Application of Value of Information-Based Approaches in Road Inspection Processes and Asset Management: A Literature Review

Abstract

Modern road infrastructure asset management faces increasing pressure to improve the quality of decision-making processes, also due to limited public resources. The field of road diagnostics is no exception. The aim of the research is to analyze, through a literature review, the possibilities of applying the theoretical concept of information value. The selected point of interest is the tasks associated with the selection of specific sections intended for inspection, monitoring the level of information gain that this inspection can bring. Methodologically, the research is based on a systematic bibliometric analysis of the literature from the Web of Science and SCOPUS databases for the period January 2010 to June 2025. This is supplemented by a non-systematic content review, while the identified publications were processed by the Bibliometrix and VOSviewer tools and subsequently qualitatively interpreted. The result of the research is a synthesis of knowledge from the finally analyzed set of relevant scientific papers. The findings point to a growing interest in linking the process of planning and performing road infrastructure diagnostics with asset management decision-making processes. At the same time, they point to the development of data-oriented and digital approaches, as well as the limited application of the concept of information value in planning inspections before their implementation. The findings indicate that the assessment of expected information benefit represents a promising tool for reducing uncertainty, determining priorities, and allocating resources more efficiently, while its implementation in road infrastructure management requires further methodological research and practical verification.

1. Introduction

Current developments in the field of road infrastructure asset management indicate the need to re-evaluate some traditional approaches and assess them from new methodological perspectives [1,2,3,4]. Increasing requirements for the safety, reliability and economy of road network operation, as well as pressure for efficient use of public resources, lead to a growing emphasis on improving the quality of decision-making processes in the field of maintenance, renewal and modernization of these large assets [5,6,7,8]. The field of determining the structural and technical condition of assets through inspections also has a role to play in these tasks [9,10]. The diagnostic process itself should be perceived not only as a technical activity aimed at identifying failures, but also as an integral part of the decision-making framework on the part of road administrators. It is undeniable that information obtained through inspections represents a key input into planning maintenance interventions, determining priorities and assessing risks [11,12]. Recently, research has focused on the limits of data in terms of uncertainty, different quality and subjectivity of the obtained data. The current scientific literature also documents the dynamic development of diagnostic technologies, data approaches and analytical tools used in monitoring transport infrastructure, including non-destructive methods, sensor systems or machine learning methods [13,14,15]. A set of questions is gradually being developed, what value individual diagnostic information has for decision-making itself and how it can be systematically integrated into asset management processes [16,17]. Nevertheless, in practice, there is always a new demand for generating approaches that would enable a better connection of diagnostics with management and planning processes within the asset management of the road network.

A topic that has not been developed yet is the idea of assessing the level of information benefit obtained by performing diagnostics before the actual performance. Such an approach would allow assessing the extent to which a specific inspection can reduce uncertainty about the technical condition of the infrastructure and determine priorities for the inspection of the construction and technical condition. In practice, it can be assumed that not all sections of the road network provide the same information value after the inspection and that some of them can bring significantly more relevant knowledge for decision-making than others.

2. Background

One of the promising directions may be to consider introducing a concept based on the theory of information value, which allows quantifying the benefit of the obtained data for decision-making. At the theoretical level, this is an approach that is used in various areas of decision-making under conditions of uncertainty, while its application in the field of road infrastructure management can represent a potential tool for optimizing the scope and frequency of inspections, determining intervention priorities and more efficient allocation of resources [18,19,20]. Taking into account the expected information value of individual inspections could thus represent an additional criterion in determining priorities and planning diagnostic activities.

Considering which sections have the potential to bring more significant information and which, on the contrary, only limited benefit, opens up space for more efficient use of financial, technical and human resources and a more systematic connection of diagnostics with road infrastructure asset management processes. The application of this approach could contribute to increasing the transparency and quality of decision-making processes, but their application in this area cannot yet be considered sufficiently researched or methodologically established. On this basis, it can be assumed that the concept of information value has the potential to become a relevant analytical framework for decision support in the field of inspection planning and road infrastructure management. This is primarily the case in situations characterized by uncertainty about the actual condition of assets. At the same time, it can be assumed that existing research addresses this issue only partially and does not yet offer a unified methodological approach to assessing the information benefit of diagnostic activities before their implementation. Another hypothesis is the assumption that approaches based on information value probably appear in the literature indirectly. It is estimated that this happens through models supporting decision-making, prioritization or uncertainty analysis and not as an explicitly developed framework applied to road infrastructure management.

Based on this, there is a need for a deeper examination of existing research focused on the use of the concept of information value in the context of road infrastructure inspection and management. For the above reasons, the aim of this work is to carry out a systematic research search focused on the identification and analysis of scientific works that deal with the application of approaches based on the value of information in the processes of diagnostics and asset management. The work focuses on examining whether and to what extent these approaches appear in existing research, what methodological frameworks they use, and what their potential is for supporting decision-making, prioritizing, and optimizing infrastructure asset management processes. The result is a synthesized overview of knowledge that contributes to a better understanding of the possibilities of using the concept of information value in this area and identifying directions for further research. The research formulates partial objectives based on the trends and hypotheses identified above. Attention is therefore focused on assessing how the connection between road infrastructure diagnostics and asset management decision-making processes is developing in the monitored period. Efforts are being made to identify the dominant methodological and technological approaches applied in this area, in particular, data-oriented, sensor, digitalization and modeling tools used in monitoring, condition assessment and decision-making support. At the same time, the research focuses on assessing the extent to which the concept of information value appears in existing research, whether explicitly or implicitly through approaches focused on working with uncertainty, prioritizing interventions and optimizing resource allocation. Special attention is paid to identifying methodological limitations and research gaps that currently prevent a more systematic use of the concept of information value in planning inspections before they are carried out. On this basis, the research seeks to assess whether the assessment of the expected information benefit of inspections can be considered a promising tool for reducing uncertainty, setting priorities, and more effectively supporting decision-making in road infrastructure management.

3. Methodology

The methodological approach of the research is conceived on two mutually complementary levels, which make it possible to capture the issue under investigation from both a quantitative and a qualitative–interpretative perspective. This dual framework enables the research not only to map the broader structure of scientific knowledge in the selected field, but also to interpret its substantive content, theoretical foundations, and practical implications in greater depth. The first level is based on a systematic literature review carried out through a targeted bibliometric analysis of scientific sources indexed in the Web of Science and Scopus databases, with the aim of identifying the scope, structure and development of scientific knowledge in the research area. This level makes it possible to identify publication trends, dominant thematic orientations and the evolution of scientific interest over time. The second level represents a non-systematic literature review, oriented towards a content synthesis of key theoretical and empirical knowledge. This approach is without strictly defined search criteria, which allows for a broader analysis of the issue at hand and to build on existing scientific discourses. In contrast to the systematic level, this part of the research allows for greater flexibility in incorporating conceptually relevant sources that may not be captured through formally defined bibliographic procedures alone. The procedure for choosing and selecting publications is also presented through a PRISMA flow diagram (Figure 1), which transparently shows the individual phases of identification, sorting, assessment and final inclusion of sources in the analytical file.

3.1. Systematic Analysis

The methodology of bibliographic research is based on a multi-step systematic procedure, which can be summarized as follows:

• Selection of databases, time period and definition of search criteria;
• Data collection through scientific databases;
• Optimization of search strings with an emphasis on the relevance of the outputs;
• Export of the obtained results and their processing into analytical form;
• Analytical and interpretative processing of the results.

The monitored period was set in the interval from January 2010 to June 2025, and this time frame is fully sufficient in terms of the development of knowledge and publication activity in the given area. It includes more than fifteen years of research, during which there was a significant development of approaches to road infrastructure diagnostics, digitalization of processes and application of data-oriented methods in decision-making. The chosen interval thus allows us to capture the main trends, methodological shifts and the current direction of scientific discourse. Subsequently, the identified sources were processed, sorted, analyzed and evaluated, as well as the subsequent interpretation of the results and the formulation of conclusions. This process included bibliometric data processing, content analysis of publications and synthesis of knowledge, which required adequate time to ensure the quality, consistency and professional relevance of the research outputs.

The bibliometric analysis uses the Web of Science and Scopus databases, which are considered the most reliable and comprehensive sources of indexed scientific publications for the relevant scientific field. The specifics of the search strings for both databases are listed in

Table 1. Descriptions of formulated query strings.

Database	Search String
Web of Science	((((TS = (road)) OR TS = (pavement))) AND TS = (inspection)) AND TS = (value) and 2025 or 2024 or 2023 or 2022 or 2021 or 2020 or 2019 or 2018 or 2017 or 2016 or 2015 or 2014 or 2013 or 2011 or 2010 or 2012 (Publication Years) and Article or Proceeding Paper (Document Types) and English (Languages) and Article or Proceeding Paper (Document Types) and Engineering or Construction Building Technology or Materials Science or Transportation or Science Technology Other Topics or Remote Sensing (Research Areas)
SCOPUS	(TITLE-ABS-KEY (road) OR TITLE-ABS-KEY (pavement) AND TITLE-ABS-KEY (inspection) AND TITLE-ABS-KEY (value)) AND PUBYEAR > 2009 AND PUBYEAR < 2026 AND (LIMIT-TO (PUBSTAGE, “final”)) AND (LIMIT-TO (DOCTYPE, “ar”) OR LIMIT-TO (DOCTYPE, “cp”)) AND (LIMIT-TO (SUBJAREA, “ENGI”) OR LIMIT-TO (SUBJAREA, “SOCI”) OR LIMIT-TO (SUBJAREA, “MATE”) OR LIMIT-TO (SUBJAREA, “ENVI”)) AND (LIMIT-TO (LANGUAGE, “English”)) AND (LIMIT-TO (SRCTYPE, “j”) OR LIMIT-TO (SRCTYPE, “p”))

.

