A Virtual Simulation Of Autonomous Construction Equipment Operations For Enhancing Safety Awareness

Abstract

The Goal Of This Study Is To Simulate A Sensor-Integrated Construction Equipment That Can Detect Obstacles In A Simulated Construction Site. Various Approaches Of Construction Site Automation Attempted To Enhance The Level Of Autonomy For The Construction Management Purposes, Such As Safety Control, Site Monitoring, Or Progress Control, Rather Than For The Construction Operations. However, It Is Very Difficult To Achieve A Perfect Real-Time System Because It Requires A Series Of Information Process For Converting Into The Contextual Information Regarding The Management Tasks. Therefore, This Study Wanted To Focus On Fully Autonomous Operations Of Construction Equipment That Can Catch The Scene, Understanding The Situation, And Sending The Contextual Information To The Management Personal. The Scope Of This Presented Paper Is Limited To Providing The Overview Of Creating The Virtual Pickup Truck Model With 3D Laser Scanners And Cameras And Simulating The Collection And Processing Of The Sensing Data To Detect Obstacles. Based On Robot Operating Systems (ROS) Framework, A Semantic Description Of The Sensor-Integrated Truck Was Created And Point Cloud Library (PCL) Algorithms Were Applied To Detect Point Cloud Clusters. With The Capabilities To Recognize Objects Such As Workers In The Blind Spots, This Study Demonstrates The Potential To Enhance Worker Safety In Congested Construction Sites. Further Studies Include Autonomous Controls Based On The Sensor-Based Perception And Evaluation Of The Equipment In The Real Workplace.