Enhancing environmental sustainability and operational efficiency in a case study of limestone quarry in an arid climate

Abstract

This study explored innovative approaches to enhance environmental sustainability and operational efficiency in quarry operations in Ethiopia, focusing on the limestone quarry of the National Cement Factory. This open-air quarry, dedicated to produce gravel and aggregate for construction materials, presented significant environmental challenges, including air pollution, dust emissions, and noise pollution. The study examined the integration of automated environmental monitoring systems, dust and noise control techniques, and sustainable site rehabilitation. Automated environmental monitoring systems, incorporating IoT sensors and machine-learning algorithms, provide real-time data on air quality, dust levels, and noise pollution. This data-driven approach enables predictive analytics for early hazard detection and proactive mitigation. In parallel, eco-friendly practices such as non-toxic dust suppressants and sustainable land reclamation strategies mitigate the ecological impacts of quarrying. These strategies, tailored to arid environments, include the use of indigenous plant species to restore biodiversity and prevent soil erosion. This study employs a case study methodology, collecting and analyzing real-time data before and after interventions. The findings demonstrate significant improvements, including a 45% reduction in dust levels, a 15% reduction in noise levels, and a 266% increase in vegetation cover. Additionally, the results highlighted the role of local regulations, industry collaboration, and workforce training in achieving sustainable quarrying operations. The proposed approach enhanced environmental compliance, safety, and operational efficiency, offering a model for similar quarrying sites in arid climates.

Introduction

Mining operations play an important role in the economies of mineral-rich countries, contributing significantly to Gross Domestic Product (GDP) and creating job opportunities^1–3. However, the mining process frequently creates significant environmental and operational challenges⁴. The arid climate, coupled with the region’s unique geological characteristics, such as its loose soil composition, flat topography, and high mineral content, exacerbates issues related to air quality, dust levels, and noise pollution. These factors contribute to higher levels of air pollution, dust, and noise, complicating efforts to meet environmental standards.

The lack of vegetation and hard ground surfaces in this arid environment exacerbate noise pollution, extending sound’s reach and increasing the overall environmental impact of quarrying activities. For example, the limestone quarry at the National Cement Factory in Eastern Ethiopia faces significant challenges due to the arid climate and lack of vegetation. Addressing these challenges requires novel approaches that combine advanced technologies with sustainable practices tailored to the region’s specific conditions^5,6.

In recent years, there has been an increased emphasis on sustainable mining practices around the world^7,8. This trend is driven by stricter environmental regulations, increased public awareness, and the mining industry’s commitment to reducing its environmental footprint⁹.

Automated environmental monitoring systems, which use Internet of Things (IoT) sensors and machine learning algorithms, provide real-time information about critical environmental parameters^10–12. These systems enable continuous monitoring of air quality, dust levels, and noise pollution, providing actionable insights that allow for faster detection and mitigation of environmental hazards¹³. The application of advanced monitoring technologies improves environmental compliance and operational efficiency by allowing for early hazard detection and proactive mitigation¹⁴. While previous research has explored various aspects of sustainable mining and environmental monitoring, several critical gaps remain unaddressed. Despite growing efforts toward sustainable quarrying, the integration of real-time environmental monitoring with targeted dust and noise control strategies remains underexplored, particularly in arid environments such as Saudi Arabia and Ethiopia. Many quarries continue to rely on periodic manual monitoring, which delays hazard detection and response. Additionally, while green practices such as sustainable land reclamation have been implemented in various regions, their effectiveness in extreme climates remains insufficiently studied. This study aims to bridge this gap by evaluating the effectiveness of IoT-based environmental monitoring systems and sustainable site rehabilitation in reducing dust, noise pollution, and ecological degradation in quarry operations.

In addition to advanced monitoring, green practices such as the use of non-toxic dust suppressants and sustainable land reclamation techniques are critical for reducing the environmental impact of quarrying^15–18. In arid regions such as Eastern Ethiopia, these practices include drought-resistant vegetation, water-efficient irrigation, and surface treatments, all of which are critical for restoring biodiversity and preventing soil erosion following quarrying activities. These methods help to stabilize the soil by establishing deep-rooted vegetation and using windbreaks, which address both wind and potential water erosion¹⁹.

By adopting these practices, the quarrying industry in Ethiopia can align itself with global sustainability standards and enhance its environmental stewardship. Green mining practices, including the use of eco-friendly materials and sustainable land reclamation techniques, reduce harmful environmental impacts while promoting biodiversity and ecosystem health^20–24. Recent advancements in machine learning algorithms have further improved the utility of monitoring systems, enabling predictive analytics that allows for early detection of potential environmental hazards and proactive mitigation measures^25–27.

This study focuses on the application of innovative technology and sustainable practices in the limestone quarry of the National Cement Factory in Eastern Ethiopia. This study evaluates the impact of integrating IoT-based environmental monitoring systems with sustainable quarrying practices to address the challenges posed by quarry operations in an arid environment. Unlike previous studies that focus on either monitoring or sustainability independently, this research explores their combined effect on dust reduction, noise pollution control, and ecosystem rehabilitation. Through a case study at the limestone quarry of the National Cement Factory in Ethiopia, this work assesses the practical implications of real-time data analytics in improving environmental compliance and operational efficiency. The incorporation of green practices, particularly in arid environments such as Saudi Arabia and Ethiopia, illustrates their profound impact on restoring biodiversity, supporting native species, and increasing ecosystem resilience.

This study proposed novel perspectives on enhancing compliance, safety, and ecological restoration in arid mining environments by combining real-time environmental monitoring with sustainable quarrying methods. The remainder of the paper is structured as follows: Sect."Literature review"provides a review of existing literature on environmental monitoring systems and sustainable quarrying practices. Section"Methodology"presents the methodology, including case study selection, sensor deployment, and data processing techniques. Section"Results"presents and analyzes the results, focusing on the effectiveness of dust and noise control measures, predictive analytics, and land reclamation practices. Section"Discussion"concludes with key findings, limitations, and recommendations for future research.

Literature review

Environmental impacts of quarrying

The environmental impacts of quarrying activities, particularly in arid and semi-arid regions, have been investigated^28–30. Studies highlight the effects of dust generation, noise pollution, and soil erosion associated with mining operations, emphasizing the need for effective mitigation strategies^31–33.

Gazi et al.³⁴ presented a method for evaluating energy and environmental concerns associated with small-to-medium enterprises involved in marble quarrying and processing. They quantified total energy consumption for particular products and processes, alongside relevant environmental indicators. Their findings indicated that the proposed model effectively enhances energy efficiency in plants and measures environmental impacts.

