Smart Highway Construction Site Monitoring Using Artificial Intelligence

The main goal of this project is to develop a method to detect, classify, monitor, and track the equipment, workforce, and other surrounding objects during construction, maintenance, and rehabilitation of transportation infrastructure by using artificial intelligence (AI) and a deep learning approach.

In this study, we will evaluate the performance of AI and deep learning algorithms to compare their performance in detecting and classifying the equipment in various construction scenes. Several edge-case scenarios with crowded scenes, where the target objects are occluded with other objects, will also be investigated. Figure 1 shows the process of image database collection and labeling. The detection accuracy and performance of the preliminary model will be improved once the proposed image database is developed in this study. We will provide a glossary of various roadway construction and maintenance/rehabilitation equipment and categorize them by activity type. Once the process of training and validation of the proposed models is complete, the algorithm will be able to detect, classify and track the trajectory of the most critical objects. Based on the availability of actual construction data, the applicability of the algorithm to both stationary and moving video sources will be evaluated. The models will be calibrated based on the properties of each image and video frame source. It should be noted that although higher video quality (i.e., higher resolution and number of frames per second) can improve the detection accuracy and tracking capabilities of the model, it will require advanced computational power and may introduce a lag in real-time tasks. Our goal is to find the optimized balance between the model capabilities in real-time detection and memory processing requirements.

USDOT Priorities:

This project supports one of the main USDOT RD&T strategic goals which is “Improving the durability and extending the life of transportation infrastructure.”