Based on the defined search criteria, 281 records were identified in the Web of Science database, of which 202 were articles published in journals. The Scopus database provided 342 outputs, of which 206 were journal articles. The development of the number of publications and citations in the period January 2010–June 2025 (Figure 2) indicates a growing trend of interest in the given research topic. After a relatively stable period between 2010 and 2017, the number of publications has started to increase significantly since 2018, with the trend culminating in the years 2022 to 2024 with more than 30 documents per year. This development is accompanied by a dynamic increase in the number of citations, which confirms the increasing relevance and topicality of the research issue. The decrease in 2025 can be attributed to the time limit of data collection (June 2025). Overall, it can be stated that the research topic under study shows the character of a dynamically developing scientific trend.

Table 2. Overview of publications in the Web of Science and SCOPUS databases by research area and publishing house.

SCOPUS				Web of Sciences
Research Area	Nr. of Publications	Publishing House	Nr. of Publications	Research Area	Nr. of Publications	Publishing House	Nr. of Publications
Engineering	280	Elsevier	56	Engineering Civil	132	Elsevier	58
Computer Science	80	MDPI	36	Construction Building Technology	62	MDPI	42
Social Sciences	68	Amer Soc Civil Engineers	30	Transportation Science Technology	49	Taylor & Francis	24
Materials Science	63	IEEE	17	Materials Science Multidisciplinary	41	Asce-Amer Soc Civil Engineers	18
Environmental Science	52	Taylor & Francis	15	Engineering Multidisciplinary	39	Springer Nature	16
Physics and Astronomy	41	SPIE-INTERNATIONAL SOCIETY FOR OPTICAL ENGINEERING	14	Engineering Electrical Electronic	29	IEEE	14
Earth and Planetary Sciences	37	CRC press	13	Transportation	24	Amer Soc Civil Engineers	13
Mathematics	27	Sage	10	Engineering Mechanical	18	SPIE-INTERNATIONAL SOCIETY FOR OPTICAL ENGINEERING	9
Energy	21	Emerald Group Publishing	6	Materials Science Characterization Testing	18	Emerald Group Publishing	8
Chemistry	10	Springer	4	Environmental Sciences	17	Sage	8
Biochemistry, Genetics and Molecular Biology	9			Instruments Instrumentation	15	Trans Tech Publications Ltd.	8
Business, Management and Accounting	9			Remote Sensing	15	Iop Publishing Ltd.	6

provides an overview of publications indexed in Scopus and Web of Science by research area and publisher.

Table shows the number of publications in the main thematic categories and identifies the publishers with the highest representation in each database. The table also shows that the thematic structure of publications in both databases is somewhat different, with Scopus showing a broader representation of technical and interdisciplinary areas. Web of Science is more oriented towards specific engineering and technological directions.

The

Table 3. Overview of citations in the SCOPUS and Web of Science databases.

Article Title	Number of Citations		Rank
	SCOPUS	Web of Sciences	SCOPUS	Web of Sciences
Nhat-Duc, H.; Nguyen, Q.-L.; Tran, V.-D. Automatic Recognition of Asphalt Pavement Cracks Using Metaheuristic Optimized Edge Detection Algorithms and Convolution Neural Network. Automation in Construction 2018, 94, 203–213, doi:10.1016/j.autcon.2018.07.008.	318	269	1	1
Ji, A.; Xue, X.; Wang, Y.; Luo, X.; Xue, W. An Integrated Approach to Automatic Pixel-Level Crack Detection and Quantification of Asphalt Pavement. Automation in Construction 2020, 114, 103176, doi:10.1016/j.autcon.2020.103176. [21]	230	198	2	2
Cafiso, S.; Di Graziano, A.; Di Silvestro, G.; La Cava, G.; Persaud, B. Development of Comprehensive Accident Models for Two-Lane Rural Highways Using Exposure, Geometry, Consistency and Context Variables. Accident Analysis & Prevention 2010, 42, 1072–1079, doi:10.1016/j.aap.2009.12.015.	193	166	3	3
Akagić, A.; Buza, E.; Omanovic, S.; Karabegović, A. Pavement Crack Detection Using Otsu Thresholding for Image Segmentation. 41st international convention on information and communication technology, electronics and microelectronics (MIPRO) 2018 (pp. 1092–1097). IEEE. 10.23919/MIPRO.2018.8400199	136	108	4	6
Yılmaz, B.; Asyalı, M.H.; Arıkan, E.; Yetkin, S.; Özgen, F. Sleep Stage and Obstructive Apneaic Epoch Classification Using Single-Lead ECG. BioMedical Engineering OnLine 2010, 9, 39, doi:10.1186/1475-925x-9-39.	130	101	5	7
Kanafi, M.M.; Kuosmanen, A.; Pellinen, T.K.; Tuononen, A.J. Macro- and Micro-Texture Evolution of Road Pavements and Correlation with Friction. International Journal of Pavement Engineering 2014, 16, 168–179, doi:10.1080/10298436.2014.937715.	106	92	6	8
Cardellicchio, A.; Ruggieri, S.; Nettis, A.; Renò, V.; Uva, G. Physical Interpretation of Machine Learning-Based Recognition of Defects for the Risk Management of Existing Bridge Heritage. Engineering Failure Analysis 2023, 149, 107237, doi:10.1016/j.engfailanal.2023.107237.	82	75	7	9
Shahnazari, H.; Tutunchian, M.A.; Mashayekhi, M.; Amini, A.A. Application of Soft Computing for Prediction of Pavement Condition Index. Journal of Transportation Engineering 2012, 138, 1495–1506, doi:10.1061/(asce)te.1943-5436.0000454.	80	68	8	10
Biçici, S.; Zeybek, M. An Approach for the Automated Extraction of Road Surface Distress from a UAV-Derived Point Cloud. Automation in Construction 2020, 122, 103475, doi:10.1016/j.autcon.2020.103475.	62	47	9
Munawar, H.S.; Ullah, F.; Shahzad, D.; Heravi, A.; Qayyum, S.; Akram, J. Civil Infrastructure Damage and Corrosion Detection: An Application of Machine Learning. Buildings 2022, 12, 156. https://doi.org/10.3390/buildings12020156 [22]	59		10
Tsangaratos, P.; Ilia, I.; Hong, H.; Chen, W.; Xu, C. Applying Information Theory and GIS-Based Quantitative Methods to Produce Landslide Susceptibility Maps in Nancheng County, China. Landslides 2016, 14, 1091–1111, doi:10.1007/s10346-016-0769-4.	158	149		4
Okujeni, A.; Van Der Linden, S.; Tits, L.; Somers, B.; Hostert, P. Support Vector Regression and Synthetically Mixed Training Data for Quantifying Urban Land Cover. Remote Sensing of Environment 2013, 137, 184–197, doi:10.1016/j.rse.2013.06.007.	60	131		5

allows to identify the most cited scientific works in the monitored area and at the same time compare their impact according to the Scopus and Web of Science databases.

The publication focused on the automatic detection of road cracks using artificial intelligence algorithms and neural networks achieved the highest number of citations in both databases, which ranked first in both rankings. Works focused on the automated detection of road damage, road safety modeling and road condition assessment are also in the leading positions. This indicates a strong research emphasis on digitization, image processing and data-oriented approaches. Despite the demonstrable agreement in the databases regarding the most cited publications, in some cases, their order differs. Some works were placed in only one database or achieved significantly different positions, which reflects the different indexing coverage and citation background of individual databases. From a content perspective, it is clear that the most cited publications focus mainly on the use of artificial intelligence, machine learning, image analysis and advanced modeling approaches in the field of transport infrastructure diagnostics. Works directly using the concept of information value are less represented, indicating the potential for further research in this area.

The outputs from the Scopus and Web of Science databases were subsequently merged and processed in the RStudio software version 2025.08.0 environment using the Bibliometrix package, which ensured the removal of duplicates and the creation of a single dataset containing 390 unique records. Bibliometrix is an open-source tool for quantitative research in the field of scientometrics and bibliometrics, which allows the application of the main analytical methods, as well as the standardization of data formats. The tool also includes the Biblioshiny interface, designed for statistical processing and graphic visualization of data. For the purposes of network and thematic analysis, the VOSviewer software tool version 1.6.20 was also used, which allows the identification of relationships between authors, keywords and citation links. Subsequently, a qualitative selection of publications was carried out through the analysis of the titles and abstracts of individual articles. The aim of this step was to identify contributions that thematically correspond most to the objectives and research questions of the work.

3.2. Non-Systematic Analysis

A non-systematic literature review complements a systematic bibliometric analysis with a broader theoretical context of the research issue. Its goal is not to quantitatively map publication activity, but to identify and synthesize key concepts, approaches, theories, and empirical findings that have shaped the development of a given research area. The selection of sources within a non-systematic review is not tied to a strictly defined search algorithm or database limitations. The analysis includes, in particular, the following:

• Monographs and chapters in professional books;
• Review and theoretical studies;
• Significant research with a high citation response;
• Documents from international organizations and professional institutions;
• Relevant national sources and legislative documents, if related to the subject of the research.

Methodologically, this review is based on an analytical–synthetic comparison of the approaches of individual authors, their methodological starting points and the presentation of the main research results. The emphasis is placed on identifying dominant research directions, comparing different approaches to solving the problem under study and identifying insufficiently researched areas and open research questions. A non-systematic review also allows connecting the results of a systematic bibliometric analysis with a content interpretation of knowledge. It serves as a tool for understanding broader contexts that cannot be captured exclusively by quantitative methods and contributes to formulating the conclusions of this work. The output of this part is a synthesized review of the most important theoretical and empirical knowledge, which identifies results, research gaps and problematic areas.

