Rescue and Safety of Workers at Construction Sites by Implementing AI and RSSI Technologies

Abstract

Construction sites pose significant risks to workers, particularly during structural collapses, where trapped individuals become difficult to locate under debris. To address this, we propose a system integrating Received Signal Strength Indicator (RSSI) technology to determine the precise location of workers using signals from smartphones or wearable devices, enabling rapid rescue operations. Our solution not only accelerates disaster response but also incorporates real-time monitoring to alert supervisors via audible alarms if workers deviate from designated safe zones, thereby preventing unauthorised movements. Additionally, we extend safety protocols to prefabricated construction, a sector challenged by safety vulnerabilities due to large-component assembly and high-risk lifting operations. By employing Internet of Things (IoT) sensors, we analyse factors influencing safety risks, such as unsafe worker behaviours, excessive workload, and insufficient supervision. The system also leverages historical and real-time data to predict hazards (e.g., impending structural failures), enabling pre-emptive interventions

Introduction

Construction sites are inherently dangerous due to dynamic environments, heavy machinery, and human factors like fatigue or poor supervision. A significant portion of injuries (e.g., 20% from falling objects or being trapped) is exacerbated by delayed emergency response and limited real-time visibility. Traditional safety protocols often fall short due to their reactive nature and lack of real-time data integration.

Proposed Solution

This study introduces a technology-driven, proactive safety framework integrating:

• IoT (Internet of Things)

• RSSI (Received Signal Strength Indicator)

• AI-based analytics

These components work together to provide real-time monitoring, predictive warnings, and automated emergency responses on construction sites.

Key Technologies and Features

1. Wearable IoT Devices & Sensors

   • Track heart rate, oxygen levels, and worker location

   • Detect PPE compliance using AI on edge devices (ESP32 + TensorFlow Lite)

   • Generate emergency alerts (e.g., SMS, buzzers) in under 5 seconds

2. RSSI-Based Location Tracking

   • BLE tags provide real-time, high-accuracy worker positioning, even in obstructed zones

   • Helps locate trapped workers within ±1.5 meters during emergencies

3. AI and Predictive Analytics

   • Use of machine learning (Random Forest Classifier) to predict hazards based on real-time and historical data

   • Cloud dashboards visualize live heatmaps, worker vitals, and risk zones

4. Edge & Cloud Integration

   • ESP32 microcontrollers process sensor data locally

   • Cloud platforms (e.g., AWS IoT) perform deeper analysis and generate predictive insights

Literature Insights

The literature supports the use of:

• Computer Vision for behavioral monitoring (Li et al., 2024)

• HRV-based stress analytics (Tran & Pentek, 2023)

• IoT-based multi-sensor systems for real-time monitoring (Sabu & Kumar, 2022)

• Integrated safety frameworks combining wearables and cloud dashboards (Hayward et al., 2022)

• Autonomous alert systems for emergency detection (Elhassan et al., 2018)

• Immersive, scenario-based safety training (Chan et al., 2015)

• Leadership's role in shaping safety culture (Griffin & Hu, 2013)

Methodology

A. Module Description

• Location Tracking Module: Uses BLE RSSI to localize workers.

• Health Monitoring Module: Tracks vital signs, triggers alerts on abnormalities.

• AI Safety Gear Detection: Recognizes missing PPE items with ~98% accuracy.

B. Working Principle

• Data from wearables and cameras is filtered and analyzed locally (ESP32), then sent to the cloud.

• AI models detect patterns like overexertion, equipment strain, or heat risks.

• Supervisors are alerted via dashboards and SMS with real-time, actionable insights.

