Studies of Autonomous UAV-UGV Teams in Construction Applications: A Survey from Advances and Challenges Perspective

Abstract Research studies on Unmanned Aerial Vehicles (UAV) and Unmanned Ground Vehicles (UGV) collaborative systems have been explored since the 1990s 1. The goal of a collaborative system, especially this heterogeneous system, is to combine the positives of both UAV and UGV, creating a diverse set of capabilities. These systems have mainly been used for the military and surveillance sectors, but nowadays, they are being researched to improve efficiency in the construction industry. Autonomous UAV-UGV systems can potentially increase efficiency by collecting data, completing automated construction tasks, and testing various other aspects. Based on this motivation, we present this survey paper on studying the different team approaches in construction applications. In addition, open research issues and research challenges are discussed and presented. One of the issues is that currently, data collection is completed manually, which is time-consuming and prone to errors. Other challenges for automation are that construction sites are often remote, uneven, and dynamic, construction tasks require dexterity and strength, and the tasks are rarely repetitive. Another significant issue is the effective autonomous navigation in congested GPS-denied situations where some locations are inaccessible to UGVs. The teams have conducted different experiments. One team collected data to map an indoor construction site, but they were unable to use GPS, and there were dynamic obstacles. Another team simulated their task management system, but the limitation was that certain bricks have unique requirements, which adds more complexity. The type of UAV varies from each study, but the main goal of the UAV is to either act as an external eye for the ground unit or to work side by side with the ground unit. While working side by side, the task hierarchy (order of tasks given to a specific robot) is essential, and optimizing this aspect can significantly improve the efficiency of task completion. All these experiments need to be conducted in different environments, indoors and outdoors, and the type of obstacles also need to vary to create a realistic environment. The importance and desirability of developing vision-based mobile robotic systems are explored. Based on these studies, it is possible to develop and deploy an automated multi-robot system to help in construction applications. The most prominent issues are the computational load on the CPU and obstacle avoidance on the blimp. The GPU can be arranged to compute heavy calculations, so the load on the CPU will be reduced. By adding more sensors, like ultrasonic sensors, to the blimp, obstacle avoidance can be achieved, as well as more accurate data. The hierarchical task solution can also be applied and tested in a real heterogeneous robotic system rather than a simulation.