4. Results

4.1. Results of the Systematic Analysis

The result of the selection process was a final set of 82 scientific publications, which were evaluated as relevant and subsequently subjected to in-depth content and interpretive analysis.

Figure 3 presents the summary metric obtained after processing the selected set of scientific publications. It contains key indicators quantifying the scope, structure and basic characteristics of the analyzed publication set. This provides a general picture of the state of the publication base in the monitored area, which is useful in assessing the relevance and importance of the research issue in the scientific literature. Of the total number of 82 analyzed publications, 20.73% were created within the framework of international co-authorship, which indicates the active involvement of researchers in international research networks and the ability of the research issue to go beyond the national and regional framework. This value can be considered significant in the field of applied research on the topics of transport infrastructure management, as it reflects the need for the exchange of data, methodologies and experiences between countries with different technical standards, legislative frameworks, climatic conditions and organizational approaches to road network management. The average number of 4.1 co-authors per publication indicates a predominantly team-based nature of research, typical of interdisciplinary studies that combine knowledge from civil and transport engineering, information technology and decision-making methods. The year-on-year growth rate of publication activity at 7.62% documents the dynamic development of research interest in the topic under study and its gradual establishment as a relevant research area. This trend is in line with the growing demands for effective transport infrastructure management, the need for ever-wider digitalization of diagnostic processes and the implementation of data-based decision-making tools.

The average age of publications of 3.22 years points to the high topicality of the analyzed publication set, indicating that a significant part of the scientific production was created in the recent period and thus reflects current technological and methodological trends. This fact is also supported by the average number of 11.17 citations per publication. This indicates an adequate scientific response and indicates that the published results are actively used and developed in further research. The total number of 2593 references also document the broad theoretical and methodological background of the analyzed works, as well as the connection of the researched issue with several related scientific disciplines.

Figure 4 shows the substantive trend and temporal continuity of keywords from selected publications in the areas defined by the selection criteria in three time periods: 2017–2021, 2022–2023 and 2024–2025. In the period 2017–2021, keywords such as asphalt mixtures, geological surveys, maintenance and pavement dominated. These formed the basis from which new and more specific words of interest were gradually formed. Asphalt mixtures were transformed into the more general term asphalt, which became the central theme. Maintenance developed towards a more modern approach designated as condition-based maintenance. Also new in this period were keywords such as crack detection and the use of laser technologies, which points to the growing use of technologies and sensory approaches in road monitoring. The keyword pavement persisted and remained an important part in the following period. In the last monitoring period, 2024–2025, the keyword asphalt evolved into the more specific term asphalt pavements. Condition-based maintenance remained a key theme. Geological surveys and crack detection reappear, confirming their importance in assessing the technical condition of the infrastructure. The new keyword is highway management, which may signal a change in trend towards an interest in improving the organization and management of transport infrastructure. The diagram shows a gradual transition from basic concepts such as materials, maintenance, infrastructure to specialized technologies and strategic road management, with an emphasis on digitalization, modern diagnostics and efficient infrastructure management.

Based on multivariate statistical analysis, a projection of keywords (Figure 5) was created, which shows their relationships in two main dimensions. The first dimension (Dim 1) representing the main difference direction on the horizontal axis explains 30.66% of the variance, while the second dimension (Dim 2) representing the secondary difference aspect on the vertical axis covers 13.22% of the variance.

Moreover, the more variance a dimension explains, the more informative it is. Based on the location of individual keywords in the quadrants, it is possible to deduce their thematic proximity, which can serve as a basis for formulating new research objectives or interdisciplinary hypotheses. From the map, it is possible to recognize the main research areas defined by keywords such as geological surveys, radar measurements and non-destructive testing methods. The upper right quadrant is represented by topics such as maintenance, quality control, crack detection, machine learning, data processing and sensor systems, which reflect the connection of civil engineering with digital and computational technologies. These keywords represent modern approaches to road infrastructure management with an emphasis on automation, data analytics and predictive modeling.

Figure 6 shows a graph representing a strategic map of keywords, used to illustrate their importance and level of development in a scientific field. The graph is divided into four quadrants according to two main axes: centrality (Relevance degree) on the horizontal axis and development (Development degree) on the vertical axis. More central concepts are those that are more interconnected with other keywords, while more developed topics have a deeper level of processing and research depth. In the upper-right quadrant are the so-called carrier topics, which are also well developed (e.g., crack detection, inspection, highway management, georadar). The upper-left quadrant represents areas defined by keywords that are highly developed but less interconnected with others (radar measurement, drone use). In the lower-right quadrant are the basic topics. These defined areas are very relevant, but are not yet fully developed and their development potential is long-term (roads). Conversely, the lower-left quadrant shows emerging or declining areas defined by keywords (e.g., road infrastructure, failures, condition monitoring).

The Sankey diagram (Figure 7) shows the connection between research areas defined by keywords, the countries of origin of the authors and the scientific journals or conferences where the outputs are published. The diagram thus allows identifying not only global research priorities, but also regional specializations and preferences in the publication strategies of individual countries.

Based on the data presented, it is clear that in the field of crack detection and computer processing, authors in China, Japan and the USA are extremely active. The given outputs are often published in technology-oriented journals such as Sensors or Automation in Construction. From the overall overview, it is clear that China dominates as the country with the widest spectrum of research topics, which are also applied in various international scientific journals.

By analyzing the frequency of co-occurrence of keywords in publications, a network diagram with nodes and links was generated, presented in Figure 8. This approach makes it easier to understand the structure of the relevant research area, identify missing connections, or assess the potential for new collaborations. The size of the node reflects the intensity of the occurrence of the given keyword, the thickness of the connecting lines represents the strength of the connection, and the colored clusters express thematic affinity. Several conclusions can be drawn from the visualized network map, particularly regarding intelligent infrastructure monitoring and its maintenance. The subject areas are located at the intersection of several disciplines and thus connect knowledge and methods from information technology, civil engineering, and the field of transport infrastructure. The dominant research directions defined by keywords are mainly focused on the use of modern computing and data technologies. The most frequent and most interconnected key terms include machine learning, crack detection, and structural health monitoring. Emphasis is also placed on digitization, sensor systems, automatic fault detection and advanced data analytics. This indicates that research is increasingly focused on predictive and autonomous systems that can significantly improve the efficiency of transport infrastructure operations and management. A notable area that cannot be overlooked is structural condition monitoring, which deals with the continuous monitoring of the technical condition of bridge structures and roads. This area is linked to research on technical condition assessment, which often uses outputs from sensor systems, visual inspections or satellite and drone data. The diagram also points to the diversity of topics that are part of the given research spectrum, from visual damage detection, through image data processing, to the design of decision-making algorithms for maintenance planning. This creates thematic clusters composed of concepts that together form logically and technically connected research areas.

Figure 9 shows the occurrence of keywords using a color spectrum. Blue shades indicate older research directions, while red signals current or growing trends. It is clear that concepts (or areas of interest expressed by keywords) such as “pavement”, “pavement management”, “condition assessment”, or “maintenance” belong to traditional areas of research and appeared in professional outputs several years ago. They represent basic pillars and form a long-term stable base of research interest. At the center of the visualization are concepts such as “machine learning”, “deep learning”, “neural network” and “health monitoring”, which are depicted in neutral or transitional colors. These terms represent technologies that have become more prominent in recent years. They bring a wider use of automation tools, intelligent data processing and prediction models to the field of infrastructure management. Their presence in the central part of the network points to their key role as a connecting element between classic and new topics. The latest and fastest-growing research directions are shown in shades of red. For example, these are terms such as “damage detection”, “visual inspection”, “sensor fusion”, “optimization” or “urban environment”. These keywords represent current research trends that focus on increasing the accuracy of damage detection, integrating various sensor data and applying advanced next-generation neural networks. They are characterized by higher technical complexity, multidisciplinary overlap and high innovation potential.

Table 4. Content overview of the resulting selection of the systematic literature review.

