DROPS: A Drone-based Cyber-Physical-Human System for Workers’ Safety in Road Construction Sites

Author/Presenter: Pestelli, Matteo; Pascucci, Federica; Kyriakou, Charalambos; Laoudias, Christos; Kyrkou, Christos; Cavone, Graziana
Abstract:

This work presents the design of DROPS (Drone-based Roadwork Operators Protection System), a drone-base Cyber-Physical-Human System for enhancing road construction site workers’ safety, integrating computer vision, smart wearables, and wireless communication. Traditional construction site safety solutions often lack real-time hazard assessment and worker-specific alerts, leading to delayed responses to dangerous situations. Our system addresses this issue by combining aerial and ground sensing with adaptive warning mechanisms.

A drone equipped with a camera and computer vision algorithms detects vehicles, estimates their speed and distance, and relays this information to a master unit based on an Arduino Mega. A smart traffic cone, featuring an Inertial Measurement Unit (IMU) and a Doppler-based speed sensor powered by an Arduino Nano, detects vehicle speed and impacts. Additionally, smart wristbands, each equipped with an Arduino Nano, RGB LEDs, a buzzer, and a vibration motor, provide tiered alert levels (visual, auditory, and haptic) based on a dynamically computed danger index. The master unit aggregates data from the drone and cone, evaluates risk levels, and distributes alerts through a dedicated PC interface, an onboard display, and worker-worn wristbands. Workers can also manually trigger an SOS alert via their wristbands in case of emergencies.

By leveraging wireless communication between all components, our system ensures real-time responsiveness without requiring extensive infrastructure. Initial experimental validations demonstrate effective hazard detection and timely warning distribution, highlighting the system’s potential to reduce accident risks in high-exposure work environments.