Environmental Effects in Life Cycle Assessment of Machine-Vision-Driven Spall Repair Material Estimation for Sustainable Road Maintenance

Abstract

Portland cement concrete is widely used in road construction due to its durability and minimal maintenance needs. However, its susceptibility to spall highlights the drawbacks of conventional repair methods, including cost inefficiencies, delays, environmental impacts, and safety risks from road closures. To address these challenges, this study evaluated the environmental benefits of a spall detection and repair method employing artificial-intelligence-based computer vision technology. By utilizing machine vision techniques, this approach detects spall damage without road closures and automates the calculation of repair areas and material requirements through a proprietary estimation program. Environmental impact assessments were conducted using life cycle assessment across three frameworks, TRACI, ReCiPe, and ILCD, to compare this method with conventional practices. The results revealed a 79% reduction in the overall environmental impacts, including significant decreases in global warming due to shorter road closures and reduced material waste. Resource usage improved through optimized processes, and air pollution decreased, with lower emissions of smog and particulates. This study highlights the potential of machine-vision-driven repair material quantity takeoff as a more efficient and sustainable alternative. The results of this study will help institutional engineers and practitioners adopt sustainable strategies for green infrastructure repair and integrate them into various infrastructure maintenance practices to contribute to the development of sustainable urban environments.

1. Introduction

Portland cement concrete (PCC) is a foundational material for road infrastructure due to its durability and low maintenance requirements. However, spall damage—characterized by cracking, flaking, or chipping—presents persistent challenges that compromise structural integrity, increase repair costs, and exacerbate operational inefficiencies [1]. Addressing spall damage effectively is crucial for maintaining reliable infrastructure, but existing detection and repair methods often fall short due to inefficiencies, environmental impacts, and inaccuracies in determining the required amount of repair materials for long-lasting interventions.

Existing spall detection methods rely heavily on manual visual inspections conducted by on-site workers. While this method is straightforward, it is inherently subjective and prone to human error, as the outcomes depend heavily on the inspector’s experience and judgment [1]. This inconsistency often results in inaccurate assessments of the damage extent, potentially exacerbating structural degradation if early-stage spall is missed. Alternative methods such as ultrasonic testing have been developed to improve the detection accuracy; however, their high cost and resource-intensive nature limit their scalability for large-scale applications [2]. Moreover, even when spall areas are correctly identified, the estimation of necessary repair materials often relies on rough approximations, leading to either underestimation, which compromises the repair durability, or overestimation, which increases the material waste and costs.

The limitations of existing repair methods further compound these challenges. Techniques such as partial depth repair (PDR) involve removing and replacing damaged concrete with new materials. The repair sequence for concrete spall typically involves marking the repair boundaries, removing deteriorated concrete, preparing the repair area, applying a bonding agent, placing the repair material, curing, and, if needed, diamond grinding and resealing joints [3]. These steps are labor-intensive and require skilled workers, which increases project timelines and costs [4]. Although widely used, PDR is susceptible to errors in determining the correct size of repair areas and the precise volume of repair materials, leading to stress concentrations at the repair boundaries and subsequent failures [5]. The longevity of these repairs is heavily dependent on the quality and quantity of the applied materials. Inadequate bonding and improper material volume calculations often result in premature failures, necessitating further interventions [6]. Furthermore, existing repair methods focus on addressing visible damage rather than the underlying causes, such as moisture infiltration or thermal expansion [7], making accurate and efficient material allocation even more critical to prevent recurring issues.

These conventional practices also have significant environmental impacts. The production of PCC contributes approximately 5% of global carbon dioxide (CO₂) emissions due to the energy-intensive calcination process [8]. Additionally, the road closures required for manual inspections and repairs lead to traffic congestion, increased vehicle idling, and higher emissions. According to studies, construction work zones (CWZs) causing heavy congestion at 5 mph can potentially increase fuel consumption by 85% and CO₂ emissions by 86%. Reducing congestion to medium levels lowers fuel use by 40% [9]. Furthermore, excessive material use and waste generation, exacerbated by inaccurate material estimations, amplify the environmental footprint of spall management. Construction waste, including concrete, represents a significant portion of urban waste, accounting for 10–30% of the total waste received at landfill sites globally, highlighting the urgent need for more efficient and sustainable management approaches [10]. Improving the precision of repair material estimation can thus significantly reduce both the financial and environmental costs associated with spall maintenance.

In addition to addressing the material inefficiencies, energy-saving strategies play a critical role in minimizing the environmental impacts of transportation infrastructure maintenance. Integrating energy-efficient practices, such as optimizing vehicle operations during road repairs and leveraging low-energy technologies, can significantly reduce fuel consumption and emissions [11]. These strategies complement the advancements in repair precision by targeting indirect sources of environmental degradation, such as prolonged vehicle idling during road closures.

Recent advancements in artificial intelligence (AI), machine vision (MV), and computer vision (CV) have introduced promising solutions for improving spall detection and repair. These technologies leverage data-driven algorithms to provide objective and consistent assessments, reducing the subjectivity associated with existing visual inspections. Techniques such as convolutional neural networks (CNNs) and region-based convolutional neural networks (R-CNNs) have demonstrated high accuracy in detecting cracks and spall on concrete surfaces [12,13]. For instance, Wu et al. [14] utilized unmanned aerial vehicles (UAVs) combined with CNNs to classify damage types and assess structural health, achieving efficient crack and spall detection. However, early implementations, such as edge detection combined with CNNs, were limited by the residual noise and low precision for finer cracks, as demonstrated by Dorafshan et al. [15], where the accuracy ranged from 53% to 79%. To address these limitations, deep CNNs were introduced, achieving 86% accuracy in crack detection and significantly reducing noise. Additionally, deep ensemble learning models have been applied to road damage detection, enabling the classification of various types of road damage, including spall, with improved accuracy and robustness [16]. Furthermore, to overcome computational inefficiencies, lightweight models such as YOLO were explored. Mandal et al. [17] demonstrated the application of YOLO for road damage detection, which achieved high accuracy for larger defects. However, YOLO struggled with smaller objects and lacked the fine-grained segmentation capability necessary for precise spall detection. More advanced segmentation models, including UNet and EfficientNet, addressed these shortcomings by providing pixel-level segmentation for detailed damage characterization. Wang et al. [18] proposed a cascaded network combining VGG-19 and ResNet-50 to effectively segment concrete spalling and exposed steel bars, achieving an F1-score of 0.925 for spalling detection. Despite these improvements, challenges remained in terms of detecting smaller defects and processing images efficiently. Building upon these efforts, Mask R-CNN was employed by Kumar et al. [19] to enhance the detection precision, especially for spalling at the pixel level.