Topic area: Road condition and quality indicators
[23]	The research examines the relationship between different types of pavement damage and their roughness, as measured by the IRI. The strongest influences on the increase in the IRI are network cracks, permanent deformations and transverse cracks. The study also highlights that not all defects that affect the IRI are equally captured by traditional visual inspections or PCI assessments.
[24]	This study explores an innovative approach to monitoring road quality using smartphones, specifically through IRI measurement. The results of the analysis showed that the application is able to generate consistent IRI values over repeated passes over the same sections. Moreover, these values showed a strong correlation with data measured by official measurement vehicles.
[25]	The article deals with the development and implementation of a complex road damage detection system, which is based on processing a large amount of data obtained from multiple types of sensors. The authors developed a detection algorithm based on the decision tree classification method.
[26]	The authors present a method that uses vertical acceleration recorded by sensor systems in connected vehicles. These data reflect the vehicle’s response to surface irregularities such as bumps, deformations or cracks. The classification system was then compared with traditional metrics such as the IRI and verified through visual inspections. The results showed a strong correlation between vehicle data and IRI measurements, confirming that connected vehicles can be a reliable source of information about the technical condition of roads.
[27]	The paper deals with the evaluation of the efficiency of road maintenance using Markov chains. The main performance indicator used by the authors is the IRI. The findings show that with exclusively routine interventions, the probability of remaining in the same condition is low, especially in worse condition categories. Periodic maintenance and repairs significantly increase the chance of maintaining the current condition, making them more effective.
[28]	The research focuses on using Bayesian network models to understand the relationships between factors that influence road deterioration. The results showed that the most important indicator of overall road condition is the IRI. Bayesian networks have proven to be a suitable tool for such analysis because they allow working with uncertainty, missing data and complex relationships between variables.
[29]	The study examines the assessment of pavement condition using the Pavement Condition Index. The research confirmed that the biggest factors in pavement damage are heavy freight traffic and the presence of groundwater. These conditions increase the rate of structural and functional failures. The research also confirms that preventive maintenance in the early stages of deterioration is more economically advantageous.
[30]	The research aims to analyze the relationship between the IRI and PCI and propose a new pavement condition classification system based on the IRI that could improve the accuracy of maintenance decision-making.
Topic area: Bridges—inspections, diagnostics and maintenance
[31]	This study is devoted to the assessment and classification of the condition of bridge structures in Malaysia based on a standardized assessment system. The BCI assessment system is based on visual inspections of bridge elements, which are classified according to the degree of damage and their impact on the overall functionality of the structure. Individual bridge elements are divided into major and minor components, and each category is assigned a weight according to its importance.
[32]	The aim of the study is to assess the vulnerability of existing bridges and propose possible alternatives for their reinforcement according to performance criteria that are analogous to the new bridge design. At the same time, the importance of regular monitoring and inspection of existing bridges, especially those designed according to older standards, is emphasized.
[33]	The paper deals with modernizing approaches to bridge maintenance through predictive models and optimization strategies. To this end, they apply a model that allows for more realistic modeling of bridge degradation, taking into account irregular inspections, maintenance time delays, and different levels of damage. The model combines discrete damage states and repair planning with continuous monitoring of the time until the next inspection or repair.
[34]	The article deals with the issue of the reliability of data obtained by visual inspection of bridge structures. The authors draw attention to the fact that, despite the widespread use of this method, there is a significant degree of uncertainty resulting from the subjective assessment of inspectors. A comparison of 988 defects showed that only approximately 21% of the assessments completely coincided in the type, severity and extent of the defect. These differences can affect not only the assessment of the condition of specific bridges, but also strategic decision-making in the management of the entire bridge network.
[35]	The authors propose a methodology for assessing the condition of bridges based on risk analysis. In the new model, the degradation states of individual elements are linked to the probability and consequences of their failure on the static safety of the bridge. Moreover, the analysis revealed that classic visual inspections often assess the condition of the structure too pessimistically, mainly because they do not distinguish between “appearance” damage and real static hazards. Such exaggerated assessments can lead to unnecessary and costly interventions that are not justified in terms of real risk.
[36]	The paper focuses on an integrated monitoring system for assessing the deformation behavior of a bridge. The procedure for creating an accurate observation method using a 3D laser scanner, identifying and quantifying the size of defects and deformation on the road surface, and determining the displacement of monitoring points using the transformation of selected parameters is applied.
[37]	This paper deals with improving the accuracy of ground-penetrating radar simulations for bridge deck diagnostics. Traditional GPR simulations often simplify the pavement model as a homogeneous material, which is not realistic, since asphalt concrete is a heterogeneous mixture. This inaccuracy leads to differences between simulations and real measurements. The authors propose a new methodology called Realistic Composite Medium Modeling (RCMM), which creates a detailed 3D model of the bridge deck based on the physical properties of its components.
[38]	The article deals with the proposal of a more efficient bridge condition assessment procedure in combination with the use of machine learning methods. The main goal is to improve the accuracy and consistency of bridge condition prediction, thereby strengthening the decision-making process in maintenance planning and extending the service life of these infrastructure elements.
[39]	The research describes a modern approach to managing and monitoring the condition of reinforced concrete bridges using digital twins. A digital twin is a virtual 3D model of a real object that is connected to sensory data from the actual bridge.
[40]	The aim of the study was to create an objective and adaptive bridge condition assessment model that overcomes the subjectivity of traditional methods. The model uses machine learning techniques to analyze historical inspection data, allowing for more accurate and consistent assessment of the technical condition of bridges.
[41]	The research deals with the assessment of bridge reliability using data obtained from structural monitoring. The monitoring results were compared with computational bridge models, and good agreement was achieved between the measured and simulated deformations.
[42]	The authors present research based on the combination of GPR, laser scanning and FEM calculations, which represents an effective tool for comprehensive assessment of the condition of the pavement in the bridge transition zone. Such an integrated approach allows not only diagnosing current damage, but also understanding its causes and preventing future failures.
[43]	The paper presents the effective use of non-destructive methods, namely ground-penetrating radar and laser scanning, in the diagnosis of pavement damage in bridge transition areas, especially in places where transverse cracks occur at bridge abutments. Invasive tests were also performed to verify the results of non-destructive methods. The conclusion of the article points out that the combination of ground-penetrating radar measurements and laser scanning represents an effective tool for the early diagnosis of pavement defects in the bridge area.
[44]	The paper focuses on the design and validation of an evaluation system for assessing the technical condition of bridge structures. The proposed evaluation system is based on a combination of visual observations and quantifiable technical parameters. Specific evaluation indicators are determined for each of the monitored bridge parts, which reflect the extent, type and severity of failures. Subsequently, weight coefficients are assigned to these indicators according to their importance for the overall safety and functionality of the bridge.
[45]	In this study, the quality of asphalt pavement was analyzed using the pavement quality index. The aim was to compare the absolute values of the pavement quality index and the differences between them depending on the type of section. The conclusion of the study recommends that these segments, especially large bridges and long tunnels, be considered separately when planning routine maintenance work, determining the amount of necessary interventions, and calculating costs.
[46]	The paper presents an advanced model for optimizing bridge maintenance, which includes not only maintenance costs but also user costs. The authors develop the previous Markov model of bridge degradation and take into account imperfect repairs, which have different success rates and impacts on the condition of the bridge. The model allows for multi-level inspections with different intervals depending on the current condition of the structure. The result is not a single solution, but a set of Pareto-optimal strategies, among which the bridge manager can choose according to preferences, for example, to prioritize lower costs for users at the cost of higher investment in maintenance or vice versa.
[47]	The proposed model allows for inspections to be performed at different time intervals depending on the current state of the system. This approach is preferable to fixed periodic inspections because it better reflects the actual maintenance needs and can significantly reduce redundant inspection interventions. The results show that the best results in terms of cost and efficiency are achieved by the model that combines PH (phase-type distributions—PH) and condition-dependent inspections.
[48]	The paper describes how modern digital technologies, especially drones, are changing the way bridges and roads are inspected and maintained. The result is the design and development of a comprehensive software solution for intelligent bridge inspection.
[49]	The paper is devoted to the development of a procedure for identifying cracks in reinforced concrete bridges using semantic segmentation using deep neural networks. The aim is to make bridge maintenance more efficient in terms of time and money.
Topic area: Assessment and classification of road damage
[50]	The aim of this study was to propose calibration factors for predicting pavement degradation in Moroccan conditions, specifically for four basic types of damage: structural cracks, surface breakdown, potholes, and resultant unevenness.
[21]	The article presents a comprehensive approach to automatic crack detection on asphalt pavements, emphasizing segmentation accuracy at the pixel level and detailed damage quantification.
[51]	The aim of the research was to solve the problem of traditional methods for segmenting pavement crack images, which often fail to accurately capture the edge structures of cracks. The authors proposed a flexible system for crack detection and identification that increases segmentation accuracy even in challenging conditions.
[52]	The research presents an innovative approach to assessing the condition of road markings using LiDAR technology. It demonstrates the practical use of existing sensor technologies, which are already found in many autonomous vehicles or infrastructure data collection vehicles today. The results can be integrated into road network management systems, which will allow for early identification of the need for marking renewal, reducing costs and increasing safety.
[53]	This paper deals with the automatic detection and classification of cracks on pavements using image processing methods, specifically based on the Hough transform. The system achieved high accuracy in both crack detection and classification compared to manual evaluation.
[54]	The publication addresses the problem of the automatic classification of cracks on roads, proposing an innovative approach that combines intelligent thresholding optimization with hybrid image data extraction. Experimental results indicate an efficiency of the proposed hybrid model of up to 98.10%.
[55]	The paper describes a methodology combining advanced computational modeling, stochastic modeling based on field data, and easily measurable meteorological variables into one integrated forecasting tool with the aim of better predicting the moment of crack formation on the road surface.
[56]	The article deals with the automated detection and sorting of cracks on highway surfaces using aerial images from unmanned aircraft and artificial intelligence technologies.
Topic area: AI/computer vision—crack and failure detection
[57]	The article deals with the design and verification of a new method for visualizing and detecting damage to airport runways. As a solution, the authors propose a method based on a combination of BIM technology and CCD imaging (Charge-Coupled Device imaging). This combination allows for detailed imaging of airport runway surface defects and at the same time creates space for intelligent processing of these data in an integrated digital model.
[58]	The research focuses on the development of fully automated software for recognizing road damage through image data analysis. The author’s goal was to streamline and automate the process of assessing road condition, which is currently in many cases dependent on manual or semi-manual methods. The research focuses on processing video footage from common dashcams installed on vehicles.
[59]	Manual inspections are labor-intensive, costly, and error-prone. Existing automated systems also struggle with false positives. Research is focused on architectures for detecting and scoring dents using photographic evidence and AI to improve processes and reduce costs.
Topic area: Traffic signs and their quality
[60]	The article focuses on the development of an intelligent system for detecting damage to road markings using deep learning methods. The proposed system uses deep neural networks to process road images in order to detect damage to markings, such as fading, cracks, interruptions or complete disappearance of markings.
[61]	The article deals with the assessment of the quality of road marking implementation and its night visibility using computer vision and machine learning technologies. The high potential for automated, accurate and efficient assessment of road marking visibility is confirmed.
Topic area: Data, sensors, monitoring, digital twin