Mensah et al.³⁵ explored the effects of stone quarrying on sustainable livelihoods and environmental quality in Ghana. Their study identified major environmental consequences such as vibrations, noise, building cracks, dust in air and water, land degradation, and wildlife migration. Given these diverse impacts, the authors recommended that stakeholders adopt more environmentally friendly quarrying practices and improve socio-economic justice. However, while these studies highlight the environmental damage caused by quarrying, they primarily focus on general mining activities rather than the unique challenges faced in arid regions, where dust control and ecological restoration require specialized approaches.

Monitoring technologies in quarry operations

Traditional environmental monitoring methods rely on periodic manual sampling and analysis. For example, air quality is often monitored by collecting samples with physical filters, which are then analyzed in a laboratory. Noise pollution might be measured at regular intervals using portable sound level meters. While these methods provide valuable data, they are limited by their sporadic nature, as they only capture environmental conditions at specific times, leading to potential gaps in monitoring³⁶. Automated systems incorporating IoT sensors and machine learning algorithms enable continuous tracking of air quality, dust levels, and noise pollution^37,38. However, traditional monitoring approaches remain prevalent in many quarrying regions due to high implementation costs and limited technical expertise^39,40. In contrast, advanced monitoring systems rely on automated IoT sensors. These sensors continuously monitor various environmental parameters, such as air quality, dust levels, noise pollution, and water contamination, in real time. Data from these sensors is transmitted to centralized systems where it is analyzed instantly. If any parameter exceeds predefined thresholds, immediate alerts are generated, allowing for prompt corrective actions⁴¹.

Wang et al.⁴² discussed the geological assurance theory and techniques for green and sustainable coal mining. Their study emphasized the importance of conducting thorough environmental assessments before, during, and after quarrying operations to minimize environmental damage. Brodny and Tutak⁴³ explored the application of artificial neural networks (ANNs) in analyzing greenhouse gas and air pollutant emissions from the mining sector in the European Union. Their findings suggested that ANN models provide highly accurate predictions for emission levels and enable proactive mitigation measures.

Sustainable quarrying and rehabilitation practices

Sustainable quarrying requires the implementation of environmentally friendly dust suppression and land rehabilitation techniques. Quarrying in arid regions presents unique challenges due to low precipitation, limited vegetation, and increased soil erosion risk. Sustainable practices such as revegetation, the use of non-toxic dust suppressants, and efficient water management are essential for mitigating environmental degradation.

Worku⁴⁴ evaluated the environmental and socio-economic effects of quarry mining in Addis Ababa, Ethiopia, and proposed mitigation strategies. The study identified pre-planning quarrying operations, adopting resource-efficient mining practices, and improving stakeholder engagement as key measures to minimize environmental damage.

Lèbre et al.⁴⁵ analyzed sustainable practices in managing solid waste in urban areas, emphasizing the importance of community involvement and technology integration. Their study highlighted the role of recycling and proper waste disposal in reducing environmental harm. Although these strategies improve environmental restoration, their effectiveness in arid environments remains unclear. Given the extreme conditions of quarry sites with minimal rainfall and high soil degradation, the long-term sustainability of these practices requires further validation through data-driven monitoring.

Integration of monitoring and sustainability practices

Kariri et al.⁴⁶ proposed a novel digital twin framework for optimizing mining operations. Their framework integrates real-time data, machine learning, and predictive analytics to enhance decision-making and sustainability in mining environments.

Despite significant advancements in environmental monitoring and sustainable quarrying practices, few studies have examined their combined effects in arid environments. Existing research on quarrying impacts has largely focused on general environmental consequences, while technological solutions such as IoT-based monitoring have been studied independently from sustainable rehabilitation strategies. This fragmentation creates a research gap in understanding how real-time environmental monitoring can enhance the effectiveness of sustainable quarrying techniques. This study seeks to bridge this gap by integrating predictive analytics, real-time monitoring, and eco-friendly land rehabilitation in the limestone quarry of the National Cement Factory in Ethiopia. By addressing the limitations of previous studies, this research provides a comprehensive framework for improving environmental compliance, operational efficiency, and ecosystem restoration in quarrying operations.

Methodology

This study employed a hybrid approach, combining both qualitative and quantitative methodologies. A case study methodology was chosen to enable an in-depth examination of the quarrying operations at the National Cement Factory in Eastern Ethiopia. The research integrated experimental design principles by deploying IoT-based monitoring systems and evaluating the impact of dust and noise control interventions through comparative analysis of pre- and post-intervention data. Additionally, machine learning models were used to predict environmental parameters, enhancing decision-making in quarry management. The quantitative component focused on analyzing sensor data, while qualitative insights were gathered through observational assessments of rehabilitation efforts.

Case study (Limestone quarry of the National cement factory in Eastern Ethiopia)

The limestone quarry at the National Cement Factory in Eastern Ethiopia has been chosen as the study area. This quarry is located at 9°34′18″N latitude and 41°51′21″E longitude, approximately 515 km east of Ethiopia’s capital, Addis Ababa. The quarry is located in the Oromia region, near the town of Dire Dawa, one of the country’s important industrial hubs (Fig. 1).

Fig. 1.

Location and production operations of limestone quarry of the National Cement Factory, Ethiopia (Map data: Google, Airbus).

The area has an arid climate, with temperatures frequently exceeding 30 °C during the hottest months. Annual rainfall is sparse, averaging around 500 mm, with the majority falling during brief seasonal rains. The quarry is located on a limestone-rich plateau, with relatively rugged topography surrounding it. The geological composition is primarily composed of large limestone deposits that are actively mined for cement production. The loose and sandy soil in this region presents significant challenges for dust management during quarrying operations. Furthermore, the scarcity of vegetation caused by arid conditions exacerbates dust and noise pollution, as natural barriers to mitigating these environmental impacts are minimal. The environmental challenges associated with these quarrying activities, including dust generation, noise pollution, and potential impacts on local biodiversity, have necessitated the implementation of innovative environmental management practices⁴⁷.

This region, having witnessed increasing industrial activity, faces growing concerns about the sustainability of its environmental conditions. The lack of natural water bodies and reliance on limited groundwater resources add to the complexity of managing environmental impacts in this area. Moreover, the region is prone to soil erosion, driven by both wind and occasional water runoff during the brief rainy season, complicating efforts to stabilize the land post-quarrying. Given these challenges, the limestone quarry at the National Cement Factory in Eastern Ethiopia serves as an ideal case study for examining the effectiveness of advanced environmental monitoring systems and sustainable quarrying practices in an arid environment.

Implementation of monitoring systems

Recently, environmental monitoring in mining operations utilizes a combination of traditional methods and advanced technologies. The transition from manual to automated systems has been driven by the need for more accurate, real-time data that can ensure compliance with increasingly stringent environmental regulations^48–50.