The integration of AI with practical applications has expanded significantly. Idjaton et al. [20] proposed that advanced deep learning techniques were applied to 3D survey images for limestone spalling detection, achieving pixel-level precision while addressing the challenges of 3D spatial data processing. This study demonstrated the ability to integrate high-resolution data with AI algorithms to detect and classify structural defects in complex environments. Yasmin et al. [21] applied semantic segmentation using deep architectures for spall severity classification in concrete structures. This approach provided a more granular understanding of the damage severity, enabling targeted repair strategies that optimize material usage and reduce unnecessary interventions. AI-based systems can also automate processes such as repair planning, optimizing operational workflows, and minimizing resource waste [22]. The existing studies have been summarized in

Table 1. Summary of the literature review.

Authors	Methods	Results	Limitations	Research Gap
Wu et al. [14]	Combined UAVs with CNNs for damage classification and structural health assessment	Achieved efficient crack and spall detection	Potential limitations in UAV image resolution and data quality; scalability issues not addressed	Need for scalable and accurate UAV-based inspection methods
Dorafshan et al. [15]	Compared deep CNNs and edge detectors for image-based crack detection in concrete	Achieved higher accuracy using deep CNNs compared to traditional edge detectors	May require extensive training data; higher computational cost for CNNs compared to simpler edge detectors	Optimize computational efficiency and reduce dependency on large labeled datasets
Manda et al. [17]	Automated road crack detection using deep CNNs	Achieved high accuracy in classifying road cracks	High computational demand; sensitivity to image quality and lighting conditions	Develop more robust models that handle varying image conditions and reduce computational requirements
Doshi and Yilmaz [16]	Implemented deep ensemble learning models for road damage detection	Enhanced accuracy and robustness in classifying various road damages	May require significant computational resources; ensemble methods can be complex to deploy	Enhance ensemble methods to cover broader damage classifications effectively
Le et al., [12]	Developed CNNs for crack recognition	Achieved high accuracy in detecting cracks	Dependency on labeled data; potential overfitting to specific datasets	Need for objective and consistent crack detection methods, reducing reliance on subjective manual inspections
Kumar et al. [19]	Employed Mask R-CNN for multiclass instance segmentation of concrete damage	Enhanced detection precision, especially at the pixel level	High computational cost; may require extensive training data for accurate segmentation	Improve detection precision for spalling at the pixel level to ensure more accurate assessments
Yu et al. [13]	Utilized deep convolutional neural networks optimized by the enhanced chicken swarm algorithm for crack detection	Demonstrated effective crack detection capabilities	Optimization process may be time-consuming; applicability to varied datasets not tested	Enhance detection algorithms to achieve better precision and reliability
Idjaton et al. [20]	Applied advanced deep learning techniques to 3D survey images for limestone spalling detection	Achieved pixel-level precision; integrated high-resolution data with AI algorithms to detect and classify structural defects in complex environments	Challenges of 3D spatial data processing; potential computational intensity	Enhance integration of high-resolution 3D data with AI algorithms for more efficient processing and detection
Yasmin et al. [21]	Utilized semantic segmentation using deep architectures for spall severity classification in concrete structures	Provided granular understanding of damage severity; enabled targeted repair strategies that optimize material usage and reduce unnecessary interventions	Dependency on large annotated datasets	Optimize material usage and reduce unnecessary interventions through targeted repair strategies
Wang et al. [18]	Automatic detection method for concrete spalling and exposed steel bars in reinforced concrete structures based on machine vision	Achieved high accuracy in automatically detecting and classifying concrete spalling and exposed steel bars	High computational cost for processing high-resolution images; limited model generalization across varied environmental conditions	Enhance computational efficiency for real-time application and improve model generalization across diverse environments

.

However, existing studies have focused only on the detection accuracy without addressing operational inefficiencies such as delays due to the overuse of materials and inaccurate calculations. Therefore, this study proposes a process that can go beyond the existing studies to calculate the amount of maintenance materials.

Building upon prior studies, the proposed method utilizes a line-scan camera mounted on a vehicle, combined with deep learning algorithms like Mask R-CNN, to automate spall detection and repair material estimation. This innovative approach eliminates the need for road closures during data collection, allowing for non-intrusive, high-precision imaging of pavement conditions. By analyzing the captured images in a controlled indoor environment, this method not only reduces traffic disruptions and associated emissions but also ensures consistent material estimation. These advancements help reduce the need for prolonged road closures, thereby alleviating traffic congestion and related emissions [23].

Compared to conventional, manual inspection methods, the proposed AI-driven approach offers notable advantages when the goal is to accurately determine the quantity of repair materials. First, automated spall detection reduces the likelihood of under- or overestimation, minimizing wasted PCC and ensuring that repairs are conducted only where necessary. Second, because line-scan cameras can capture detailed imagery of the pavement surface while the vehicle is in motion, data acquisition does not require extended road closures. This feature not only decreases traffic congestion and associated emissions but also lowers labor costs and shortens project timelines. Third, the consistent, data-driven output from AI algorithms mitigates the subjectivity inherent in manual inspections, allowing engineers to standardize repair volume calculations and reduce the variability in results. Nevertheless, the conventional manual approach retains some utility in specific scenarios—particularly those involving small-scale or localized repairs—due to its low initial equipment investment and simple on-site decision-making. However, as the project scale and the need for accurate material budgeting grow, the drawbacks of relying on human perception (e.g., potential for human error, safety risks to inspectors, longer closure times) become more pronounced. In contrast, the MV-driven method effectively addresses these concerns by systematically capturing objective, high-resolution data across large roadway segments in a single pass. Consequently, for agencies and contractors aiming to optimize both environmental outcomes and cost-efficiency through precise PCC volume estimations, the proposed system represents a significant improvement over traditional manual inspection.

