Construction work is one of the least intelligent and automated industries in the world. With the rapid development of robotics technology, a lot of companies and institutes explore how robotics can be involved in construction work to improve efficiency, overcome labor shortages and save cost. Since a large proportion of construction operations are incredibly repetitive, construction robotics is expected to solve the disadvantage of manual labor and manipulate more precisely. Automation is the key to meeting demand and autonomous mobile robots are an integral part of construction automation which has a great number of application scenarios. Different actuators can be embedded on the unmanned mobile chassis to fulfill corresponding needs, such as laying, painting, leveling, inspection, e...[ Read more ]

Construction work is one of the least intelligent and automated industries in the world. With the rapid development of robotics technology, a lot of companies and institutes explore how robotics can be involved in construction work to improve efficiency, overcome labor shortages and save cost. Since a large proportion of construction operations are incredibly repetitive, construction robotics is expected to solve the disadvantage of manual labor and manipulate more precisely. Automation is the key to meeting demand and autonomous mobile robots are an integral part of construction automation which has a great number of application scenarios. Different actuators can be embedded on the unmanned mobile chassis to fulfill corresponding needs, such as laying, painting, leveling, inspection, etc.

Localization plays a key role in the automation of mobile robots and it is an essential step for further navigation and manipulation. This thesis addresses the problem of localizing a mobile robot on the construction site for both indoor and outdoor environments. Lidar, RTK GPS, inertial measurement units, and wheel encoders are combined to output accurate position information with sensor fusion. Moreover, building information modeling(BIM) has been increasingly adopted in the construction process and it contains geometric and semantic information which can be exploited by construction robotics for a better understanding of the environment. We also explore combining information from BIM for localization. Experiments in both indoor and outdoor environments show a reasonable positioning result which can enable further navigation.