To ensure accurate and continuous environmental monitoring, this study deployed a network of IoT-based sensors at strategic locations within the limestone quarry of the National Cement Factory in Eastern Ethiopia. The monitoring system was designed to capture real-time data on air quality, dust concentrations, and noise pollution, enabling immediate corrective actions and long-term environmental compliance.

Sensor placement was determined based on a comprehensive site survey that analyzed prevailing wind patterns, operational hotspots, and historical pollution data. The sensors were installed at critical locations, including crusher zones, loading areas, and haul roads, to maximize spatial coverage. The data collected was wirelessly transmitted to a central processing unit, where machine learning models analyzed trends and generated predictive alerts for hazardous conditions.

Installation and operation of IoT sensors for real-time data collection

The installation and operation of IoT sensors were carried out at the limestone quarry of the National Cement Factory to monitor air quality, dust levels, and noise pollution in real-time. The process involved several stages, from planning and sensor selection to deployment and data integration (Table 1).

Table 1.

IoT sensors installation and operation details.

Sensor type	Parameter monitored	Location	Power source	Data transmission	Calibration frequency	Maintenance activities
PM2.5 and PM10 sensors	Particulate matter	Near crushers and storage	Solar panels	Wireless network	Monthly	Cleaning, recalibration, replacement if faulty
Gas sensors	NO2, CO, SO2	Loading and transportation areas	Backup batteries	Wireless network	Quarterly	Routine inspections, recalibration
Noise level sensors	Noise pollution	Throughout complex	Solar panels	Wireless network	Monthly	Calibration, data integrity checks, replacement if necessary

• Planning and Sensor Selection: Initially, a comprehensive site survey was conducted to identify optimal locations for sensor placement, determined by analyzing the site’s geography, prevailing wind patterns, and historical environmental data. A total of 15 sensors were deployed across the site, with placements concentrated in areas of highest dust and noise generation, such as near crushers, loading zones, and storage areas. The sensors were spaced approximately (150–200) meters apart to ensure comprehensive and overlapping coverage, thereby optimizing the effectiveness of the monitoring network. Areas with the highest levels of dust and noise generation, such as near the crushers, loading zones, and storage areas, were prioritized for sensor placement. The sensors were installed at a height of approximately 1.5 to 3 m for dust and air quality monitoring, and 1.5 to 4 m for noise monitoring, ensuring accurate capture of relevant environmental data. Based on the environmental parameters to be monitored, appropriate IoT sensors were selected. These included particulate matter (PM) sensors, gas sensors, and noise level sensors, which not only record noise levels in dB but also monitor for example when noise exceeds predefined thresholds.

• Installation Procedure: The selected sensors were installed at the predetermined locations. The installation process involved mounting the sensors on sturdy structures to ensure stability and prevent interference from external factors. Each sensor was connected to a central data collection unit via a secure wireless network. Power sources, such as solar panels and backup batteries, were provided to ensure uninterrupted operation.

• Data Collection and Integration: Once installed, the sensors were calibrated to ensure accurate measurements. Calibration was performed using standardized reference materials and protocols. Calibration was performed using standardized reference materials, including NIST-traceable particulate matter standards (e.g., SRM 1648a) for dust sensors, certified gas mixtures for gas sensors, and acoustic calibrators for noise sensors. These materials and devices ensure that the sensors provide accurate and reliable measurements according to internationally recognized standards. The sensors continuously collected data on air quality, dust levels, and noise pollution, which were transmitted in real-time to a central server. Advanced machine learning algorithms were applied to process and analyze the collected data, enabling predictive analytics for early hazard detection and proactive mitigation.

• Monitoring and Maintenance: Regular maintenance checks were conducted to ensure the sensors’ optimal performance. This included routine inspections, cleaning, and recalibration. Any faulty sensors were promptly detected and replaced to maintain data integrity, utilizing automated diagnostic checks, real-time data consistency analysis, and cross-verification with adjacent sensors. Additionally, routine calibration and maintenance schedules provided early feedback on sensor performance, allowing for proactive interventions before complete sensor failure occurred. The real-time data were monitored using a custom-built dashboard, providing actionable insights for environmental management.

The IoT sensors enables quarry operators to continuously monitor environmental conditions, leading to faster responses to hazardous air quality and excessive noise pollution. These sensors help reduce equipment downtime, optimize dust suppression strategies, and ensure compliance with occupational safety standards. The use of predictive analytics based on sensor data also enhances decision-making by forecasting environmental risks, allowing for proactive mitigation strategies.

The comprehensive deployment of IoT sensors facilitated real-time monitoring and data collection, providing critical insights into environmental conditions at the limestone quarry of the National Cement Factory. For example, when the sensors detected a significant rise in dust levels near the crushers, real-time alerts enabled immediate interventions such as enhanced dust suppression and operational adjustments. These timely actions not only ensured compliance with environmental regulations but also contributed to the site’s operational efficiency and environmental sustainability.

Data processing methods using machine learning algorithms

In this methodology section, IoT sensors were strategically deployed throughout the limestone quarry of the National Cement Factory to continuously collect environmental data. This data underwent rigorous preprocessing, including cleaning, normalization, and feature extraction, to ensure its quality and relevance for subsequent analysis. Machine learning algorithms, including ANN was employed to train the model on historical data⁵¹. This model utilized features such as operational activity levels and meteorological conditions to predict environmental parameters like air quality, dust levels, and noise pollution. Cross-validation was utilized to validate these models and prevent overfitting, ensuring their robustness and generalizability. The outcome was a sophisticated predictive analytics framework that enabled early hazard detection and proactive mitigation strategies based on real-time environmental insights. Visualizations and alerts were generated to facilitate intuitive monitoring and reporting, empowering stakeholders with actionable information for sustainable quarry operations (Table 2).

Table 2.

Data processing workflow and techniques.

Step	Description	Outcome
Data collection and storage	Continuous collection and storage of sensor data.	Large dataset of real-time environmental data.
Data preprocessing	Cleaning, normalizing, imputing missing values, and handling outliers.	High-quality dataset ready for analysis.
Feature extraction	Extraction of key environmental and temporal features from preprocessed data.	Relevant features for machine learning analysis.
Model training and validation	Training and validating machine learning models using cross-validation.	Robust and generalizable models.
Predictive analytics	Applying trained models to predict future environmental conditions.	Early detection of potential environmental hazards.
Visualization and reporting	Creating dashboards and alerts for real-time monitoring and reporting.	Intuitive and actionable insights for stakeholders.

The dataset used for machine learning model training was derived from continuous monitoring over a 12-month period, during which 15 strategically placed IoT sensors collected real-time environmental data. Each sensor recorded data at predefined intervals, resulting in a total of approximately 129,600 raw readings across all sensors. Following preprocessing steps such as noise removal, outlier detection, and feature selection, a representative subset of 50,000 sensor readings was extracted for model training and testing. The dataset was split into 80% for training and 20% for testing, with model performance evaluated using normalized root mean square error (NRMSE) as the primary accuracy metric.