While these technologies show potential for improving not only the efficiency and accuracy of spall management but also the precise calculation of repair material volumes, their environmental benefits compared to existing methods remain insufficiently quantified. Prior studies, such as Aryan et al. [24], have emphasized the need for comprehensive LCA approaches that consider all the life cycle phases, including the use and end-of-life stages, and account for multiple environmental impact categories beyond the global warming potential and energy demand. Lei et al. [25] demonstrated significant reductions in energy consumption and carbon emissions through life cycle assessment (LCA) of optimized pavement renovation methods like resonant rubblization in situ regeneration (RRIR) and emulsified asphalt cold recycling (EPCR). Additionally, they highlighted the importance of incorporating sensitivity analysis and localized assessments into LCA studies. However, these evaluations have focused on large-scale pavement renovation and new material recycling, leaving gaps in the specific context of spall detection and repair methods. To address this gap, this study applies LCA to compare the environmental impacts of MV-driven and conventional spall detection and repair material estimation methods. The goal is to determine whether MV-driven methods, through more accurate material quantification, offer measurable reductions in environmental burdens and operational inefficiencies, thereby bridging the existing research gap in the field.

This paper is structured as follows. Section 2 details the methodology, including the LCA framework and system boundaries. Section 3 presents the results and discussion of the environmental impact assessment using the LCA framework. Finally, Section 4 contains the conclusions and future research.

2. Methodology

This study seeks to address the issues associated with existing spall material quantity takeoff repair methods, such as time delays, cost inefficiencies, toxic emissions, safety concerns, and delays for road users, by proposing a new spall detection and repair process utilizing a vehicle equipped with MV.

2.1. Overview of Proposed Spall Material Quantity Takeoff Method

The proposed method introduces an MV-driven spall detection and repair system utilizing a vehicle equipped with a high-performance line-scan camera. This innovative approach aims to enhance the operational efficiency and greatly reduce the environmental impacts associated with traditional spalling repair methods. The system follows a streamlined three-step process designed to ensure precision, sustainability, and minimal disruption to road users.

First, a line-scan camera, mounted on a moving vehicle, captures high-resolution, detailed images of the pavement surface as the vehicle traverses the road. This setup eliminates the need for road closures or extensive traffic management during the detection phase, ensuring smooth traffic flow and reducing vehicle idling times. The use of line-scan cameras instead of conventional imaging techniques addresses issues related to the image quality and scale, enabling accurate detection of spalling across extensive road sections.

Second, the captured images are analyzed using advanced AI algorithms, specifically leveraging Mask R-CNN, a state-of-the-art deep learning model known for its high accuracy in object detection and segmentation at the pixel level. These algorithms detect spalling with precision, identifying both the location and the dimensions of damaged areas. Unlike traditional manual inspections, which are prone to errors due to human fatigue or inconsistent criteria, this automated approach ensures consistent and reliable results. By incorporating AI-driven analysis, the system substantially reduces the time and resources required for detecting spalling, while simultaneously improving the detection accuracy.

Finally, the detected spalling areas are fed into a developed calculation program, which determines the repair area and calculates the precise amount of PCC needed for the repair. The program adheres to the standards outlined by the Federal Highway Administration (FHWA), ensuring globally recognized accuracy and reliability. This precision minimizes material waste, reduces overestimation and underestimation errors, and ultimately leads to more efficient resource utilization. The ability to accurately estimate material quantities also mitigates the environmental impact associated with excess material production and transportation. This proposed method addresses critical challenges inherent in traditional spalling repair processes. By eliminating the need for road closures during detection, it significantly reduces the fuel consumption and emissions caused by idling vehicles. The integration of AI-driven analysis and automated calculation ensures greater consistency and reliability in repair planning, overcoming the inaccuracies often associated with manual methods. Moreover, the program visualizes spalling data and calculates repair quantities rapidly, enabling engineers to make informed decisions in less time. As demonstrated in Figure 1, this system effectively estimates repair areas and material quantities with unparalleled precision, paving the way for an environmentally sustainable and operationally efficient approach to spall detection and repair.

To assess the environmental benefits of the proposed process, an LCA was conducted. LCA is a systematic framework for evaluating the environmental impacts of a product or service throughout its life cycle and has become a widely used tool in sustainability assessments [26]. According to ISO 14044:2006, the term “product” in the context of LCA encompasses both goods and services. However, LCA primarily focuses on environmental impacts and typically does not address the economic or social dimensions of a product or service [26,27]. The four distinct phases of LCA—goal and scope definition, life cycle inventory analysis (LCI), life cycle impact assessment (LCIA), and interpretation—ensure a comprehensive evaluation of the inputs, outputs, and environmental impacts [26,27]. This study’s LCA focused on quantifying global warming, resource usage, and energy consumption to evaluate the sustainability of the proposed method compared to existing methods.

In this study, the LCA was conducted to demonstrate whether the proposed method for calculating the required materials for spall repair offers actual environmental benefits, as shown in Figure 2. LCA provides a comprehensive approach for assessing the environmental impact of spall repair by quantitatively analyzing the global warming, resource usage, and energy consumption at various stages [28].

The analysis was structured using OpenLCA v2.3.0, a tool that supports multiple databases and impact assessment methodologies, such as the tool for reduction and assessment of chemicals and other environmental impacts (TRACI), ReCiPe, and international life cycle data system (ILCD) [29]. OpenLCA’s user-friendly interface and flexible functionality made it well suited for validating the effectiveness of the proposed method. It facilitated a direct comparison between existing and MV-driven spall repair material quantity takeoff methods, enabling a detailed evaluation of the environmental benefits. The analysis focused on key metrics such as energy consumption, material inputs, and emissions, which were collected through LCI analysis.

