2026· International journal of advanced engineering and management research· Vol 11, pp. 189-197· 0 citations· 15 references
TL;DR
This study quantitatively evaluates the performance of a YOLO-based computer vision system for real-time hazard detection across construction, manufacturing, and healthcare environments in New York State and indicates that YOLO-based systems significantly outperform manual inspection across all metrics.
Abstract
This study quantitatively evaluates the performance of a YOLO-based computer vision system
for real-time hazard detection across construction, manufacturing, and healthcare environments
in New York State. The analysis compares YOLO-based detection with traditional manual
inspection using key performance metrics, including mean average precision (mAP), recall,
precision, time-to-detection, and personal protective equipment (PPE) compliance rates. Results
indicate that YOLO-based systems significantly outperform manual inspection across all metrics,
demonstrating higher detection accuracy, faster response times, and improved compliance
monitoring. The findings provide empirical evidence supporting the effectiveness of artificial
intelligence–enabled safety systems in enhancing hazard detection performance and advancing
proactive safety management practices.
Results indicate the vision model achieves 92% PPE detection accuracy and practical implications highlight the need for robust encryption user training and continuous model retraining, and future research should examine long-term behavioral impacts and cross-plant scalability.
A. Mahajan, K. Dharwadkar· International Journal of Inn...· 0 citations
Highlights What are the main findings? Improved Precision and Localization Quality: Compared with YOLO26n, YOLO-SCC improves precision by 6.1 percentage points and mAP50-95 by 3.3 percentage points, while recall and mAP50 increase by 0.9 and 0.1 percentage points, respectively. Validation of Module Effectiveness: The designed SCC enhancement structure (SPDConv, CBAM, ContextAggregation) effectively addresses three core challenges in underground detection: detail loss, background interference, and insufficient contextual information. The efficacy of each module is validated through ablation studies. What are the implications of the main findings? Provides a Feasible Solution for Lightweight Detection in Complex Industrial Scenarios: This study demonstrates that through carefully designed lightweight enhancement modules, the robustness of models in extreme environments can be significantly improved without substantially increasing computational costs. This offers important guidance for deploying high-performance safety monitoring systems on resource-constrained edge devices. Directly Advances Intelligent Safety in Coal Mining: By enabling more accurate and reliable detection of underground personnel, this method can help reduce safety incidents caused by missed detections or false alarms. It provides core technological support for developing a new generation of intelligent mine safety monitoring platforms and emergency response systems. Abstract To address the challenges in personnel detection in underground coal mines, such as uneven lighting, dust occlusion, cluttered backgrounds, and significant scale variation of targets, this paper proposes a lightweight object detection method named YOLO-SCC. Based on the lightweight detection network YOLO26n, the SCC lightweight enhancement structure, composed of SPDConv, CBAM, and ContextAggregation, is constructed to improve feature representation and target recognition capability in complex underground environments while striving to maintain model compactness. Specifically, SPDConv optimizes the downsampling process to reduce the loss of feature details, which helps preserve edge and texture information of miners under low-light conditions, at long distances, or at small scales. CBAM adaptively weights features from both channel and spatial dimensions, enhancing the model’s focus on the main body of miners and suppressing irrelevant interference from light reflections, equipment structures, and dust noise. The ContextAggregation module aggregates richer contextual semantic information, strengthening the model’s discriminative ability for personnel targets under occlusion, dense distribution, and complex backgrounds. Experimental results show that with only a modest increase in parameters (from 2.38 MB to 2.88 MB) and computational cost (from 5.2 GFLOPs to 6.0 GFLOPs), YOLO-SCC achieves precision of 88.8%, recall of 82.2%, mAP50 of 85.3%, and mAP50-95 of 51.2%. These represent improvements of 6.1, 0.9, 0.1, and 3.3 percentage points, respectively, over the baseline YOLO26n. The results suggest that YOLO-SCC primarily improves false-positive suppression and localization quality in complex underground environments while maintaining a lightweight model scale.