This section details the methodological approach used for processing sensor data with machine learning algorithms, ensuring a thorough and systematic analysis of environmental conditions at the limestone quarry of the National Cement Factory.

The structure of an ANN consists of interconnected layers of nodes or neurons, typically including an input layer, one or more hidden layers, and an output layer. Each connection between neurons has an associated weight that is adjusted during the learning process to minimize prediction errors. The learning process involves feeding data into the network, processing it through the hidden layers, and comparing the output against known results. The network iteratively adjusts the weights to improve its accuracy over time^52–54.

1

.

In the Eq. 1, to show the applied weight on hidden layer w_nm has been used that connects the mth neuron of input layer to the n^th neuron of hidden layer and also w_hn represents the weight in output layer which connects the n^th neuron of hidden layer to the h^th neuron of output layer. These weights are different and in network training process can change. Then, w_na and w_ha depict the bias for nth neuron of hidden layer and h^th neuron of output neuron respectively. In addition, x_m shows the m^th input variable for input layer. Also, f_p stands for the activation function in hidden neuron and f_a displays that in output neuron. M_m is the number of input neurons and N_n is for hidden neurons.

In this study, ANN models were utilized to analyze real-time data collected from IoT sensors monitoring various environmental parameters such as air quality, dust levels, noise pollution, and water contamination. The ANN was trained on historical data to recognize patterns and predict high-risk periods for environmental hazards. This enabled proactive measures to be taken before conditions became critical, thus improving the overall safety and efficiency of the mining operations.

Dust and noise control measures

In response to the complexities of managing both dust and noise emissions during quarrying operations at the limestone quarry of the National Cement Factory, a variety of tailored techniques were implemented. Although dust and noise present distinct challenges, The proposed approach integrates their management to achieve a comprehensive environmental strategy, leveraging the benefits of addressing these issues simultaneously (Table 3).

Table 3.

Dust and noise control measures at limestone quarry of the National cement factory.

Measure	Description	Numerical report
Water sprinkling	Controlled application of water to suppress dust emissions.	Average reduction of airborne particulate matter by 45%.
Enclosures and barriers	Structural installations to contain noise emissions and reduce propagation.	Average noise reduction by 30 dB.
Non-toxic dust suppressants	Application of environmentally safe chemicals to prevent dust from becoming airborne.	Average dust emission reduction by 50%.
Operational optimization	Adjustments in operational practices to minimize dust and noise generation.	Optimization resulting in improved efficiency.

Dust control measures

To minimize dust emissions, several methods were employed. Water sprinkling systems were strategically installed across operational areas, achieving an average reduction of airborne particulate matter by 45%. Additionally, non-toxic dust suppressants, including Calcium Magnesium Acetate (CMA) and synthetic organic polymers, were applied using spraying equipment to prevent dust from becoming airborne. These suppressants, selected for their environmental safety and effectiveness, contributed to a reduction in dust emissions by an average of 50%. The use of operational practices such as adjusting equipment speeds and scheduling also played a role in further reducing dust generation.

At the limestone quarry of the National Cement Factory, non-toxic dust suppressants were carefully chosen to minimize environmental impact. For example, CMA was applied due to its hygroscopic properties, effectively reducing dust emissions. Synthetic organic polymers, known for their adhesive qualities, formed a durable film over surfaces, preventing fine particles from dispersing into the air. Vegetable oils, selected for their viscosity and tackiness, were also applied to surfaces prone to dust emissions, further enhancing the site’s sustainability. Non-toxic dust suppressants were applied using sprayers and additives to minimize airborne dust particles. These suppressants work by binding fine particles together, preventing them from becoming airborne during operations. Operational practices were enhanced to minimize dust generation, including optimized blasting techniques, controlled loading and transportation of materials, and covering stockpiles to reduce wind erosion.

Noise control measures

To address noise emissions, structural enclosures and barriers were erected around noisy equipment and zones, limiting noise propagation and achieving average noise reductions of 30 dB. These physical barriers were complemented by operational practices designed to reduce noise generation, including the optimization of equipment speeds and operational schedules. While noise control measures differ in their application from dust control, the integrated approach at the limestone quarry of the National Cement Factory highlights the effectiveness of a comprehensive strategy that addresses multiple environmental challenges concurrently.

To quantify the impact of dust and noise control interventions on production efficiency, key operational metrics were collected before and after the implementation of these measures. The data included equipment downtime due to dust-related failures, worker productivity losses associated with noise disturbances, and overall production throughput. Operational logs from the quarry management system were analyzed over a 12-month period, comparing the frequency of maintenance-related delays, interruptions in material transport, and worker-reported disruptions. The downtime hours due to equipment clogging and noise-related workflow inefficiencies were recorded and statistically compared to assess the improvements following intervention. These measures and suppressants successfully mitigated environmental impacts associated with dust and noise at the quarry site, aligning with sustainability goals and improving overall operational conditions.

Site rehabilitation practices

Several innovative land reclamation techniques were employed at the limestone quarry of the National Cement Factory to restore disturbed areas and promote ecological sustainability. Unlike traditional methods, which often involved only basic leveling and minimal vegetation efforts, these techniques introduced significant changes to standard operating procedures. The reclamation process at the complex included advanced soil stabilization using biodegradable geotextiles, the strategic planting of drought-resistant native vegetation, and integrated water management practices such as the creation of micro-catchments to maximize water retention in the arid environment. These methods not only prevented soil erosion, primarily driven by wind in the region, but also improved soil fertility and supported long-term biodiversity restoration. Continuous monitoring and adaptive management were also implemented, allowing for adjustments to be made as needed, a step beyond the traditional reactive approaches. These comprehensive strategies have significantly enhanced the effectiveness of land reclamation at the site, aligning with ecological sustainability goals and setting a new standard for quarry operations in similar environments.

Excavated pits and disturbed lands were graded and contoured to facilitate natural drainage and prevent erosion. Soil stabilization measures, including the application of biodegradable erosion control mats and blankets, were implemented to prevent soil loss and promote vegetation establishment. Additionally, strategic placement of mulch and compost helped improve soil fertility and moisture retention, supporting plant growth and biodiversity restoration.

The rehabilitation process at the limestone quarry of the National Cement Factory focused on selecting and planting indigenous plant species well-adapted to the arid environment. Native shrubs, grasses, and trees were carefully chosen based on their ability to thrive in local conditions and contribute to biodiversity restoration. Planting methods included direct seeding, transplanting seedlings, and installing mature plants, ensuring diverse vegetation cover across the rehabilitated areas. Monitoring programs were established to assess the survival and growth of planted species, with periodic evaluations conducted to track ecosystem recovery and soil stability (Table 4).