This study performed a cradle-to-gate LCA focusing on spall repair material quantity takeoff methods in concrete pavements. Two primary methods were analyzed: the existing spall repair material quantity takeoff method and the proposed method utilizing a vehicle equipped with a line-scan camera. The functional unit for analysis was defined as a 1 km section of a two-lane concrete pavement road [30,31]. This selection supports standardized environmental impact assessments in pavement life cycle studies by providing a consistent basis for comparison. By adopting this well-defined functional unit, this study improves the applicability and transferability of its findings to other infrastructure evaluations. The system boundaries were drawn from raw material acquisition for PCC production to the completion of spall repair activities, excluding end-of-life recycling or long-term maintenance cycles. This cradle-to-gate perspective maintains consistency with standard road maintenance LCA approaches while highlighting the immediate environmental burdens associated with detection and repair operations. The key boundary conditions and parameters include the following:

• Energy and fuel consumption: The existing method involves high fuel consumption due to road closures and heavy machinery usage, while the proposed method minimizes fuel use during the spall detection phase without requiring road closures.
• Material use: The proposed method allows for precise calculation of the required repair materials, reducing waste compared to existing methods.
• Traffic emissions: Emissions from idling vehicles due to road closures are included for the existing method, whereas the proposed method assumes no idling emissions.
• Impact assessment framework: Environmental impacts, global warming, resource usage, and pollution are evaluated using OpenLCA software v2.3.0 [32].

This system boundary design allows for a detailed comparison of the environmental impacts, highlighting the sustainability improvements offered by the proposed spall material quantity takeoff repair method.

The proposed method was evaluated using OpenLCA, where various processes were set up to define the system boundaries and assess the environmental impacts at each stage. OpenLCA supports various databases and impact assessment methodologies, making it suitable for comprehensive environmental impact analysis. Additionally, its user-friendly interface and flexible functionalities are ideal for verifying the effectiveness of the proposed method. This allowed for a comprehensive comparison of the environmental impacts between the existing and proposed methods, validating the environmental effectiveness of the proposed approach.

2.2. Inventory Analysis

The inventory analysis quantified the significant inputs and outputs for both methods within clearly defined boundaries. The first boundary addresses the interaction between the technical system and the environment, capturing resource use and emissions. For instance, the MV-driven system consumes 10 L of diesel during detection, emitting 100 kg of CO₂, compared to 50 L and 160 kg for the existing method. The AI system also reduces labor demands, requiring only one worker instead of eight. These efficiencies highlight its lower environmental burden during detection and repair.

The second boundary differentiates significant processes, such as fuel consumption, emissions, and traffic delays, from less impactful ones like administrative tasks. Significant processes are prioritized due to their direct contribution to global warming, energy use, and congestion, ensuring the analysis remains focused on key factors.

The third boundary considers shared resources and processes, such as traffic management and material use. The AI system reduces the total road closure time by 32 h or 4 working days by eliminating the need for road closures during detection, thereby reducing traffic-related emissions. While both systems use PCC, the AI system minimizes waste by optimizing the repair precision.

In the LCI analysis, data on energy consumption, material inputs, and emissions for each method were collected. To ensure the reliability and reproducibility of the LCI data, this study employed a combination of primary and secondary data sources. The primary data included operational logs, direct fuel consumption measurements, and material input records obtained from the actual spall detection and repair operations. For the secondary data, such as emission factors and upstream resource extraction impacts, this study utilized the OpenLCA LCIA Methods v2.0.2 package within the OpenLCA platform. Where data gaps existed, conservative estimates or proxy values from similar road maintenance projects were adopted [33]. This multi-source data collection strategy not only enhanced the robustness of the inventory analysis but also ensured that the assessment aligns with internationally recognized standards and best practices. The existing spall material quantity takeoff repair method is characterized by high fuel consumption due to heavy machinery use and additional emissions from traffic congestion caused by road closures. In contrast, the proposed method focuses on minimizing fuel consumption by utilizing a vehicle with a line-scan camera, and on reducing environmental impacts by minimizing the quantity of materials required during the repair process. The collected data were used as input for OpenLCA to calculate and compare the environmental impacts of each stage.

To enhance the accuracy of the analysis, this study conducted the LCA based on standard design and construction practices for concrete pavements. The existing spall material quantity takeoff repair method involves fully removing the damaged area and restoring it with PCC. PCC production generates significant CO₂ emissions during the calcination and fuel combustion processes. This study estimated the quantity of PCC required for each method, reflecting the fact that the existing method demands a higher amount of PCC compared to the proposed method.

Based on this 1 km section, the key assumptions for the LCI analysis are as follows:

• Vehicle fuel consumption: In the existing method, the idling time caused by traffic congestion during road closure is included as a factor affecting fuel consumption. In the proposed method, energy consumption by the line-scan camera and vehicle is only used during the spall detection phase, with no idling assumed.
• Energy consumption and material input: Given that the line-scan camera and vehicle consume far less energy compared to heavy machinery, it is assumed that the proposed method uses minimal energy and materials during the repair process.
• Material usage: Material use is based on the amount of PCC required by each method, and it is assumed that the proposed method results in more accurate material quantities, reducing waste.
• Traffic congestion and emissions: The analysis assumes an average traffic density in a congested urban section to estimate the additional emissions during road closure. The increase in global warming due to traffic congestion is factored in, quantifying the environmental impact of congestion on emissions.

2.3. Life Cycle Impact Assessment

The LCIA phase evaluates the environmental impacts of the MV-driven and existing spall material quantity takeoff repair methods using three established frameworks.

The TRACI is a widely used LCA methodology developed by the U.S. Environmental Protection Agency (EPA) [34] to assess the environmental impacts of various processes, including construction and infrastructure repair. It provides impact characterization factors that help quantify the potential environmental effects of emissions and resource use [35]. The TRACI categorizes impacts into several areas, including global warming potential, and human health impacts, making it a versatile tool for evaluating the environmental performance of infrastructure repair methods [36]. The TRACI is often considered the de facto standard for life cycle assessments in the United States. Given that many references and emission factors used in this study are North-America-oriented, the TRACI provides a reliable framework aligned with U.S. environmental policies. The method includes multiple categories (e.g., acidification, ecotoxicity, smog) that are directly relevant to transportation infrastructure projects. In the context of infrastructure repair, the TRACI can be employed to assess the emissions associated with different repair materials and techniques. For instance, a study by Chester and Horvath utilized the TRACI to evaluate the environmental impacts of high-speed rail infrastructure, demonstrating its applicability in assessing transportation projects [35]. By using the TRACI, decision-makers can identify repair methods that minimize environmental harm while maintaining structural integrity.