Zhe Wu, Xuanrui Zhang, Yanping Cui et al.· Italian National Conference...· 0 citations
Overloading in public utility buses remains a persistent transportation safety and regulatory problem, particularly in urban environments, where monitoring is often conducted through manual inspection. This study developed and evaluated a non-contact overloading detection system for public utility buses using the YOLOv8 object detection algorithm, Raspberry Pi 4, camera module, GPS, and real-time alert transmission. The study employed a descriptive-developmental quantitative design and Agile software development methodology. Fifty respondents participated in the evaluation, consisting of 44 passengers and 6 authorities. System performance was assessed through controlled trials, ISO/IEC 25010-based survey questionnaires, and object detection metrics, including precision, recall, and mean average precision (mAP). Test trials showed detection accuracies ranging from 67% to 82%, with one case of over-detection. Real-time processing achieved an average speed of 0.04 to 0.09 seconds per frame. Survey-based evaluation indicated high acceptance across hardware detection (overall mean = 4.40), alert system performance (overall mean = 4.41), and real-world system performance (overall mean = 4.37). ISO/IEC 25010 evaluation also showed strong results in usability (4.46), security (4.44), maintainability (4.44), reliability (4.37), efficiency (4.37), and functionality (4.19). YOLOv8 performance metrics reported precision = 0.94, recall = 0.91, mAP50 = 0.95, and mAP50-95 = 0.88. Findings indicate that the proposed system is feasible for real-time overloading monitoring, though performance remains sensitive to occlusion, lighting, crowd density, and network stability.
Dhane Kian Hellie, John Arnie Barba, Ann Paulina Calida et al.· Asian Research Journal of In...· 0 citations
Operator fatigue remains a significant occupational risk factor in the mining industry, affecting both safety outcomes and operational efficiency. This study aims to evaluate the effectiveness of a camera-based fatigue alertness system in reducing fatigue-related events and improving productivity at one of Indonesia’s gold mining operations. A mixed-methods approach was applied, integrating four years of historical operational data (2021–2024) with qualitative insights from dump truck operators and managerial personnel. Quantitative analysis focused on the frequency and severity of fatigue events and delay durations, while qualitative data were obtained through structured questionnaires and semi-structured interviews. The results indicate a substantial reduction in critical fatigue events, down to near zero by 2024, along with a 75% decrease in total fatigue-related delays. Notably, moderate fatigue alerts dropped by 97%, and controlled fatigue alerts declined by 78%, reflecting the system’s effectiveness in promoting behavioural change. Additionally, overspeed events as one of key parameters monitored also decreased consistently. Management responses confirmed high satisfaction with the system’s impact and suggested further integration of alert notifications and system accuracy enhancements. Operators generally acknowledged the system’s benefits but raised concerns about occasional false positives. This study concludes that real-time fatigue alertness technology, when embedded within a comprehensive safety management framework, can effectively enhance safety, reduce operational risk, and generate measurable financial value. The findings support the integration of smart fatigue detection tools as a strategic component of behaviour-based safety systems in the mining industry.
Tasya Millenia, E. Suryani, Arif Johar Pratama· IOP Conference Series: Earth...· 0 citations
This research investigates the feasibility and effectiveness of implementing localized artificial intelligence-based system for enhancing defensive driving through automated urban hazard detection. For this purpose, a lightweight You Only Look Once version 11 nano (YOLOv11-nano) model containing 2.6 million parameters was fine-tuned on a localized dataset consisting of 18,132 images collected over 300 days in Chiang Mai, Thailand. These results reveal a mean Average Precision (mAP) of 0.96 when using an intersection over union (IoU) threshold of 0.5. Even if the precision rate is quite high in ideal conditions, there is an observed decrease of 13.1 percent in the Average Detection Confidence due to monsoon rain and night-time conditions. Thus, a safety margin has been established for the application of Advanced Driver Assistance Systems. The reliability test conducted across seven localized object categories reveals consistent detection performance for vulnerable objects, such as motorcycles and pedestrians. Furthermore, a cost-benefit analysis based on localized accident cost data confirmed the economic viability of implementation in commercial transporting services.
Seungyeop Kang, Jira Yammeesri· ECTI Transactions on Compute...· 0 citations
As of 2022, the United States had 620,669 bridges that require regular inspections to ensure their safety and serviceability. These inspections are conducted in accordance with Federal Highway Administration regulations and are scheduled based on bridge condition and mandated inspection intervals. During each inspection, more than 100 bridge characteristics including location, material, geometric, and operational data are collected and used to determine the Structural Evaluation Rating (SER), a key indicator of overall bridge condition. This study investigates the application of machine learning techniques, including Pearson’s Correlation, Decision Trees, and Random Forest models, to identify and visualize relationships between bridge characteristics and the SER. The proposed methods are expected to provide rapid and reliable predictions of bridge condition based on selected characteristics and historical inspection data. In addition to improving prediction accuracy, these models offer intuitive visual interpretations of the factors influencing bridge performance. When integrated with existing inspection scheduling practices, these techniques may support more efficient and data-driven bridge management and may also assist engineers in selecting design characteristics that enhance long-term bridge performance.
Akm Anwarul Islam, Natalie Rose· International Journal of Bri...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