Table 4.

Site rehabilitation practices at limestone quarry of the National cement factory.

Practice	Description	Numerical report
Grading and contouring	Adjusting terrain to facilitate drainage and prevent erosion.	Effective soil stabilization and drainage.
Soil stabilization techniques	Application of biodegradable mats and blankets to control erosion.	Significant reduction in soil loss.
Mulching and composting	Application of organic materials to enhance soil fertility and moisture retention.	Improved soil quality and plant growth.
Selection of indigenous plant species	Choosing native species adapted to local arid conditions for rehabilitation.	Successful establishment of diverse flora.
Planting methods	Direct seeding, transplanting, and installation of mature plants for vegetation restoration.	High survival rates and growth of species.

During the rehabilitation process, several challenges were encountered. The harsh arid conditions of the region posed significant difficulties in establishing vegetation, with high temperatures and low rainfall limiting plant survival. The selection of indigenous plant species required extensive research to ensure the chosen species were not only well-adapted to the environment but also resilient enough to survive the initial critical stages of growth. Additionally, soil compaction in some areas, due to prolonged quarrying activities, impeded root penetration and water infiltration, necessitating extra soil conditioning and aeration efforts. Despite these challenges, the implementation of targeted soil stabilization and planting techniques, combined with ongoing monitoring and adaptive management, enabled successful vegetation establishment and ecosystem recovery.

Efficiency criterion

In this study, the NRMSE is employed as a key efficiency criterion to evaluate the performance of predictive models. NRMSE provides a normalized measure of the differences between predicted and observed values, making it a robust indicator of model accuracy across different scales of data. The NRMSE is calculated by taking the square root of the mean squared error (MSE) between predicted () and observed (y) values, and then normalizing it by the range of the observed data (y_max - y_min). The formula for NRMSE is (Eq. 2):

2

where n is the number of observations. This metric is particularly useful in this study as it allows for a standardized comparison of prediction errors, ensuring that the results are not biased by the scale of the data. By using NRMSE, It could be effectively assessed the predictive accuracy of these models in monitoring and controlling various environmental parameters within the mining operations⁵⁵.

NRMSE is between 0 and 1 when it is normalized by the range of the observed data. This normalization ensures that the error metric is dimensionless and allows for comparison across different scales and units. A NRMSE value of 0 indicates a perfect fit, where the predicted values match the observed values exactly. As the NRMSE value approaches 1, it indicates increasing discrepancies between the predicted and observed values, with 1 suggesting that the average prediction error is equal to the range of the observed data. Values greater than 1 are possible, but they would indicate very poor model performance, with errors larger than the range of the data⁵⁵.

Results

Environmental monitoring data

Analysis of air quality, dust levels, and noise pollution before and after intervention

Figure 2 presents the variations in particulate matter (PM2.5 and PM10) and nitrogen dioxide (NO₂) concentrations before and after blasting at the limestone quarry of the National Cement Factory. In the pre-blasting phase (Fig. 2a), pollutant concentrations fluctuated irregularly, with PM10 displaying the highest values, followed by PM2.5 and NO2. These variations likely resulted from localized dust disturbances caused by vehicle movement, material handling, and atmospheric conditions influencing particle dispersion. After blasting (Fig. 2b), a pronounced increase in pollutant concentrations was observed, with PM10 and PM2.5 levels rising sharply. Unlike the pre-blasting phase, the data exhibited a more structured and linear trend, indicating a steady and consistent release of pollutants into the atmosphere. The significant increase in PM levels after the blasting process is due to the high-energy fragmentation of limestone, which produced and spread fine particulate matter across a broader area. Additionally, the increase in NO₂ concentrations suggested the contribution of combustion-related emissions from explosives and operational machinery. The findings demonstrate the significant influence of blasting activities on air quality, highlighting the importance of continuous monitoring and effective mitigation strategies. The execution of dust suppression strategies, including precise water spraying and the use of non-toxic suppressants, was essential in managing the spread of particulate matter following blasting activities. Moreover, predictive modeling could optimize intervention timing to reduce airborne pollution and achieve regulatory air quality standards.

Fig. 2.

Growth of matter concentration in the limestone quarry of the National Cement Factory a before blasting (30 min), b after blasting (60 min).

Table 5 summarizes the concentrations of these pollutants before and after intervention. The monitoring of air quality parameters (PM10, PM2.5, and NO₂) showed consistent reductions after intervention. The results indicate significant reductions in PM10 (48%), PM2.5 (47%), and NO₂ (40%) levels following the implementation of dust control measures and improved operational practices. These improvements reflect the effectiveness of measures such as the application of non-toxic dust suppressants and enhanced ventilation systems in minimizing airborne pollutants.

Table 5.

Air quality parameters before and after intervention.

Parameter	Before intervention (µg/m³)	After intervention (µg/m³)	Variation (%)
PM10	75	39	− 48%
PM2.5	51	27	− 47%
NO2	20	12	− 40%

Table 6 presents the dust concentration levels measured at various monitoring points across the quarry site before and after intervention. The data reveals consistent reductions in dust concentrations across all monitored areas post-intervention. These reductions can be attributed to the effective implementation of dust suppression techniques and improved operational practices, which minimized dust generation during activities such as loading, crushing, and transportation.

Table 6.

Dust concentration levels before and after intervention.

Location	Before intervention (mg/m³)	After intervention (mg/m³)	Variation (%)
Loading area	0.16	0.11	− 31.3%
Crushing area	0.23	0.13	− 43.8%
Haul roads	0.19	0.10	− 47.3%

Noise levels were measured using decibel (dB) meters positioned strategically around the quarry site. Table 7 summarizes the noise levels recorded before and after the implementation of noise control measures.

Table 7.

Noise levels before and after intervention.

Location	Before intervention (dB)	After intervention (dB)	Variation (%)
Crushing area	89	72	− 19.1%
Loading area	84	70	− 16.6%
Perimeter	74	62	− 16.2%

Figure 3 compares limestone quarry noise levels before and after noise mitigation at various locations. In Fig. 3a, noise measurements were taken across multiple operational zones. Before intervention, noise levels in these areas were significantly high, with several locations exceeding 85 dB. Following the implementation of noise reduction measures (the installation of sound barriers, optimized scheduling of machinery operations, and the use of noise-dampening enclosures) a noticeable decline in noise levels was observed. Significant reductions were observed in high-impact areas such as the mining checkpoint and batch plant. In Fig. 3b, noise levels before and after intervention were assessed in indoor facilities. Before intervention, noise levels in these areas remained elevated, primarily due to the transmission of vibrations from nearby heavy machinery and industrial processes. After the application of noise insulation materials and the reconfiguration of workspace layouts to minimize direct exposure to high-noise sources, a considerable reduction in decibel levels was achieved. The control room and process and mining office exhibited the highest noise reductions, underscoring the role of acoustic insulation and strategic placement of noise barriers in improving indoor work environments.