The ReCiPe is another prominent LCA methodology that combines midpoint and endpoint indicators to provide a comprehensive assessment of environmental impacts. It was developed by the Dutch National Institute for Public Health and the Environment (RIVM) and other European research institutions, and it is widely recognized for its robustness and flexibility [37,38]. The ReCiPe categorizes impacts into several areas, including climate change, human health, and ecosystem quality, allowing for a nuanced understanding of the tradeoffs associated with different repair methods. The ReCiPe offers a broader, internationally recognized perspective, making it complementary to the TRACI’s North American focus. By covering both midpoint (e.g., particulate matter formation, photochemical oxidant formation) and endpoint (e.g., ecosystem quality, human health) indicators, the ReCiPe enables a more holistic analysis of potential environmental consequences. The relevance of the ReCiPe to infrastructure repair is evident in its ability to assess the long-term impacts of repair strategies on both the environment and human health. For example, Palacios-Munoz et al. [39] employed the ReCiPe to compare the sustainability of refurbishment versus new constructions, highlighting the importance of considering the entire life cycle of infrastructure projects. By applying the ReCiPe, stakeholders can make informed decisions about repair methods that align with sustainability goals.

The ILCD is a framework developed by the European Commission to promote consistency and transparency in LCA studies [40]. It provides guidelines for conducting LCAs and offers a comprehensive database of life cycle inventory data [41]. The ILCD methodology emphasizes the importance of data quality and completeness, making it a valuable resource for infrastructure repair assessments. Aligned with European environmental policies, the ILCD evaluates impact categories such as climate change, freshwater acidification, ecotoxicity, and photochemical ozone creation. This ensures that assessments are robust and policy-relevant. In infrastructure repair, the ILCD ensures reliable and transparent data for comparing the environmental impacts of different materials and techniques, aiding stakeholders in selecting the most sustainable repair strategies [42]. Although it originated within the European Union (EU), the ILCD can be applied globally. This complements both the TRACI and ReCiPe, which also have well-established user bases and datasets, offering a diverse yet harmonized suite of impact indicators.

By integrating the TRACI, ReCiPe, and ILCD in parallel, this study demonstrates benefits for environmental assessment. The advantage is that these frameworks collectively allow early-stage information on potential environmental hotspots—such as excessive CO₂ emissions, resource depletion, or toxic releases—without waiting for the entire life cycle of a repair strategy to conclude. Specifically, the TRACI can highlight the stages of spall repair that produce the greatest fuel consumption or contribute most to smog formation, while the ReCiPe captures broader ecosystem and human health implications. The ILCD standardized data requirements and strong alignment with European policy further ensure that the assessments remain transparent, comparable, and well grounded in reliable inventory data. In this way, agencies can conduct a robust LCA-based analysis to identify cost-effective improvements—like reducing material overestimation or minimizing road closures—early on, eliminating the need for extensive, long-term field monitoring. Through doing so, stakeholders gain the ability to refine maintenance schedules, allocate budgets more effectively, and enact proactive interventions that enhance both environmental sustainability and operational efficiency in spall repairs.

In order to conduct a comprehensive LCIA using the TRACI, ReCiPe, and ILCD methods, it is essential to gather consistent input data related to the materials, energy consumption, and transportation requirements of both the existing and proposed MV-driven spall material quantity takeoff repair methods. These key input parameters are summarized in

Table 2. Input data for spall repair for both the existing and proposed methods.

		1. Detection		2. Repair		3. Traffic Impact
Flow	Unit	Proposed	Existing	Proposed	Existing	Proposed	Existing
Diesel	L	10	50	38	38	1000	1624
Labor	Person	1	8	10	10	-	-
Electricity	kWh	1.5	-	-	-	-	-
Line-Scan Camera Truck	Vehicle/h	1/0.5	-	-	-	-	-
Air Compressor Cutter	h	-	-	20	20	-	-
PCC Dry	kg	-	-	346	352	-	-
Traffic Signal Truck	Vehicle/h	-	3/32	3/72	3/72	-	-

, providing the necessary foundation for the subsequent impact assessment.

3. Results and Discussion

The LCA conducted using the TRACI, ReCiPe, and ILCD methodologies highlights the significant environmental benefits of the proposed spall material quantity takeoff repair method compared to traditional approaches. By quantifying the impacts across multiple categories—ranging from the global warming potential to toxicity and resource depletion—these results offer a comprehensive view on how optimizing spall repair procedures can yield substantial sustainability gains. In the TRACI assessment, the proposed spall repair material quantity takeoff method exhibited markedly lower environmental impacts compared to the traditional approach across all the evaluation categories.

The TRACI categorizes its environmental impact assessments into the following major groups:

• Global warming potential: Global warming (kg CO₂ equivalent (eq)).
• Energy use: Fossil fuel depletion (megajoule (MJ)).
• Air pollutants: Acidification (kg sulfur dioxide (SO₂) eq), respiratory effects (kg particulate matter (PM) 2.5 eq), and smog (kg ozone (O₃) eq).
• Resource and ecosystem impact: Ecotoxicity (comparative toxic unit equivalent (CTUe)).

In terms of the global warming potential, the proposed method decreased emissions from 64.16 kg CO₂ eq to 13.29 kg CO₂ eq—a reduction of 79.28%. This substantial reduction demonstrates how shortened road closures and the elimination of idling vehicles directly lower fuel consumption and CO₂ emissions. Notably, this improvement carries broader climate change mitigation significance, particularly in urban or heavily trafficked areas where vehicular emissions are already a public concern.