Results

• 60% reduction in near-miss incidents

• 70% faster emergency response time

• Random Forest Classifier used for multi-class hazard prediction, evaluated using:

  • Accuracy, Precision, Recall, F1-score

  • Confusion Matrix for class-wise insight

Conclusion

This project establishes a transformative framework for construction safety by integrating RSSI-based localisation, AI-driven PPE compliance, and real-time health monitoring into a unified IoT system. By employing ESP32 microcontrollers for edge processing, the solution achieves rapid worker tracking during emergencies like collapses, reducing rescue times from 30+ minutes to under 10 seconds. The AI models demonstrate 98% precision in detecting safety gear violations, while health sensors enable proactive medical interventions through instantaneous alerts for anomalies like cardiac events. Future enhancements include deploying digital twin models would simulate site-specific risks, enabling preemptive adjustments to safety protocols. Edge computing optimizations, such as quantizing AI models for ESP32, could further reduce latency in emergency alerts. Additionally, adopting 5G connectivity and blockchain-based data logging would enhance real-time communication and auditability across global construction networks. Collaborations with AR/VR platforms could also transform safety training, using real-time site data to simulate emergencies for immersive worker preparedness.

References

[1] R. M. Choudhry and D. Fang, “Why operatives engage in unsafe work behavior: Investigating factors on construction sites,” Saf. Sci., vol. 46, no. 4, pp. 566–584, 2008. [2] A. Adibfar, A. Costin, and A. Wehle, “Leading indicators—A conceptual IoT-based framework to produce active leading indicators for construction safety,” Safety, vol. 5, no. 4, Art. no. 86, 2019. [3] P. Abrahamsson et al., “Security challenges in IoT development: A software engineering perspective,” in Proc. XP2017 Sci. Workshops, Art. no. 11, 2017. [4] D. J. Edwards and G. D. Holt, “Case study analysis of risk from using excavators as ‘cranes’,” Autom. Constr., vol. 19, no. 2, pp. 127–133, 2010. [5] M. A. Griffin and X. Hu, “How leaders differentially motivate safety compliance and safety participation: The role of monitoring, inspiring, and learning,” Saf. Sci., vol. 60, pp. 196–202, 2013. [6] S. Hayward, K. van Lopik, and A. West, “A holistic approach to health and safety monitoring: Framework and technology perspective,” Internet Things, vol. 20, Art. no. 100606, 2022. [7] K. Arifin et al., “Occupational safety and health management in the construction industry: A review,” Int. J. Occup. Saf. Ergon., vol. 24, no. 4, pp. 493–506, 2018. [8] K. Kartam, N. A. Flood, and P. Koushki, “Construction safety in Kuwait: Issues, procedures, problems, and recommendations,” Saf. Sci., vol. 36, no. 3, pp. 163–184, 2000. [9] O. Elhassan, R. Kanan, and R. Bensalem, “An IoT-based autonomous system for workers’ safety in construction sites with real-time alarming, monitoring, and positioning strategies,” Autom. Constr., vol. 88, pp. 73–86, 2018. [10] G. Chan, H. Li, M. Lu, and M. Skitmore, “Proactive training system for safe and efficient precast installation,” Autom. Constr., vol. 49, pp. 163–174, 2015. [11] W. Chen et al., “Safety climate, safety behavior, and worker injuries in the Chinese manufacturing industry,” Saf. Sci., vol. 78, pp. 173–178, 2015. [12] S. Mohamed, “Safety climate in construction site environments,” J. Constr. Eng. Manage. , vol. 128, no. 5, pp. 375–384, 2002. [13] M. J. Skibniewski, “Information technology applications in construction safety assurance,” J. Civ. Eng. Manage., vol. 20, no. 6, pp. 778–794, 2014. [14] J. Kivisto-Rahnasto, N. Nenonen, and S. Tappura, “Managers’ viewpoint on factors influencing their commitment to safety: An empirical investigation in five Finnish industrial organizations,” Saf. Sci., vol. 96, pp. 52–61, 2017. [15] T. Cheng, Y. Fang, and J. Teizer, “Location tracking and data visualization technology to advance construction ironworkers’ education and training in safety and productivity,” Autom. Constr. , vol. 35, pp. 53–68, 2013. https://doi.org/10.1016/j.autcon.2013.03.004 . [16] M. Bhasi and M. N. Vinodkumar, “Safety management practices and safety behavior: Assessing the mediating role of safety knowledge and motivation,” Accid. Anal. Prev. , vol. 42, no. 6, pp. 2082–2093, 2010.

Copyright

Copyright © 2025 M Mercella Christie, Dr. G. Kulanthaivel, Dr. R. Suja Mani Malar . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.