[62]	The research presents terrestrial laser scanning as a promising technology for a modern road management system. In combination with appropriate software tools and automation algorithms, it can significantly contribute to the digitalization and streamlining of road condition control processes.
[63]	The research focuses on evaluating the use of satellite imagery as an alternative and cost-effective tool for monitoring the technical condition of roads. They showed that the use of satellite data can lead to cost reductions of up to 6.5% over the life cycle of a road, with the greatest savings being seen on roads that have not been regularly monitored.
[64]	The article aims to present a practical methodology using a ground-based laser scanner for defect detection and condition assessment of concrete pavements as an alternative to traditional visual inspections.
[65]	The research deals with the use of mobile laser scanning as an effective technology for assessing road surface damage. As part of the study methodology, a monitoring unit was installed on a vehicle that moved along selected road sections. The obtained 3D data were subsequently processed using algorithms to calculate micro- and macrotexture parameters, longitudinal and transverse profiles and surface variations. The study also highlights the challenges associated with the use of MLS, such as the need for standardized data processing procedures, solving shading problems when collecting data in urban environments, or integration with other types of sensors.
[66]	The article deals with the development and implementation of a custom sensor system for measuring and predicting temperature trends in road construction layers. The aim of this work is to make winter road maintenance more efficient, especially in remote areas where reliable information on current road conditions is lacking. The lack of data often leads to unnecessary trips of gritting vehicles, which causes environmental and economic losses.
[67]	The article deals with modern approaches to monitoring the condition of road and airport surfaces. The latest advances in 3D scanning and integration of data from GNSS systems allow these routine inspections to bring much higher added value in the form of creating so-called “3D digital twins”. The outputs from the software can then be transferred to laser measurement systems that control 3D pavers and milling machines.
[68]	The aim of the study is to develop an effective, cost-effective and rapid system for assessing the condition of roadways using satellite remote sensing.
[69]	The paper focuses on improving the quality of roadway survey data. The authors present a comprehensive data cleaning framework that combines data mining and deep learning techniques. In conclusion, it can be stated that the proposed framework provides an innovative and efficient solution for processing large and heterogeneous datasets from road inspections.
[70]	The research deals with advanced assessment of the structural condition of pavements and estimation of their remaining service life through a combination of two modern methods—accelerated load testing and ground-penetrating radar survey. The authors conclude that the combined approach significantly improves the ability of infrastructure managers to accurately determine the current condition of the pavement and predict its development.
[71]	Conventional inspection techniques for detecting underground cavities, such as GPR, are effective but expensive and time-consuming. Alternatively, a laser dynamic deflectometer can be applied, but its reliability is unclear and needs to be investigated.
[72]	The authors propose a novel autonomous positioning system that combines two laser devices and an image sensor. This hybrid approach should enable accurate positioning even in environments without a GNSS signal, paving the way for the full deployment of digital technologies in infrastructure monitoring and management even in challenging conditions.
[73]	The research focuses on monitoring the impact of an interactive digital twin on various forms of interaction between users and monitored objects. The goal is to provide insights into how these interactions can improve infrastructure inspection and maintenance processes.
[74]	The study presents that the combination of non-invasive testing and geostatistical analysis is an effective tool for assessing the condition of asphalt pavements. This approach enables efficient maintenance planning, reduces costs, and minimizes the need to damage the pavement during testing.
[75]	The research focuses on the automated processing of data obtained using ground-penetrating radar in order to identify the interfaces of asphalt layers and determine their thickness. A significant innovation is the introduction of a fully automated algorithm for identifying these interfaces, which minimizes the need for expert intervention. The system avoids traditional manual marking and uses boundary detection and segmentation procedures of radar signals in the form of B-scans (images created by a series of radar measurements). The process also applies automatic computational determination of the dielectric constant of asphalt.
[76]	The article presents a modern and effective method of diagnosing airport pavements using a high-resolution radar system. It allows airport operators to obtain a comprehensive overview of the condition of the structure, identify potential weak spots before they appear on the surface, and thus reduce the risk of accidents and extend the life of expensive infrastructure.
[77]	The proposed inspection model, using advanced mathematical and machine learning methods, quantifies how much should be invested in various sources of information and whether these investments will yield the expected benefits. The proposed methodology provides a framework for road maintenance decision-making under conditions of uncertainty and limited resources.
Topic area: System approaches/asset management/other
[78]	The paper deals with the possibilities of using digital image processing in assessing road damage as an effective alternative to traditional methods of visual and mechanical inspection. The starting point of the research is the finding that conventional visual inspection is practical and economically available, but at the same time very subjective, prone to errors caused by human factors, such as fatigue or different interpretations between inspectors. More accurate mechanical methods are expensive and often measure only one specific property of the road. Therefore, the authors propose an algorithmic approach based on the processing of digital images of damaged roads.
[79]	Road condition prediction models play an important role in predicting road maintenance costs. The study proposes an approach that helps managers predict annual maintenance costs in the context of road quality.
[80]	The research shows that using the modified YOLOv7 model in combination with appropriate image processing and high-quality data, a powerful system for automated crack detection and classification can be created that is more accurate, faster, and safer than traditional approaches.
[81]	The paper presents a new approach to assessing road damage using a signal processing technique on graphs, where individual nodes represent specific road sections and edges express their spatial or functional connections. The method offers not only higher accuracy and robustness in damage estimation, but also flexibility in incorporating different types of data.
[82]	The study proposes a strategy for using the IRI as a basic indicator in planning and monitoring road construction and maintenance. The authors recommend measuring the IRI regularly and using these data as a basis for decisions on prioritizing repairs and investments.
[83]	The paper presents a new system for the quality control of concrete pavements based on data mining principles. The aim is to build a multi-purpose and efficient detection structure that would significantly improve the reliability and efficiency of quality assessment through automated data processing.
[84]	This study aims to improve the quality control methodology for concrete pavement construction. The author proposes a new quality control method based on data mining techniques, which enables more accurate, faster and multidimensional evaluation of the quality of concrete surfaces.
[85]	The article presents a methodology that links long-term monitoring data with condition assessments of infrastructure objects through a reliability framework. The aim is to quantify reliability indices corresponding to the lower limit states described in existing condition assessments. This approach can improve decision-making in the field of maintenance and renewal of infrastructure objects.
[86]	The article presents an innovative method for measuring rut depth on roads using a conventional camera installed in a vehicle. The authors use an advanced neural model without the need for keypoint detection, which is able to reliably identify corresponding points even on homogeneous surfaces. Based on the differences in the position of these points between two images, it is possible to reconstruct the shape of the road cross-section and derive the rut depth in pixels.
[87]	The study highlights the fact that the actual service life of infrastructure can be significantly different from the theoretical one. The main sources of observations are data on pavement deflections, historical records of pavement layer composition and visual inspections. Factors that could influence this state include, for example, climatic conditions, type of surface treatment and age of the pavement.
[88]	The paper focuses on the development of an artificial neural network-based model for predicting the Pavement Condition Index. The model is able to predict the PCI values based on easily accessible technical and geometric parameters such as damage type, damage intensity, pavement dimensions, and operational factors.
[89]	The research is focused on developing a complex system that uses mobile devices and computer vision algorithms to automatically detect cracks in structures, determine their exact location, and quantify their size.
[90]	The research deals with the design and development of an improved algorithm for detecting cracks in pavements, which aims to increase the accuracy and efficiency of defect identification. The authors proposed a modification of the well-known Yolov5s detection model.
[91]	The research deals with the design and development of an improved algorithm for detecting cracks in pavements, which aims to increase the accuracy and efficiency of defect identification. The authors proposed a modification of the well-known Yolov5s detection model.
[22]	The authors present that traditional inspection methods, such as visual inspections or manual measurements, are often time-consuming, subjective, and not always reliable. In response, they apply several machine learning algorithms to classify and predict damage based on a database of damage and corrosion defect images.
[92]	The study deals with the use of deep learning methods for the automated detection of asphalt pavement damage. The results showed that the model is able to identify different types of damage with high accuracy, even under complex lighting conditions or different asphalt textures. The system was able to work in real time, which allows its integration into mobile platforms, such as vehicles equipped with cameras for continuous monitoring of road infrastructure.
[93]	The research aims to improve automatic crack detection in pavements using computer vision methods. The results confirm that multi-resolution adaptive thresholding significantly increases the efficiency of automatic crack detection.
[94]	The text is devoted to setting priorities for infrastructure repairs under limited finances. Modern technologies in the field of sensors, communication systems and fault detection algorithms create new opportunities for data-driven decision-making. These approaches allow traditional visual inspection methods to be at least partially, if not completely, replaced by continuous monitoring. However, despite these possibilities, there are several fundamental technical, organizational and practical problems that prevent their widespread deployment in real infrastructure networks.
[95]	The aim of the research was to design a comprehensive and efficient system that, using image data processing, can accurately identify cracks on road surfaces, reconstruct their spatial shape, and then classify them by type and risk. In experiments, the system achieved more than 95% classification success and approximately 84% crack detection accuracy, with processing of one image taking approximately two seconds.
[96]	The research compared the road condition index calculated from drone data with values obtained from traditional ground inspections. The differences between the two methods ranged from 1.3 to 9.1 points, which is a relatively small difference. Furthermore, the Pearson correlation coefficient R (0.79 to 0.90) showed a strong correlation between the two approaches, confirming the high potential of integrating drone inspections as a complement to traditional inspections.
[97]	The study addresses the optimization of airport pavement inspection planning based on the concept of value of information (VoI), a theory of decision-making under uncertainty. This approach allows quantifying the benefit of inspection data as the expected risk reduction or benefit increase resulting from performing measurements before making decisions on infrastructure maintenance or renewal. The authors show that their approach allows for efficient, targeted inspection planning with limited resources.
[98]	The paper presents an innovative approach to the detection and classification of asphalt pavement rutting using 3D technology and machine learning. The authors purposefully developed an automated system based on 3D data collected from the pavement surface and the use of machine learning algorithms to classify rutting severity.
[99]	The research focuses on the management and maintenance of unpaved roads with low traffic intensity. The results show that the implementation of GIS as a decision support tool in the field of road management is effective, especially in resource-limited environments. An integrated approach combining field data collection, visualization, analytical evaluation and intervention design significantly increases the efficiency of decision-making processes.
[100]	The study presents an alternative and low-cost approach that uses accelerometers embedded in smartphones to estimate road condition by measuring vibrations while driving. The research involved the development of an Android application that records vertical acceleration and processes it to estimate road surface quality. The results of this study confirm that vibration data from smartphones represent an effective and affordable way to continuously monitor road condition.
[101]	The article focuses on the evaluation of the technical condition of concrete pavements at airports using multi-level criteria analysis. The proposed method consists of quantifying the degree of degradation based on 13 types of damage and repairs that are detected during inventory inspections. The evaluation is carried out by calculating the so-called degradation indicators, which take into account the extent of damage, repairs performed, as well as their impact on the safety of air traffic. The impact of individual parameters is weighted by coefficients determined by experts in the field of airport structure research.
[102]	The paper addresses the problem of incomplete road condition data. The authors propose an innovative approach using a Graph Convolutional Network (GCN) based on collective learning. This method combines information about the characteristics of adjacent road sections and the dependencies between their observed conditions to effectively fill in the missing values.