Reductions in air pollutants were also observed, including smog-related emissions, which decreased by 79.44%, and particulate matter emissions contributing to respiratory effects, which saw a 79.21% reduction. Similar patterns were seen with regard to the acidification potential, which decreased by 79.53%, and ecotoxicity, which dropped by 79.94%. These findings underscore that even modest changes—such as minimizing traffic congestion or refining the precise quantity of concrete used—can drastically cut emissions linked to nitrogen oxides (NOx), SO₂, and other harmful pollutants. These pollutants, typically exacerbated by traffic delays and prolonged road closures due to existing methods, were significantly mitigated through the optimized operation of the proposed system.

• The integration of machine vision (MV) technology allows for real-time detection of spall without necessitating road closures. This efficiency not only reduces the time required for inspections but also minimizes the operational downtime of repair crews, leading to significant fuel savings and lower CO₂ emissions.
• Accurate estimation of repair materials through MV technology minimizes overestimation and underestimation, leading to reduced material waste and lower environmental footprints. This precision ensures that only the necessary amount of Portland cement concrete (PCC) is used, which directly contributes to the observed reductions in ecotoxicity and resource depletion.
• By streamlining the repair process, the proposed method reduces the reliance on heavy machinery and extensive labor hours, further decreasing the energy consumption and associated emissions. This optimization not only enhances environmental performance but also improves overall project timelines and cost-effectiveness.

The TRACI results emphasize several key environmental benefits of the proposed method. By preventing vehicle idling, the method directly reduces fuel consumption and CO₂ emissions, aligning with climate mitigation objectives. Furthermore, the method significantly decreases emissions of air pollutants, such as NOx and particulate matter, which contribute to smog formation and respiratory health issues. Additionally, the reduction in road closure times minimizes traffic congestion, leading to further reductions in emissions from vehicles caught in gridlock.

The findings demonstrate that the proposed method not only improves the efficiency of spall detection and repair processes but also significantly alleviates environmental burdens, particularly in densely populated urban areas with high traffic volumes. From a practical standpoint, implementing the MV-driven spall detection system may require initial investments in equipment, staff training, and integration with existing maintenance protocols. Nevertheless, once established, the system’s rapid detection capabilities and reduced need for prolonged road closures can streamline maintenance scheduling and coordination. Over larger road networks, cumulative environmental savings may become even more pronounced, as agencies could strategically deploy the technology in high-traffic corridors to maximize congestion relief and emission reductions. The scalability and adaptability underscore the method’s potential as part of a broader sustainable asset management strategy in transportation infrastructure. The outcomes align with broader sustainability goals in terms of infrastructure maintenance and underscore the method’s potential to serve as a more environmentally responsible alternative to existing spall repair practices.

Table 3. TRACI results.

Impact Category	Existing Method	Proposed Method	Reduction (%)
Acidification (kg SO2 eq)	0.419996325	0.086008587	79.53
Ecotoxicity (CTUe)	42.83626922	8.592969417	79.94
Fossil Fuel Depletion (MJ)	359.0908967	72.2753667	79.87
Global Warming (kg CO2 eq)	64.16456518	13.29106932	79.28
Respiratory Effects (kg PM2.5 eq)	0.019476133	0.004050296	79.21
Smog (kg O3 eq)	3.445408793	0.707922127	79.44

and Figure 3 quantitatively support these observations by showing the magnitude of the reductions across impact categories, reinforcing the notion that small operational adjustments can yield large-scale environmental benefits.

While the TRACI assessment focuses primarily on regional impacts and is well suited for analyzing U.S.-based environmental factors, the ReCiPe evaluation provides a broader, global perspective.

In the ReCiPe evaluation, the proposed spall material quantity takeoff repair method exhibited significantly lower environmental impacts across most categories compared to the existing method.

The ReCiPe categorizes its environmental impact assessments into the following major groups:

• Global warming potential: Climate change (kg CO₂ eq).
• Air pollutants: Particulate matter formation (kg PM10 eq), photochemical oxidant formation (kg non-methane volatile organic compounds (NMVOC)), and terrestrial acidification (kg SO₂ eq).
• Toxic releases: Human toxicity (kg 1,4-dichlorobenzene (DCB) eq).
• Resource and ecosystem impact: Water depletion (m³).

Notably, the reductions of 79.83% in human toxicity (kg 1,4-DCB-Eq) and 20.27% in water depletion (m³) highlight the proposed method’s superior performance. These improvements are attributed to the optimized equipment usage and reduced traffic congestion, which collectively enhance resource conservation and minimize pollutant emissions.

Beyond the significant reduction in the global warming potential by 79.69% (from 0.99 kg CO₂ eq to 0.20 kg CO₂ eq), the proposed method achieves similar substantial environmental benefits by decreasing resource consumption and emissions of hazardous pollutants. This outcome is particularly meaningful in a global context, where cement production and extended machinery operation often generate considerable GHGs and hazardous by-products. The streamlined process reduces the operational time of heavy machinery and shortens road closures, effectively lowering emissions of particulate matter and photochemical oxidants. Such gains not only improve local air quality but also contribute to broader global efforts to mitigate climate change and preserve ecological integrity. These benefits translate into a positive impact on both traffic congestion and the overall efficiency of road maintenance activities.

• Although lower than the other categories, the 20.27% reduction in water use aligns with sustainable water management practices, which is especially critical in regions facing water scarcity.
• The 79.68% reduction in NMVOC emissions indicates a substantial decrease in the precursors of ground-level ozone formation, enhancing both air quality and public health outcomes.

The ReCiPe results demonstrate that the proposed method is more environmentally sustainable than existing practices. This reinforces the potential for optimizing road maintenance processes and reducing traffic-related emissions to significantly improve the environmental outcomes of infrastructure repair operations. These findings underline the importance of adopting innovative approaches that address not only climate-related impacts but also broader sustainability objectives, making road repair operations more efficient and environmentally friendly. These improvements are quantitatively detailed in the ReCiPe evaluation outcomes, as shown in

Table 4. ReCiPe results.

