimating Wood Resistance. Maderas Cienc. Tecnol. 2018, 20, 531–542. [Google Scholar] [CrossRef] [Scilit]
  71. Li, C.; Chen, X.; Zhang, L.; Wang, S. Determination of Coniferous Wood’s Compressive Strength by SE-DenseNet Model Combined with Near-Infrared Spectroscopy. Appl. Sci. 2022, 13, 152. [Google Scholar] [CrossRef] [Scilit]
  72. Shahi, S.K.; Dia, M.; Yan, P.; Choudhury, S. Developing and Training Artificial Neural Networks Using Bootstrap Data Envelopment Analysis for Best Performance Modeling of Sawmills in Ontario. J. Model. Manag. 2022, 17, 788–811. [Google Scholar] [CrossRef] [Scilit]
  73. Carratù, M.; Gallo, V.; Liguori, C.; Lundgren, J.; O’Nils, M.; Pietrosanto, A. Vision-Based System for Measuring the Diameter of Wood Logs. IEEE Open J. Instrum. Meas. 2023, 2, 5000112. [Google Scholar] [CrossRef] [Scilit]
  74. Alexandru Borz, S.; Păun, M. Integrating Offline Object Tracking, Signal Processing, and Artificial Intelligence to Classify Relevant Events in Sawmilling Operations. Forests 2020, 11, 1333. [Google Scholar] [CrossRef] [Scilit]
  75. Cheţa, M.; Marcu, M.V.; Iordache, E.; Borz, S.A. Testing the Capability of Low-Cost Tools and Artificial Intelligence Techniques to Automatically Detect Operations Done by a Small-Sized Manually Driven Bandsaw. Forests 2020, 11, 739. [Google Scholar] [CrossRef] [Scilit]
  76. Borz, S.A.; Forkuo, G.O.; Oprea-Sorescu, O.; Proto, A.R. Development of a Robust Machine Learning Model to Monitor the Operational Performance of Fixed-Post Multi-Blade Vertical Sawing Machines. Forests 2022, 13, 1115. [Google Scholar] [CrossRef] [Scilit]
  77. Sexton, J.-T.; Morin, M.; Georges, R.; Abasian, F.; Gaudreault, J. Wood Planer Control: Predictive and Prescriptive Approaches via Automatic State Matching Gaussian Processes. Eng. Appl. Artif. Intell. 2024, 132, 107843. [Google Scholar] [CrossRef] [Scilit]
  78. Haddouche, M.; Ilinca, A.; Chaouch, M. Enhancing Energy Efficiency in the Wood Furniture Sector Through Industry 4.0: Real-Time Implementation and Case Study. Energy Technol. 2026, 14, e202501297. [Google Scholar] [CrossRef] [Scilit]
  79. Sundarapandiyan, M.; Palanikumar, K.; Senthilkumar, N.; Latha, B. Artificial Intelligence-Based Optimization and Response Surface Analysis of Thrust Force and Torque in Drilling Wood–Thermoplastic Composite (WPC) Panels. Wood Mater. Sci. Eng. 2025, 1–19. [Google Scholar] [CrossRef] [Scilit]
  80. Raamets, T.; Karjust, K.; Hermaste, A.; Kelpman, K. Virtual Factory Model Development for AI-Driven Optimization in Manufacturing. Proc. Est. Acad. Sci. 2025, 74, 228–233. [Google Scholar] [CrossRef] [Scilit]
  81. Yeshiwas, T.A.; Siyoum, Y.B.; Tiruneh, A.B.; Bekele, A.A.; Nega, A.Y.; Adamu, T.K. Parameter Optimization for Enhancing Mechanical Properties of Wood-Plastic Composites Using Artificial Neural Network with Genetic Algorithm and Taguchi Method. Wood Mater. Sci. Eng. 2025, 1–16. [Google Scholar] [CrossRef] [Scilit]
  82. Melo, R.R.D.; Miguel, E.P. Use of artificial neural networks in predicting particleboard quality parameters. Rev. Árvore 2016, 40, 949–958. [Google Scholar] [CrossRef] [Scilit]
  83. Ericsson, M.; Johansson, D.; Stjern, D. AI-Based Quality Control of Wood Surfaces with Autonomous Material Handling. Appl. Sci. 2021, 11, 9965. [Google Scholar] [CrossRef] [Scilit]
  84. Ricardo, F.A.; Eizaguirre, M.; Moru, D.K.; Borro, D. Intelligent Manufacturing in Wine Barrel Production: Deep Learning-Based Wood Stave Classification. AI 2024, 5, 2018–2036. [Google Scholar] [CrossRef] [Scilit]
  85. Ruttico, P.; Bordoni, F.; Deval, M. Woodot: An AI-Driven Mobile Robotic System for Sustainable Defect Repair in Custom Glulam Beams. Sustainability 2025, 17, 5574. [Google Scholar] [CrossRef] [Scilit]
  86. Ramos-Maldonado, M.; Gutiérrez, F.; Gallardo-Venegas, R.; Bustos-Avila, C.; Contreras, E.; Lagos, L. Machine Learning and Industrial Data for Veneer Quality Optimization in Plywood Manufacturing. Processes 2025, 13, 1229. [Google Scholar] [CrossRef] [Scilit]
  87. Settimi, A.; Chutisilp, N.; Aymanns, F.; Gamerro, J.; Weinand, Y. TTool: A Supervised Artificial Intelligence-Assisted Visual Pose Detector for Tool Heads in Augmented Reality Woodworking. Appl. Sci. 2024, 14, 3011. [Google Scholar] [CrossRef] [Scilit]
  88. Explainable Artificial Intelligence|DARPA. Available online: https://www.darpa.mil/research/programs/explainable-artificial-intelligence (accessed on 27 March 2026).