provides the content orientation overview of the final set of studies included in the systematic literature review. Individual publications are divided into thematic areas according to their main research focus.

Summary of the Systematic Analysis Results

Based on a detailed analysis of the selected contributions, it can be stated that in the field of transport infrastructure management, the development of methods that enable accurate, objective and effective assessment of the technical condition of roads and bridges continues to intensify. Classic visual inspections, which have long dominated this area, are proving to be insufficient, mainly due to their subjectivity, being time-consuming and the risk of errors resulting from the presence of the human factor. The scientific environment, also due to demand from practice, is oriented towards the use of advanced sensor technologies, artificial intelligence methods, machine learning and digitization tools, with the aim of creating complex and automated systems for diagnosing, monitoring and managing the maintenance of infrastructure objects. Several studies have confirmed that the unevenness and degradation of pavements quantified by indices such as the IRI (International Roughness Index) or PCI (Pavement Condition Index) are closely related to specific types of damage, with network cracks, transverse failures and plastic deformations having a dominant influence. At the same time, it has been shown that classical visual methods often fail to identify or quantify these defects with sufficient accuracy. In response to these shortcomings, systems based on the analysis of image and sensory data are being created and improved, using deep learning algorithms or decision trees. These systems achieve high accuracy in the detection and classification of cracks, dents and other defects, while enabling automated segmentation and quantification of damage even under difficult lighting and textural conditions. Advanced sensing methods, such as terrestrial, mobile and airborne laser scanning, georadar, LiDAR or high-resolution radar systems, provide the opportunity to create detailed 3D models of road and bridge surfaces, which serve not only for diagnostics, but also for predicting the further development of degradation. These models often become the basis for the creation of digital twins, which connect physical objects with their virtual representation in real time. In this area, it is increasingly common to connect various data sources from camera recordings, radar images, sensor data from GNSS, drones or smartphones, creating integrated and adaptive monitoring platforms with a high degree of flexibility and usability.

Predictive modeling methods that combine technical indicators with statistical and stochastic tools play a crucial role. Markov chains, Bayesian networks, and value of information decision models are used to create and optimize schedules for inspection and maintenance performance with respect to technical condition, failure risks, and economic costs. Special emphasis is placed on multi-criteria optimization, which takes into account not only administrator costs, but also user costs, environmental impacts, and the level of security. Research is also devoted to data quality aspects, where methodologies for the management and standardization of large datasets from various types of sensors are proposed using data mining and deep learning tools. The goal is to create robust and reliable systems capable of operating in real time and providing data for decision-making. The research showed that there is a growing presence of research activities in the areas of automated data collection using conventional cameras and intelligent systems for detecting damage to road markings. Most of the findings ultimately aim to contribute to the transformation of the way in which the life of transport infrastructure is planned, monitored and managed, to more rational maintenance planning, to reduce unnecessary interventions and to more efficient use of financial resources. The availability of reliable information, automation, connectivity and adaptability of systems are considered key.

Key publications in the systematic review include [28,33,46,47,77,97]. Their evaluation points to the growing importance of applying the concept of value of information (VoI) in the field of repair and maintenance management of transport infrastructure. This topic is becoming extremely relevant due to limited resources, the presence of uncertainty in diagnostics and the need for effective decision-making. The work of [97] presents an advanced framework for optimizing airport runway inspections based on the value of information. This approach quantifies the extent to which new information obtained from diagnostics or partial inspections reduces uncertainty in decision-making and therefore whether the costs of these measurements are justified given the potential benefits in terms of safety and maintenance costs. The study of [77] moves this concept to the road management environment, where it combines the value of information with advanced Markov decision processes. They thus enable the integration of insufficient information from inspections and installed sensor systems into decision-making models for maintenance and repair. The approach presents the fact that even inaccurate or incomplete information can have significant decision-making value if it is correctly quantified in the context of risk and costs incurred. This issue is also followed by the works of [46,47] presenting knowledge in the field of bridge maintenance optimization through advanced stochastic models. The research focuses on realistic scenarios that take into account not only the costs of the infrastructure owner, but also user costs. The authors work with real dynamics (especially time and quality from the point of view of the work performed) of inspection and maintenance processes. Their model from 2022 is based on a Markov model enabling multi-level planning of inspections with different intervals depending on the current state of the bridge. It uses an optimization framework in which the so-called Pareto-optimal strategies minimize the trade-off between inspection and repair costs and user costs. The result is not a single solution, but a set of alternatives from which the administrator can choose according to preferences and priorities. The 2024 model further develops this concept using the so-called phase-type distributions, which describe in more detail the delays between inspection, decision and maintenance execution. Both studies present that the value of information obtained from inspections is not only a matter of theoretical interest. If this information is properly processed and its benefit is quantified in the context of maintenance decisions, it can fundamentally influence the choice of the optimal strategy and achieve significant savings while maintaining or improving safety. In terms of predicting road surface degradation, the study of [28] presents an application of the Bayesian approach to the analysis of relationships between degradation characteristics, where the international unevenness index, IRI, is presented as a key indicator. The model they propose allows for the integration of uncertainty and missing data into monitoring planning decisions, further reinforcing the need to express the value of new information. The research of [33] uses hybrid predictive models for bridge maintenance planning, emphasizing the importance of inspection information for dynamic decision-making. The models proposed by the authors support principles based on the assertion that the value of inspection information is strongly linked to the needs of reducing the risk of failure, more effectively scheduling repairs, and minimizing costs.

Topics related to the value of information represent a relevant, albeit gradually established, concept in the field of transport infrastructure management, not excluding its specific areas related to the optimization of the performance of technical condition inspections. Although it is not a dominant research trend, its importance is gradually growing, which can be observed in several studies identified by a systematic review of sources. The goal of implementing such approaches is not only to reduce managerial uncertainties associated with the selection of objects intended for the assessment of the technical condition of roads and bridges, but also to formally quantify the potential or real scope of the benefit of performing an inspection, taking into account its timing, chosen method or scope. The application of the approach based on the value of information principle contributes to better formalization of the management of technical condition inspections and at the same time improves the economic justification of their performance in decision-making models monitoring the economic side of infrastructure management. Information for managerial decisions should be complete and free from errors. This leads to better risk management, lower costs and ultimately an increase in the quality of monitored objects on the road network. The application of VoI approaches is of considerable importance in planning inspection strategies and its further research seems justified.

4.2. Results of the Non-Systematic Analysis

Effective management of road infrastructure is crucial for increasing the safety and sustainability of the road network. In this context, road management systems are important tools that provide a systematic framework for predicting road condition, planning and optimizing maintenance and repairs [103]. At the same time, the process of assessing the technical condition of roads and bridges is a fundamental pillar of effective road infrastructure management, and its results are essential inputs for the application of the proposed methodologies, calculations and simulations contained in these management tools [104,105]. In addition, the authors’ study [106] states that monitoring, technical condition diagnostics or wider inspections are included among the key factors influencing the comprehensive resilience of road infrastructure. These activities have the ability to capture facts and trends of the deteriorating quality of managed assets and thus increase the robustness of the infrastructure against signs of deterioration.