Impact Category	Existing Method	Proposed Method	Reduction (%)
Climate Change (kg CO2 eq)	0.991615037	0.201467337	79.69
Human Toxicity (kg 1,4-DCB eq)	5.65768497	1.141349174	79.83
Particulate Matter Formation (kg PM10 eq)	0.05817287	0.011958033	79.44
Photochemical Oxidant Formation (kg NMVOC)	0.04961962	0.010083596	79.68
Terrestrial Acidification (kg SO2 eq)	0.281497097	0.057449052	79.60
Water Depletion (m3)	5.369538262	4.282614698	20.27

and Figure 4.

While the TRACI provides a regional analysis and the ReCiPe offers a global perspective, the ILCD evaluation is designed to be internationally applicable with a strong alignment with European environmental policies. The ILCD offers a framework for life cycle assessment that focuses on standardized data and detailed impact factors and ensures accuracy and consistency across life cycle data. Although primarily intended for use by organizations in the EU, the ILCD is applicable on a global scale.

The ILCD categorizes its environmental impact assessments into the following major groups:

• Global warming potential: Climate change (kg CO₂ eq).
• Air pollutants: Freshwater acidification (mol Hydron (H⁺) eq), terrestrial acidification (mol N eq), and photochemical ozone creation (kg NMVOC eq).
• Resource and ecosystem impact: Freshwater ecotoxicity (comparative toxic unit (CTU)).

In the ILCD evaluation, the proposed spall material quantity takeoff repair method significantly reduces the environmental impacts across multiple categories compared to the traditional approach. Key improvements are observed across various impact categories, including human health, ecotoxicity, acidification, and air pollution, still primarily due to reduced road closure times and the efficient use of technological equipment.

The proposed method reduces the freshwater ecotoxicity (CTU) impacts by 79.94%, reflecting a substantial reduction in harm to aquatic ecosystems. Freshwater acidification (mol H⁺ eq) sees a 79.59% decrease, and terrestrial acidification (mol N eq) is reduced by 79.99%, indicating significantly lower emissions of acid-forming pollutants such as sulfur and nitrogen oxides. Moreover, photochemical ozone creation (kg NMVOC eq) is reduced by 79.68%, lessening the formation of ground-level ozone, which is linked to respiratory health issues. These large-scale reductions underscore how limiting both the duration and the intensity of roadwork can substantially lower the atmospheric emissions and chemical runoff linked to concrete or machinery operations. These quantitative results highlight the environmental advantages of the proposed repair method and are summarized in

Table 5. ILCD results.

Impact Category	Existing Method	Proposed Method	Reduction (%)
Climate Change (kg CO2 eq)	0.992339757	0.201610086	79.69
Freshwater Acidification (mol H+ eq)	0.368968452	0.0752992	79.59
Freshwater Ecotoxicity (CTU)	15.46663303	3.103744042	79.94
Terrestrial Acidification (mol N eq)	0.272625077	0.054561223	79.99
Photochemical Ozone Creation (kg NMVOC eq)	0.049935326	0.01014676	79.68

and Figure 5.

These findings emphasize the environmental benefits of the proposed method, particularly in minimizing global warming and pollutant outputs. The method achieves these reductions through accurate material usage, shorter road closure times, and efficient energy consumption, all of which limit fuel use and reduce traffic congestion during repairs. Collectively, these efficiencies translate into measurable reductions in carbon emissions, acidifying pollutants, and various toxic or ecotoxic substances, supporting a more sustainable approach to road maintenance.

The LCIA conducted using the ILCD, TRACI, and ReCiPe methodologies confirmed that the proposed method offers significant environmental advantages over traditional approaches, with the main results being the following:

• Reduced CO₂ emissions (global warming): The lower CO₂ eq in the proposed method can be attributed to less idling traffic and minimized material waste through consistent output, reducing errors in PCC usage. Existing methods often lead to excessive PCC production, where the calcination of limestone (CaCO₃ → CaO + CO₂) in cement kilns is a major CO₂ source [43]. By minimizing material overestimation, the proposed approach reduces the volume of PCC produced and transported, thus cutting down the CO₂ released during calcination and fossil fuel combustion in cement plants.
• Acidification (SO₂ and NOx): The acidification potential is driven by the release of sulfur and nitrogen oxides during fuel combustion and certain industrial processes. Reduced machinery operation times, fewer detours, and less congestion in the proposed method mean fewer instances of incomplete combustion in vehicles and machinery. This reduces NOx emissions and, consequently, the formation of nitric acid (HNO₃) in the atmosphere [44].
• Respiratory effects and smog (PM2.5, O₃ precursors): Particulate matter (PM2.5) and ozone (O₃) precursors such as volatile organic compounds (VOCs) and NOx are byproducts of combustion processes and asphalt/tire wear under congested conditions [45]. Minimizing road closures reduces the prolonged stop-and-go traffic scenario, cutting down incomplete combustion and the formation of secondary pollutants. This leads to fewer fine particulates and ozone-forming reactions in the troposphere, directly improving air quality and reducing health-related burdens.
• Human toxicity and ecotoxicity (1,4-DCB eq): The proposed method’s lower ecotoxicity and human toxicity indicators arise from the reduced use of PCC and associated raw materials. Cement and aggregate extraction involves quarrying and processing, potentially introducing heavy metals and other toxic substances into ecosystems [46]. By using a consistent estimation of PCC quantities, the method diminishes the overall cradle-to-gate resource extraction and related pollution. This reduces chemical leachates and runoff containing substances that can be toxic to aquatic and terrestrial organisms.
• Particulate matter (PM10) and photochemical oxidant formation (NMVOC): Analogous to the TRACI’s PM2.5 and O₃ categories, the ReCiPe’s PM10 and NMVOC metrics reinforce how the reduced operational times of construction machinery and lower traffic congestion curb combustion-related emissions.
• Water depletion and resource use: Accurate PCC usage reduces not only chemical emissions but also the energy and water needed for cement production. Cement milling, aggregate washing, and concrete mixing all consume water [47]. By curbing material overuse, the proposed method indirectly lowers the demand for resource-intensive upstream processes, ultimately mitigating freshwater depletion and related habitat disturbances.
• Freshwater acidification and terrestrial acidification: These metrics underscore the importance of controlling SO₂, NOx, and ammonia-related emissions. Reduced machinery operation and minimized PCC production translate into fewer nitrogen- and sulfur-based emissions. Chemically, lowering NOx and SO₂ outputs curtails the formation of acids (HNO₃, H₂SO₄) in atmospheric reactions, diminishing acid deposition in soils and freshwater bodies [48].
• Photochemical ozone creation (NMVOC eq): Similar to the ReCiPe category, the ILCD’s focus on NMVOCs emphasizes that fewer congested conditions and less heavy machinery lead to diminished VOC releases. VOCs, reacting with NOx in sunlight, form tropospheric ozone. Thus, the improved traffic flow and shorter repair durations in terms of the proposed method ensure fewer VOC precursors in the atmosphere, reducing the rate of ozone formation and thereby lessening related respiratory risks.