Fig. 3.

Comparison of noise levels in a workshop, b control rooms and logistic offices.

Presentation of real-time data and predictive analytics outcomes

Figure 4 presents a comparative analysis of collected and predicted environmental data using NRMSE as the evaluation metric. Figure 4a depicts the performance of the predictive model in forecasting air quality parameters. Figure 4b illustrates its accuracy in predicting noise levels.

Fig. 4.

Comparison of high-risk status of a air quality, and b noise levels through collected and predicted data.

ANN algorithm was employed to analyze the collected data and generate predictive models. These models forecasted trends in air quality deterioration, dust dispersion patterns, and noise level fluctuations based on historical data and current environmental conditions (Table 8).

Table 8.

Predictive analytics outcomes.

Predictive model	Parameter	Accuracy (%)	Predictive horizon
Air quality	PM10, PM2.5, NO2	81%	24 h
Dust dispersion	Dust Concentration	77%	12 h
Noise levels	dB	88%	6 h

Effectiveness of dust and noise control

This section examines the quantitative reductions achieved in dust and noise levels at the limestone quarry of the National Cement Factory following the implementation of targeted control measures. The focus is on presenting numerical data that illustrates the effectiveness of these measures in mitigating environmental impacts associated with quarry operations.

Prior to intervention, dust levels were monitored using PM10 and PM2.5 sensors located strategically across the quarry site. Following the implementation of dust suppression techniques, including the application of non-toxic dust suppressants and improved operational practices, significant reductions in dust concentrations were observed (Table 9). The results demonstrates that approximately 49% reduction in both PM10 and PM2.5 concentrations following the implementation of dust control measures.

Table 9.

Reductions in dust levels.

Dust parameter	Before intervention (µg/m³)	After intervention (µg/m³)	Variation (%)
PM10	167	86	− 48.5%
PM2.5	115	58	− 49.5%

Table 10 illustrates an 11.57% reduction in daytime noise levels and a 12.64% reduction in nighttime noise levels following the implementation of noise control measures.

Table 10.

Reductions in noise levels.

Noise parameter	Before intervention (dB)	After intervention (dB)	Variation (%)
Daytime	95	84	− 11.57%
Nighttime	87	76	− 12.64%

Figure 5 illustrates the effectiveness of dust and noise control interventions by comparing pollutant concentrations and noise levels before and after the implementation of mitigation measures. Figure 5a and b depict the temporal variations concentrations throughout a monthly cycle, while Fig. 5c presents the diurnal fluctuations in noise levels. In Fig. 5a, b and a significant reduction in PM2.5 and PM10 concentrations was observed following intervention. Prior to mitigation, a sharp spike in pollutant levels occurred at the beginning of the month, likely due to intensive quarrying activities such as blasting and material handling. However, after intervention, the peak concentration was substantially lower, and the overall dust levels remained consistently reduced throughout the monitoring period. The attenuation of airborne particulate matter can be attributed to the implementation of non-toxic dust suppressants, enhanced water spraying systems, and optimized material handling techniques, which effectively minimized dust dispersion in the quarry environment. As shown in Fig. 5c, Before intervention, noise levels exhibited a pronounced peak, reaching approximately 95 dB during peak operational hours (7:00–10:00 AM). After implementing noise reduction strategies, including the installation of noise barriers and adjustments in operational scheduling, the peak noise levels were notably reduced to approximately 88 dB.

Fig. 5.

Effect of (a and b) dust and c noise control in reduction.

In addition to environmental improvements, the impact of dust and noise control interventions on quarry production efficiency was analyzed. Table 11 presents a comparative assessment of equipment downtime and worker productivity losses before and after implementing control measures. The data indicates that after implementing dust and noise control measures, equipment downtime was reduced by 36%, primarily due to decreased dust accumulation in machinery components. Similarly, worker productivity losses due to noise disturbances dropped by 22%, attributed to a quieter work environment. Additionally, delays in material transport decreased by 38%, as improved air quality and reduced dust buildup enhanced machinery performance and operational continuity.

Table 11.

Impact of dust and noise control on production efficiency.

Parameter	Before intervention	After intervention	Improvement (%)
Equipment downtime (hours/week)	4.2	2.7	− 36%
Worker productivity loss due to noise (%)	22	17	− 22%
Material transport delays (minutes/day)	45	28	− 38%

Comparative analysis of control techniques for dust and noise reduction

Figure 6 presents an evaluation of the effectiveness of different control techniques implemented at the quarry and the performance of the ANN model in predicting environmental conditions. Figure 6a illustrates the percentage improvement achieved through various dust and noise mitigation measures. Figure 6b provides a comparison between collected environmental data and ANN-predicted values. In Fig. 6a, the effectiveness of four key mitigation strategies is compared. The application of non-toxic dust suppressants emerged as the most effective intervention, achieving an improvement of approximately 37%, demonstrating its significant role in reducing particulate matter dispersion. Figure 6b compares the normalized environmental data collected from sensors with ANN-predicted values.

Fig. 6.

Control techniques analysis a collected data, b ANN simulation.

Impact of rehabilitation practices

This section evaluating the effectiveness of these efforts in restoring biodiversity, preventing soil erosion, and improving ecological stability in the reclaimed areas. The soil in the reclaimed areas was amended with organic compost and other nutrients to enhance fertility and support the growth of native plant kinds. The soil was also leveled and contoured to prevent erosion and improve water retention. Native plant species, specifically selected for their adaptability to the arid environment, were planted. These species included drought-resistant grasses, shrubs, and trees that contribute to soil stabilization and biodiversity restoration. Efficient water management systems, including drip irrigation and rainwater harvesting, were installed to ensure the sustainable use of water resources for plant growth. The results of the land reclamation efforts are summarized in Table 12.

Table 12.

Outcomes of land reclamation efforts.

Parameter	Pre-reclamation	Post-reclamation	Improvement (%)
Vegetation cover (%)	12	44	266
Soil erosion (ton/ha/year)	10	6	66
Soil organic matter (%)	0.6	1.3	116
Soil moisture content (%)	10	18	80

To assess the percentage of vegetation cover accurately, the study employed quadrat sampling, a standard method in ecological research. Quadrats of 1 m² were randomly placed across different sections of the reclamation area. Within each quadrat, the percentage of ground covered by vegetation was estimated.

Figure 7 shows the effect of reclamation activities on vegetation cover over a one-year period, comparing pre-reclamation and post-reclamation conditions.

Fig. 7.

Effect of land reclamation in environment conservation.

