Safety Management in Construction Site by the Implementation of Image Processing on Raspberry Pi: Risk of Truck Collision Towards the Workers

The construction industry is one of the most important sectors that help in economic growth around the world. Generally, all construction sites consist of many equipment, truck, materials and variety of workers so there usually lead to accidents occurred mostly with the reasons of truck collisions towards the worker. This problem can be overcome by doing object detection and notify the workers when there is close truck towards them. The object detection process is for detecting the workers that close to the truck. Then, the speaker device will trigger audio sound as an alert tool for the workers to avoid the collision. Raspberry pi 4B with pi camera installed on it will be used as the main device to run the Node-RED flow based platform. Node-RED is flow-based programming development tool that will deploy object detection flow. Image processing is readily included in the pre-trained model of coco-SSD from Tensorflow.js to classify the image class before trigger the speaker. This system can help the workers to avoid collision with the trucks that close to them and on the future work this device may be used to measure between them and give alerts through the mobile phone or other devices.