The LCIA conducted using the TRACI, ReCiPe, and ILCD methodologies confirms the significant environmental advantages of the proposed spall material quantity takeoff repair method compared to traditional approaches. However, it is essential to acknowledge the potential limitations related to data quality and methodological constraints. The choice of databases and impact assessment methods (TRACI, ReCiPe, ILCD) may introduce regional biases or varying characterization factors. Factors such as diesel consumption rates, PCC production volumes, emission factors, regional fuel quality, cement kiln technology, quarrying practices, and traffic conditions can influence the exact magnitude of environmental improvements. For instance, the use of lower-sulfur fuels or advanced NOx reduction systems in vehicles and cement kilns could yield even greater relative reductions, while regions with less efficient technologies might observe smaller gains. Additionally, tools like OpenLCA are not sufficiently equipped to evaluate economic or social aspects, resulting in an analysis that focuses solely on environmental impacts. Comprehensive economic assessments would require detailed data on costs, including implementation, operation, and maintenance costs. Similarly, social evaluations would need to address factors such as labor conditions and health impacts on communities. Furthermore, although efforts have been made to minimize the bias in the data collected, the representativeness of the data may be affected by regional differences in fuel quality, cement kiln technology, quarrying practices, and traffic conditions. Since LCA assumptions may contain hidden biases or errors, research to reduce the data bias using methods such as some random sampling methods is necessary [49]. These limitations suggest that additional research studies or tools are necessary to comprehensively evaluate the proposed method’s overall effectiveness. Nevertheless, the LCA results affirm the significant potential of the proposed method in enhancing the environmental sustainability of road maintenance practices.

Given these advantages, the proposed method is particularly well suited for urban road maintenance, where traffic congestion and air pollution are often significant concerns. The ability to perform repairs with minimal disruption to the traffic flow can offer a more sustainable and environmentally friendly alternative to existing repair methods, especially in densely populated areas with high traffic volumes.

In conclusion, this study confirms that the proposed spall detection and repair method offers significant environmental advantages in terms of road maintenance, particularly by reducing traffic congestion, lowering fuel consumption, and decreasing global warming. Such improvements directly contribute to air quality enhancement, aligning with broader sustainability objectives and improving public health in urban settings. While these environmental gains are substantial, future research should also incorporate life cycle cost analysis (LCCA) to uncover potential economic benefits, including reduced labor costs, lower energy use, and minimized material wastage. Additionally, investigating social factors such as enhanced user experience due to fewer delays and potential decreases in accident rates would provide a more comprehensive understanding of the method’s overall value. By integrating environmental, economic, and social assessments, policymakers and transportation authorities can make more informed decisions. This integrated perspective can guide incentive structures, regulatory frameworks, and strategic funding toward more sustainable and resilient road maintenance practices. Ultimately, the proposed method emerges not only as an environmentally responsible alternative but also as a catalyst for long-term efficiency, cost-effectiveness, and social well-being in infrastructure management. The findings underscore the potential for more sustainable, efficient, and environmentally responsible road repair practices in the future.

4. Conclusions

This study demonstrates that the proposed spall detection and repair method, by optimizing material quantity takeoff and minimizing road closure times, offers substantial environmental advantages over traditional approaches. By integrating machine vision technology and advanced LCA methodologies, this research highlights the potential for significant improvements in environmental sustainability.

The key environmental benefits are as follows:

• Global warming potential: The proposed method reduced CO₂ emissions by 79.28%, primarily through shorter road closures and reduced vehicle idling. This reduction is critical for mitigating climate change impacts, particularly in urban areas with heavy traffic.
• Air pollution: Significant decreases in particulate matter (79.21%), smog-related emissions (79.44%), and acidification potential (79.53%) were achieved. These reductions enhance air quality and contribute to public health improvements in densely populated regions.
• Resource optimization: Accurate material estimation resulted in a 79.94% reduction in ecotoxicity and a 20.27% reduction in water depletion, underscoring the method’s efficiency in resource utilization.
• Traffic and operational efficiency: The elimination of road closures during the detection phases led to improved traffic flow, reducing congestion-related emissions and enhancing the overall operational efficiency.

These results demonstrate that the proposed method aligns with broader sustainability goals, addressing not only environmental concerns but also operational inefficiencies inherent in traditional spall repair processes. By combining high-precision spall detection with streamlined repair operations, this approach represents a transformative step toward more sustainable infrastructure management.

While the environmental benefits are clear, implementing this method may require initial investments in equipment, staff training, and integration with existing maintenance protocols. However, these investments are offset by long-term gains in operational efficiency, reduced material waste, and enhanced sustainability outcomes. To build on these findings, future studies should consider the following:

• Incorporate LCCA to explore the economic benefits, such as reduced labor costs and minimized energy use.
• Investigate the method’s applicability across diverse regions with varying traffic densities, climate conditions, and material availability.
• Evaluate social factors, including improved user experiences due to fewer delays and enhanced roadway safety.

In conclusion, the proposed spall detection and repair method emerges as a robust, environmentally responsible alternative to traditional practices. By aligning with sustainability objectives and offering practical benefits for infrastructure management, this approach has the potential to significantly advance the field of road maintenance, contributing to safer, more efficient, and more environmentally sustainable infrastructure systems.