  89. Tikhomirov, O.K. Philosophical and Psychological Problems of Artificial Intelligence. In Proceedings of the 4th International Joint Conference on Artificial Intelligence—Volume 1; Morgan Kaufmann Publishers Inc.: San Francisco, CA, USA, 1975; pp. 932–937. [Google Scholar]
  90. China’s AI Ambitions Target US Tech Dominance. Available online: https://www.dw.com/en/china-ai-artificial-intelligence-deepseek-us-chatgpt-semiconductors-graphics-technology-v2/a-74361630 (accessed on 27 March 2026).
  91. Almada-Lobo, F. The Industry 4.0 Revolution and the Future of Manufacturing Execution Systems (MES). J. Innov. Manag. 2016, 3, 16–21. [Google Scholar] [CrossRef] [Scilit]
Forests 17 00477 g001
Figure 1. Review flow diagram for the literature selection.
Figure 1. Review flow diagram for the literature selection.
Forests 17 00477 g001
Forests 17 00477 g002
Figure 2. Distribution of publications by year.
Figure 2. Distribution of publications by year.
Forests 17 00477 g002
Forests 17 00477 g003
Figure 3. Word cloud of author keywords.
Figure 3. Word cloud of author keywords.
Forests 17 00477 g003
Forests 17 00477 g004
Figure 4. Co-occurrence network of author keywords.
Figure 4. Co-occurrence network of author keywords.
Forests 17 00477 g004
Table 2. Leading journals by number of publications and citations.
Table 2. Leading journals by number of publications and citations.
SourceDocumentCitation
Forests751
Applied Sciences513
Wood Material Science and Engineering47
Sensors324
Note: Journals with one publication: ACTA Facultatis Xylologiae Zvolen, AI, Automation in Construction, Bioresources, Data in Brief, Drewno, Drvna Industrija, Energy Technology, Engineering Applications of AI, Geomechanics and Geoengineering—An International Journal, IEEE Access, IEEE Open Journal of Instrumentation and Measurement, International Journal of Advanced Computer Science and Applications, International Journal of Advanced Manufacturing Technology, International Journal of Structural Stability and Dynamics, Journal of Manufacturing and Materials Processing, Journal of Modelling in Management, Journal of Nondestructive Evaluation, Journal of Polytechnic—Politeknik Dergisi, Journal of Sensors, Journal of Thermoplastic Composite Materials, Journal of Wood Science, Maderas—Ciencia y Tecnologia, Materials, NDT and E International, Proceedings of the Estonian Academy of Sciences, Processes, Revista Arvore, Sensor Review, Sustainability, and Wood Science and Technology.
Table 3. Most cited publications based on WoS data.
Table 3. Most cited publications based on WoS data.
ReferencesJournalTotal CitationsAvg. Per Year
[42]International Journal of Advanced Manufacturing Technology547.71
[43]Journal of Sensors326.40
[44]Wood Science and Technology287.00
[45]Sensors186.00
[46]Materials175.67
[47]Automation in Construction175.67
Table 4. Country-level scientific production based on author appearances.
Table 4. Country-level scientific production based on author appearances.
FrequencyCountry
18China
16Canada
12Poland
9India
5Brazil, Malaysia, Slovakia
4Cameroon, Turkey
3Australia, Italy, Japan, Romania, Russia, Spain, Switzerland, USA
2Sweden
1Chile, Estonia, Ethiopia, Nigeria, Serbia, Slovenia, Thailand, United Kingdom, Vietnam
Table 5. Distribution of publications by corresponding author countries.
Table 5. Distribution of publications by corresponding author countries.
Number of PublicationsCountries
9China
5Canada, Poland
3India, Romania
2Brazil, Italy, Malaysia, Switzerland, Turkey
1Australia, Cameroon, Chile, Estonia, Ethiopia, Japan, Russia, Serbia, Slovakia, Slovenia, Spain, Sweden, Thailand, USA, Vietnam
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Kucuk, E.; Buehlmann, U. Systematic Review of Applications Using Artificial Intelligence (AI) for Wooden Materials. Forests 2026, 17, 477. https://doi.org/10.3390/f17040477

AMA Style

Kucuk E, Buehlmann U. Systematic Review of Applications Using Artificial Intelligence (AI) for Wooden Materials. Forests. 2026; 17(4):477. https://doi.org/10.3390/f17040477

Chicago/Turabian Style

Kucuk, Enis, and Urs Buehlmann. 2026. "Systematic Review of Applications Using Artificial Intelligence (AI) for Wooden Materials" Forests 17, no. 4: 477. https://doi.org/10.3390/f17040477

APA Style

Kucuk, E., & Buehlmann, U. (2026). Systematic Review of Applications Using Artificial Intelligence (AI) for Wooden Materials. Forests, 17(4), 477. https://doi.org/10.3390/f17040477

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

Article Metrics

Back to TopTop