Discussion

The results of noise levels before and after intervantion indicate notable reductions in noise levels across critical areas of the quarry site following the installation of noise barriers and operational adjustments. These measures effectively mitigated noise emissions associated with crushing and loading activities, contributing to a quieter working environment that aligns with established safety standards. Specifically, noise levels were reduced to below 85 dB (the occupational exposure threshold recommended by international health guidelines) in critical areas, a threshold considered safe for occupational exposure according to international health guidelines.

The improvements observed in air quality, dust levels, and noise pollution following the intervention underscore the efficacy of the implemented measures at the limestone quarry of the National Cement Factory. The reduction in PM10 and PM2.5 levels by 40% highlights the success of dust control strategies, including the use of non-toxic suppressants and enhanced ventilation systems. Similarly, the decrease in NO₂ concentrations by 25% reflects improved combustion processes and reduced emissions from equipment operation.

The substantial decrease in dust concentrations across loading, crushing, and haul road areas demonstrates the effectiveness of localized dust suppression techniques and operational adjustments. These measures not only minimized airborne particulates but also enhanced overall environmental conditions within the quarry site. The recorded reductions in noise levels by 14–17% validate the efficacy of noise control measures, including the installation of barriers and the strategic scheduling of activities to prevent multiple noise-generating operations from occurring simultaneously.

In this study, the ANN model was trained using historical sensor data and meteorological conditions to forecast air quality deterioration, dust dispersion patterns, and noise level fluctuations. The accuracy of these predictions was evaluated using the NRMSE metric, which ensured that the model provided reliable forecasts of environmental conditions. Specifically, the predictive model achieved 81% accuracy for air quality, 77% for dust dispersion, and 88% for noise levels, demonstrating its ability to provide actionable early warnings. Instead of reacting to pollution spikes, quarry operators optimized operational schedules by anticipating high-dust periods and scheduling dust suppression measures in advance; preemptively activating noise reduction barriers during peak operational hours to adjust mitigation strategies; reducing regulatory violations by preventing pollution levels from exceeding limits rather than using post-event adjustments.

The key advantage of predictive analysis lies in its ability to reduce reaction time and enhance decision-making efficiency, ensuring that environmental compliance is maintained continuously rather than through periodic assessments. The integration of predictive models with IoT-based real-time monitoring transforms environmental management from a reactive to a proactive approach, optimizing both sustainability and operational performance. These proactive measures helped mitigate potential risks and maintain compliance with air quality standards. The accuracy rates of 81% for air quality, 77% for dust dispersion, and 88% for noise levels underscored the reliability of the predictive models in supporting proactive environmental management strategies.

The utilization of real-time data presentation and predictive analytics at the limestone quarry of the National Cement Factory exemplifies a forward-thinking approach to environmental monitoring and management in quarry operations. The continuous monitoring of air quality, dust levels, and noise pollution enabled timely interventions such as the activation of additional air purifiers during high pollution events, the application of dust suppressants on construction sites, and the installation of sound barriers in areas of excessive noise. This proactive approach, which involved real-time data collection and immediate response, represents a significant departure from previous practices that relied on periodic checks and reactive measures.

The high accuracy of predictive models in forecasting environmental parameters reflects the effectiveness of machine learning algorithms in processing large datasets and identifying trends. NRMSE was used primarily as quality control metric, as it normalizes the RMSE by the range or mean of observed values, providing a scale-independent measure that is sensitive to large errors and allows for consistent comparison across different datasets. This makes NRMSE particularly suited for evaluating the performance of models in forecasting environmental parameters.

The integration of real-time data collection and predictive analytics has been pivotal in improving both environmental sustainability and operational efficiency at the limestone quarry of the National Cement Factory. By accurately forecasting environmental parameters and linking these predictions to specific quarry activities, the approach has enabled proactive interventions that reduce emissions, optimize operations, and ensure compliance with regulatory standards. These outcomes demonstrate the value of adopting advanced technologies in quarry operations to achieve continuous improvement in environmental management. This case study serves as a model for other mining and industrial sectors, highlighting the potential of predictive analytics to drive sustainable practices and operational excellence. While predictive analytics played a role in optimizing intervention timing, the primary focus of this study remains on quantifying the direct environmental impacts of the mitigation measures implemented at the limestone quarry. The measured reductions in PM10, PM2.5, and NO₂ concentrations, alongside significant declines in noise levels, provide concrete evidence of the effectiveness of physical interventions, independent of predictive capabilities. Specifically, the application of non-toxic dust suppressants (e.g., Calcium Magnesium Acetate and synthetic organic polymers) led to a 48% reduction in PM10 levels, a 47% decrease in PM2.5, and a 40% decline in NO₂ concentrations (Table 5). These reductions directly correlate with the targeted dust suppression strategies deployed near crushers, loading areas, and haul roads. Pre- and post-intervention assessments show that dust levels dropped by 31.3% in loading areas, 43.8% in crushing zones, and 47.3% along haul roads (Table 6). These improvements are linked to localized suppression measures such as adjusted blasting techniques, increased water sprinkling frequency, and optimized truck movement patterns. The installation of noise barriers and operational adjustments (e.g., restricted simultaneous heavy machinery operations) resulted in a 19.1% noise reduction in crushing zones, 16.6% in loading areas, and 16.2% at the quarry perimeter (Table 7). These changes significantly improved compliance with occupational noise exposure limits.

The results of land reclamation in environment conservation indicate a significant improvement in vegetation cover following the implementation of land restoration measures at the limestone quarry. Before reclamation vegetation cover remained consistently lower throughout the year, starting at approximately 18% in January and gradually increasing to around 68% by October. This increase likely reflects natural seasonal growth patterns but demonstrates limited vegetation recovery in the absence of active intervention. A slight decline is observed toward the end of the year, which may be attributed to seasonal variations such as reduced precipitation and lower temperatures affecting plant growth. Post-reclamation shows a notable increase in vegetation cover is observed, particularly from April to October, where the growth rate surpasses that of the pre-reclamation period. Although a slight decline is evident from October onwards, the overall vegetation cover remains higher than in pre-reclamation conditions, indicating improved resilience of the restored ecosystem.