The authors of the study [107] concluded that the main challenge for effective asset management is to reconcile three areas: clear definition of objectives, knowledge of the current state of assets and effective decision-making on interventions. A fundamental problem is that objectives are formulated on the basis of low expertise or multiple external influences (for example, political, expert due to regional needs or based on often conflicting user preferences). For this reason, it is necessary to balance the weaknesses associated with objective management with the remaining two areas. This means that it is necessary to have a high-quality database on the state of managed assets, which will allow justifying the outputs of decision-making on interventions. The study of [108] shares this view and emphasizes that public agencies should strengthen the systematicity of inspections on the road network and their technological level, introduce clear evaluation criteria and increase the transparency and usability of inspection data in decision-making. The study of [109] also points out that inspections and interventions must be a permanent and significant part of an integrated decision-making framework for infrastructure management. Asset condition assessment is a critical component of overall infrastructure management according to [110]. The authors’ efforts in this study led to the development of a systematic, risk-based asset management methodology for highways, which will enable effective decision-making on the maintenance, prioritization and allocation of financial resources. According to the authors, without accurate information on the condition of the infrastructure, it is not possible to effectively translate this effort into real results. The authors of the study consider an objective classification of assets according to their importance, technical condition and risk exposure as a critical prerequisite. The article of [111] presents that modern diagnostics of the technical condition of roads is an integral part of advanced road infrastructure management systems. The most important issues associated with the assessment of the technical condition include those related to the optimal time of its execution, the method and scope of its implementation, and the assessment of the value of the information obtained. It is further stated that the lack of inspections or their inappropriate planning and implementation can cause a slowdown in the restoration of managed assets and an increase in economic and social damage. One of the reasons for the optimal timing of inspection and diagnostics is to minimize the total costs over the life of the infrastructure. Inaccurate or delayed information from inspections can lead to inappropriate decisions, increased costs without adequate benefit, and to a deterioration in the safety and functionality of the infrastructure [109,112,113]. As stated by the authors of [114,115,116], too frequent assessments of the technical condition increase costs without an appropriate increase in the accuracy of decision-making, and conversely, frequent or incorrectly timed inspections can lead to failures or unnecessarily expensive repairs and maintenance.

The main focus of the study of [117] is the analysis and determination of optimal inspection intervals for infrastructure objects depending on the quality of the diagnostic methods used. The starting point is the finding that a large number of existing structures require regular assessment of the technical condition, which is financially demanding, but at the same time necessary for maintaining the safety and reliability of the assessed structures. The study uses a two-stage inspection framework, which consists of defect detection and determination of the size of the detected defects. The results show that shorter intervals increase the costs of the inspections themselves, but reduce the costs associated with failure. Longer intervals are cheaper in terms of direct inspection costs, but significantly increase the risk of failure and therefore the total costs. The authors found that the higher the quality of the diagnostics, the more the intervals can be safely extended. The study of [114] focuses on the optimization of inspection and maintenance decisions in the case of infrastructure objects using a quasi-Bayesian approach. The main objective is to minimize the total expected social costs associated with the operation of the infrastructure under conditions of uncertainty associated with its technical degradation. In practice, the processes of measuring the technical condition of infrastructure objects are not perfect. They are affected by the uncertainty resulting from the accuracy of measurement technologies. This uncertainty can lead to inappropriate decisions when choosing the type and extent of intervention, because the measured condition may not correspond to the real condition. The authors of the study propose a framework methodology that simultaneously optimizes decisions on the timing of inspections (whether and when to perform condition measurements) and the selection of the type and extent of intervention. This framework integrates updates of the object degradation model based on new data and latent (hidden) Markov decision processes. These do not assume perfect measurements and allow working with a probabilistic link between the measured and real condition of the infrastructure. The study showed that a flexible inspection plan (instead of a fixed interval) significantly reduces the total life cycle costs of the object.

The aim of [115] was to minimize the total life cycle costs of the infrastructure (inspection, maintenance and repair costs, user costs) using an adaptive optimization method that flexibly adapts the inspection schedule based on current knowledge and uncertainties. The authors combine a latent Markov decision process, where the condition of the assets is estimated using probability distributions of the condition based on historical data and current measurements, and adaptive control. This takes into account uncertainty in the prediction models and uses new information from the obtained measurements to continuously update the confidence in different possible models of infrastructure degradation. Using the example of road management, it was proven that adaptive inspection intervals reduce the total expected costs of the infrastructure and the accuracy of the initial assumptions about degradation plays a significant role. The authors showed that it is economically more advantageous to provide accurate and reliable information right from the start, as this significantly improves the long-term quality of decision-making.

Quantified results of the study of [116] confirmed that less frequent but more accurate inspections can be more effective than frequent but less accurate measurements. The authors developed a methodology for optimizing the planning of inspection and maintenance interventions on the infrastructure network in order to take into account measurement errors (uncertainties) during inspections and adapt the decision-making strategy accordingly. At the same time, in the next step, they proposed adaptive strategies for reducing the total life costs of the infrastructure network. The authors applied an extended Markov decision model, as have several previously mentioned authors.

A new approach to determining the optimal inspection time is presented in [118], which introduces a quantitative decision framework based on a semi-Markov process. The technical condition of the monitored infrastructure is divided into discrete states. If data are missing, the distributions are estimated based on expert estimation and these are updated with real data over time. An approach is considered where the time spent in one technical state before transitioning to a worse one is taken as a random variable with a known probability distribution. The optimal time for the next inspection or diagnosis is the one when the total expected discounted costs are minimal, while the decision process is repeatedly updated after each inspection based on new findings about the actual state. As can be seen, several studies emphasize that the optimal time for performing a technical condition inspection is a more significant task than has been considered so far. Already in 1991, ref. [113] addressed the issue of optimizing decisions on the maintenance and renewal of transport infrastructure in the presence of uncertainty associated with the measurement and prediction of the technical condition. He applied the latent Markov decision model as a method explicitly taking into account uncertainty in the measurement of the technical condition. This model differs from traditional Markov decision models in that it does not assume the error-free measurement of the technical condition of the infrastructure. Measurements are perceived as random variables with a certain probability of errors. This allows for better planning of maintenance and renewal with regard to the accuracy and timing of measurements.

The authors of the article [112] developed a two-dimensional Markov model based on historical data from the National Bridge Inventory database, which includes information on the condition of more than 17,500 bridges for the period 1992–2018. This model includes the current condition of the bridge and the time during which the bridge remained in its current condition. The authors recommend that bridge managers implement a flexible inspection interval system based on an accurate assessment of the current condition of bridges and the risk of their deterioration. Such an approach can bring significant savings, more efficient use of resources, and at the same time increase the overall safety and reliability of the bridge infrastructure. The authors also emphasize the need for further research that would include other factors (e.g., climatic conditions, bridge age, material characteristics), which could further refine the determination of optimal inspection intervals.

Asset condition data collection should be accurate, fast, safe and cost-effective. The methods and approaches used should optimize these criteria, but increasing accuracy should not significantly reduce speed or increase costs [108]. Accurate, fast and objective assessment of the technical condition is a key step for planning effective maintenance, reducing costs and increasing transport safety, also according to [104]. In terms of the method and scope of inspection activities, the application of technical condition assessment tools can be implemented at the network level, monitoring the impacts on long-term costs and the condition of the asset portfolio (which is rare) or at the project level intended for predicting degradation and optimizing interventions [119].

The methods of inspection and assessment of road condition have undergone a fundamental transformation over the last decade, from manual visual inspections to complex, data-oriented, intelligent systems supported by artificial intelligence tools and smart technologies. Modern approaches to infrastructure management increasingly use combined, automated and non-invasive methods of data collection and processing. The integration of automated visual technologies, geographic systems and analytical modules allows not only reducing costs, but also increasing the reliability and efficiency of decision-making processes. The result is transparent, data-based maintenance management with significant benefits for the safety and economy of road network operation. The main challenge is the standardization, high reliability and availability of tools for assessing the technical condition of roads and bridges in real operation [111]. As stated in [104], it will be crucial to address the issues of interoperability, standardization and availability of technologies for all management organizations, aiming to achieve a higher level of automation and intelligence in road network maintenance management. Challenges in technical condition diagnostics are the lack of standardization in evaluation methods and in the integration of data from different sources, budgetary and technological limits of management organizations and the need for predictive and proactive models based on real data. For example, the contribution of [120] states that non-destructive approaches are of increasing importance in the accurate diagnosis of the technical condition of infrastructure and increasing the efficiency of asset management. Challenges in the implementation of inspections during the operation of infrastructure elements are the lack of training and qualified personnel to operate and interpret data from advanced inspection technologies, the high costs of specialized equipment and maintenance, resistance to change and reluctance to implement new technologies and various technical limits (availability of GPS or internet connection in the field, technological compatibility, etc.).

In the management of road infrastructure assets, information can take on several dimensions. It can generally be viewed as an investment and an intangible asset that behaves differently from traditional tangible assets [121]. The results presented in [122] point to the growing importance of information in organizations and the need for a more systematic approach to its evaluation and understanding, management and use. At the same time, it points to the fact that information is often managed in an ad hoc manner, without a clear awareness of its value or characteristics as an economic good. Studies [123,124] emphasize that the information obtained has its value, contributing to decisions and competitive advantage, and has its manager responsible for its quality, data security and the method and extent of use. Studies indicate that information is reproducible and inexhaustible, but its relevance and reliability change over time. The work of [122] mentions that organizations invest significant resources in information systems, but rarely evaluate the information itself as a strategic asset.

The authors of the study [109] present that the highest value of the acquired knowledge lies in the reduction in uncertainty and better decision-making. Quality decision-making in the field of asset management requires not only up-to-date and accurate information, but also intelligent tools for its analysis. According to [124,125,126], the analysis of the value of information is a decision-making analytical method that quantifies the benefit of additional information in decision-making under conditions of uncertainty, following efforts to reduce its impact. This approach has theoretical roots in decision analysis, where it expresses the difference in expected value between a decision with available additional information and a decision without it. It can be calculated either as the value of perfect information or as the value of incomplete information, while decision trees, cost–benefit methods or simulation methods are often used. The authors of the study [123] demonstrate that the combination of expert inspection systems with fuzzy logic allows for the creation of an effective decision-making framework for predictive maintenance, thereby increasing organizational performance, reducing costs, and minimizing unplanned downtime.

