Autonomous Aerial Robotic Surveying and Mapping with Application to Construction Operations
This work addresses the need for systematic surveying and inventory tracking in construction operations, representing an incremental advancement in applying existing robotic methods to a specific domain.
The paper tackles the problem of autonomous inspection and mapping of construction sites by developing a flying robotic system that can operate with or without prior environmental knowledge, delivering high-fidelity 3D maps with color and thermal data, and has been field-tested in a public housing construction project.
In this paper we present an overview of the methods and systems that give rise to a flying robotic system capable of autonomous inspection, surveying, comprehensive multi-modal mapping and inventory tracking of construction sites with high degree of systematicity. The robotic system can operate assuming either no prior knowledge of the environment or by integrating a prior model of it. In the first case, autonomous exploration is provided which returns a high fidelity $3\textrm{D}$ map associated with color and thermal vision information. In the second case, the prior model of the structure can be used to provide optimized and repetitive coverage paths. The robot delivers its mapping result autonomously, while simultaneously being able to detect and localize objects of interest thus supporting inventory tracking tasks. The system has been field verified in a collection of environments and has been tested inside a construction project related to public housing.