The increase in vegetation cover (from 12 to 44%, a 266% improvement) was primarily driven by the application of soil stabilization techniques, selection of drought-resistant native plant species, and efficient water management strategies. These methods directly enhanced soil fertility and moisture retention, leading to higher plant survival rates and improved ecosystem resilience. The use of biodegradable geotextiles, compost mulching, and contouring of the land significantly reduced soil erosion by 66%, ensuring a more stable substrate for plant growth. While physical rehabilitation measures provided the foundation for vegetation regrowth, real-time monitoring and predictive analytics played a crucial role in optimizing dust suppression and improving air quality, indirectly supporting plant recovery. IoT-based sensors continuously measured dust concentration, wind patterns, and air quality, allowing machine learning models to predict high-risk dust events. This ensured that dust suppressants were applied more efficiently, minimizing particulate matter deposition on vegetation and soil. Reducing airborne dust levels by nearly 50% (Table 9) created favorable conditions for plant respiration and photosynthesis, indirectly boosting vegetation growth. In addition to dust suppression optimization, predictive analytics also supported rehabilitation efforts by guiding the selection of plant species and soil stabilization techniques. Scenario-based modeling was used to identify the most effective planting zones based on microclimatic conditions, ensuring higher survival rates for native species. The monitoring system also allowed for adaptive management, where real-time vegetation health data helped adjust irrigation schedules and soil amendment practices, maximizing the efficiency of the rehabilitation process. The distinction between rehabilitation and sensor-based interventions is significant for understanding the mechanisms driving environmental recovery. Direct rehabilitation efforts physically restored the landscape and facilitated vegetation growth, while sensor-based interventions created a favorable environment for plant establishment by controlling dust levels and informing adaptive management strategies. The utilization of both approaches resulted in a synergistic effect, maximizing ecological restoration outcomes while ensuring sustainable quarry management.

The quantitative reductions in dust and noise levels at the limestone quarry of the National Cement Factory underscore the efficacy of targeted control measures in mitigating environmental impacts associated with quarry operations. The 30% reduction in both PM10 and PM2.5 dust concentrations reflects the effectiveness of non-toxic dust suppressants and improved operational practices in minimizing airborne particulate matter.

Similarly, the reductions of 10% in daytime noise levels demonstrate successful implementation of noise control measures, including the installation of barriers and optimized scheduling of noisy activities. These measures not only enhance environmental quality within the quarry site but also contribute to fostering positive relationships with surrounding communities by reducing noise disturbances. The nearest community is located approximately 8 km from the quarry site.

Physical barriers were strategically placed around noisy equipment and operations to reduce the propagation of sound waves beyond the quarry boundaries. Noise-reducing modifications were made to machinery and equipment used in quarry operations, such as mufflers and soundproofing enclosures, to minimize noise emissions.

The comparative analysis highlights the varied effectiveness of different control techniques in mitigating dust and noise pollution at the limestone quarry of the National Cement Factory. Non-toxic dust suppressants emerged as the most effective strategy for dust reduction, achieving a substantial 50% reduction in airborne particulate matter. In contrast, noise control measures, while effective in mitigating noise emissions, showed varying degrees of effectiveness. Noise barriers and operational scheduling demonstrated significant reductions in noise levels, particularly during critical periods, thereby improving environmental conditions and community acceptance. Equipment modifications, although effective, yielded moderate reductions in noise emissions, underscoring the importance of comprehensive noise management strategies that integrate multiple approaches.

Conclusions

This study investigated the effectiveness of advanced environmental monitoring systems, dust and noise control measures, and sustainable land reclamation practices at the limestone quarry of the National Cement Factory in Ethiopia. The primary objective was to assess how these interventions contributed to environmental sustainability and operational efficiency. The study deployed automated IoT-based monitoring systems to track air quality, dust dispersion, and noise pollution in real time. To complement real-time monitoring, machine learning models were applied to enhance predictive capability. By analyzing historical trends and real-time sensor data, these models forecasted pollution spikes, allowing for early interventions. This predictive approach not only improved compliance with environmental regulations but also optimized resource allocation by enabling proactive dust suppression and noise reduction measures before conditions deteriorated. Additionally, a series of dust and noise suppression techniques, including the application of non-toxic dust suppressants and the installation of noise barriers, were implemented. The land reclamation efforts focused on biodiversity restoration, employing indigenous plant species and optimized water management strategies to enhance soil stability and prevent erosion. The results demonstrated significant environmental and operational benefits derived from these interventions. The key findings are summarized as follows:

• PM10 and PM2.5 concentrations decreased by 48% and 47%, respectively, demonstrating the effectiveness of dust suppression measures.

• NO₂ levels were reduced by 40%, highlighting improved emissions management.

• Dust concentrations at loading, crushing, and haul roads were reduced by 31–47%, reflecting the success of localized dust suppression strategies.

• Noise levels in critical areas, including the crushing and loading zones, decreased by 14–17%, enhancing workplace safety and mitigating community disturbances.

• Machine learning models achieved high accuracy (81% for air quality, 78% for dust dispersion, and 88% for noise levels), supporting real-time interventions and proactive environmental management.

• Vegetation cover increased by 266%, enhancing ecological restoration.

• Soil organic matter improved by 66%, and moisture content increased by 80%, reinforcing the effectiveness of sustainable land management.

• Soil erosion rates were reduced by 116%, demonstrating improved landscape stability.

Beyond the observed environmental improvements, the integration of predictive analytics strengthened the effectiveness of mitigation strategies. Unlike conventional monitoring that detects changes only after they occur, machine learning-based forecasting enabled quarry operators to anticipate air quality deterioration and high-noise events. This allowed for dynamic operational adjustments, such as modifying equipment schedules and activating targeted dust suppression, ensuring environmental compliance with minimal resource wastage.

While the study presents compelling evidence of environmental enhancements, it is essential to acknowledge certain limitations. The study relied on a single-site analysis, which may limit its broader applicability across different quarrying environments. Additionally, external factors such as seasonal climate variations and operational fluctuations may have influenced the measured improvements, requiring long-term assessments to ensure consistency in environmental benefits. While this study demonstrates the advantages of real-time monitoring and predictive modeling in environmental management, further refinement of machine learning-based forecasting is necessary to enhance its precision across varying climatic and operational conditions. The shift from a reactive to a predictive approach signifies a fundamental change in sustainability planning within quarry operations. Future research should explore the broader applicability of this predictive framework in diverse mining environments, assessing its long-term cost efficiency and operational scalability. In the following, other ideas are suggested for possible new studies:

• Expand multi-site evaluations to validate the effectiveness of interventions across various quarrying operations and climates.

• Enhance predictive analytics models by integrating additional environmental parameters and refining machine learning algorithms.

• Investigate the long-term ecological effects of land reclamation beyond the immediate restoration phase.

• Explore the economic feasibility and scalability of environmental monitoring and intervention strategies in different industrial sectors.

The successful implementation of real-time environmental monitoring systems and sustainable quarrying practices at the National Cement Factory provides a model for broader adoption in the mining and industrial sectors. The implementation of green mining practices in open mines, together with the placement of wine monitoring systems in hot and dry regions with higher environmental degradation, should be given significant consideration in other mines, including those in region like Saudi Arabia and Ethiopia.

Author contributions

Hussein A. Saleem: Conceptualization, methodology, validation, data curation, writing—original draft preparation, visualization. Abebe Temesgen Ayalew: Methodology, software, data curation, formal analysis, investigation, resources, writing—review and editing.

Funding

This research received no external funding.

Data availability

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.