The assessment of the contribution of information and its value is also important in the construction and infrastructure industries, especially in the context of decision-making under the influence of uncertainty. As further stated by [126], new elements of structured monitoring, big data tools and approaches based on artificial intelligence are increasingly coming to the fore in the management of objects and infrastructure. The study points to two main theoretical levels in their processing. The first of them is Shannon’s information theory, concentrating on the reduction in uncertainty. The second is Bayesian decision theory, oriented towards economic consequences. The application of the concept of the value of information allows us to answer questions such as whether it is meaningful to implement the monitoring of infrastructure elements, whether it makes sense to conduct additional technical condition checks before intervention, or to determine what type of information has the greatest impact on the decision.

The aim of the study [127] is to quantify the value of information obtained from structural monitoring in flood infrastructure management. The authors point out that many infrastructure decisions in the field of flood protection are burdened by a high degree of uncertainty regarding the current condition of dams, their performance, failure risks and expected consequences. The aim of the study is to answer the question using an analytical model: when is it worth investing in technical condition monitoring and to what extent. The value of information is defined here as the difference between the expected net benefit of a decision with information from monitoring and a decision without this information. The authors use a Bayesian decision-making framework that combines prior knowledge about the condition of dams (e.g., based on expertise), new monitoring data and probabilistic updates via Bayesian principles. The result is the expected value of different decisions, such as strengthening the dam structure, carrying out repairs or maintaining the current state. The results showed that monitoring can significantly reduce uncertainty and thus optimize decisions, especially in cases where the initial uncertainty is high and the costs of incorrect decisions (e.g., premature rehabilitation) are high. The main finding is that the value of information increases with increasing uncertainty and the consequences of failure. The study clearly supports the systematic integration of approaches based on the value of information into the asset management process of flood protection systems. The authors recommend that infrastructure managers use this method not only as an economic tool, but also as a means of transparent and evidence-based decision-making that reflects the technological possibilities of technical condition monitoring.

Several authors, including [126], state that the determination of the value of information is an important, but not yet widely used tool for making better decisions in an environment of uncertainty, providing opportunities for increasing transparency and efficiency in both technical and social terms. Several studies, on the other hand, point to complications in the form of difficulties in recognizing the type and quality of information, the high computational complexity of its processing or the different accuracy of probabilistic models and their development assumptions. The future trend and the resulting tasks should be oriented towards simplifying calculations, especially by wider use of machine learning elements or analytical approaches, and a wider involvement of VoI approaches in infrastructure management processes should be pursued.

Summary of the Non-Systematic Analysis Results

The chapter presents a summary of the knowledge gained through a non-systematic literature review. Effective management of road infrastructure assets, especially in connection with increasing the level of its sustainability and safety, is inextricably linked to a high-quality assessment of the technical condition of roads and bridges. Current, accurate and reliable information is the basis for modern road management systems, following high-quality condition prediction, planning and optimization of maintenance and repairs. A critical issue is the optimization of the scope, timing and quality of technical inspections. Research shows that an organizationally and technically inappropriate approach to these processes increases costs without adequate benefits, while inspections that are insufficient in scope or incorrectly timed lead to the risk of failure, delayed restoration and increasing economic damage. Modern methodologies, often based on latent Markov decision processes and Bayesian approaches, allow working with uncertainty and measurement errors that can significantly affect the correctness of intervention decisions. Studies show that technological progress is significantly changing the way diagnostics are performed. There is a shift from manual visual inspections to automated, non-invasive and data-driven methods supported by artificial intelligence. However, this transformation is accompanied by fundamental challenges, such as high technology costs, lack of standardization, interoperability issues and a shortage of qualified personnel. The value of the information obtained through inspection is also a central theme. Information is increasingly understood as an asset that brings value, mainly by reducing uncertainty in maintenance and investment decisions. Analysis of the value of information, most often based on Bayesian theory, allows us to determine when an investment in monitoring or additional measurements of the technical condition will actually pay off. Its importance grows with the level of initial uncertainty and the severity of the consequences of a potential failure. Studies clearly show that the systematic application of this approach can fundamentally improve the transparency, efficiency and economic optimization of the management of (not only) road infrastructure.

5. Discussion

A review of the available literature and an analysis of existing knowledge clearly indicate that the current research in the field of road diagnostics is largely focused on the technical essence of measurements, i.e., on the technological background and measurement processes. The attention of the scientific community, as well as application practice, is mainly focused on the description of new and improved measurement methods, the topics of more intensive implementation of modern algorithms (often based on the principles of artificial intelligence and machine learning) and demonstrating the benefits of diagnostics for the higher efficiency of management environments. This trend brings valuable knowledge about the accuracy and reliability of individual diagnostic approaches and tools, their technical capabilities or the possibilities of integration into infrastructure management systems. The dominant space is filled by the question of “how to perform diagnostics” and “how to ensure the highest possible information gain for maintenance and repair needs”. Figure 4, Figure 5, Figure 6, Figure 7, Figure 8 and Figure 9 show a significant shift from traditional topics such as road condition, maintenance and technical condition assessment to more modern approaches based on digitalization, sensor systems, automated fault detection and machine learning tools. This development is also in line with the results of other published works that point to the growing importance of non-destructive methods, smart monitoring and data-driven decision-making in infrastructure asset management [13,15]. An important observation is the issue of improving the efficiency of infrastructure management in the areas of mapping its technical condition in order to obtain the greatest value from the obtained data, i.e., the question of “which road sections are actually suitable for detailed diagnostics”.

From the above, it follows that the identified research gap does not lie in the lack of knowledge about the technologies themselves for diagnosing the condition of road sections, but in the absence of a connection between the technical side of diagnostics and its strategic application in places where it brings the greatest effect. This process framework should serve to prioritize the diagnostics of road sections within the entire network and answer the question: “Which sections of my entire managed network should we allocate for detailed and expensive diagnostics of its technical condition in order to maximize the benefit from the limited budget for these processes?” The identified problems are the basis for generating approaches that would be able to integrate the criteria of information value, level of uncertainty and economic efficiency of diagnostic processes. The potential for further research thus lies in the creation of methodological innovations that will allow the formulation of models and decision-making rules for optimizing the selection of road sections for diagnostics, at the strategic network level. The specific tasks of future research were primarily to create theoretical and methodological assumptions for the application of the theory of the value of information in the area of decision-making on the performance of inspections of the construction and technical condition of sections of the road network. It will certainly be interesting to analyze what types of decision-making situations this theoretical approach is able to cover and what types of data, uncertainties and expected benefits it would have to take into account. Parallel research should also focus on identifying the links between the concept of the value of information and existing approaches used in asset management, risk-based decision-making and planning of diagnostic activities. It will be equally important to gradually clarify the possibilities of creating space for the theory of the value of information to be considered as a separate methodology within decision-making systems and not as a complementary part of other decision-making frameworks. In this context, it seems appropriate to carry out more research tasks following the need for a more systematic definition of concepts, categories and relationships that would allow conceptualizing the information contribution of diagnostics not only as a technical output, but also as a factor entering into strategic decision-making at the network level.

The above-mentioned research directions have the potential to bring not only increased efficiency in infrastructure management, but also a more transparent use of public resources. Opening a new theoretical space for research can fundamentally contribute to reducing decision-making uncertainty, increasing the reliability of planning processes and maximizing the value of the information obtained.

6. Conclusions

Modern diagnostic technologies for inspecting the structural and technical condition of road sections represent investment- and operationally demanding solutions. Administrators apply them in two cases. The first case is the phase of checking the technical condition of the road before and after a planned intervention, e.g., after reconstruction (decision-making at the project level, where the location for diagnostics is already selected). These deployments are not prioritized, but are applied directly according to real need. The second case of their application is based on the need to increase the efficiency of road network management by expanding knowledge about assets. This form of technical condition monitoring cannot be applied across the board and regularly to the entire road network. Due to limited financial resources, administrators are forced to select and prioritize which sections should be focused on for detailed and often financially costly diagnostics. The problem appears to be the fact that the selection is done without a formalized methodological framework containing objective criteria for selecting locations with the greatest information benefit. On the contrary, vague expert approaches are applied that do not guarantee the greatest economic and information effect. Risks associated with road network management are managed from this point of view in an informal and rather intuitive manner. Current approaches often fail to quantify either the level of uncertainty about the condition of the asset or the potential loss (financial, social) resulting from an incorrect decision made on the basis of uncertain data. The absence of a credible framework reduces the reliability and defensibility of existing decision-making processes and complicates the deployment of a truly complex and optimized approach to solving the allocation problem associated with the selection of sections for detailed diagnostics of the technical condition.

The aim of the work was to examine the possibilities of applying the concept of information value in the processes of diagnostics and asset management of road infrastructure. Based on the analysis of the literature, it can be stated that current research shows a growing interest in linking technical condition diagnostics with decision-making processes, with data-oriented, sensory and digital approaches being dominantly developed. The result of the research is a synthesized overview of knowledge, which contributes to a better understanding of the possibilities of using the concept of information value in this area and identifying directions for further research. At the same time, it is shown that the explicit application of the concept of information value in planning road infrastructure inspections remains limited and methodologically insufficiently developed.

The main conclusion of the work is that the research gap is not related to the lack of diagnostic technologies, but to the absence of a methodological framework that would allow linking technical diagnostics with decision-making at the strategic network level. The authors’ results in this paper thus create a basis for further research focused on the theoretical and methodological elaboration of the conditions for the possible use of the concept of information value in road infrastructure management.