Task planning strategy and path similarity analysis for an autonomous excavator

Abstract

Autonomous excavators have attracted significant interest because of their ability to improve productivity and safety. To facilitate their use in construction sites, a proper task planning strategy must be established in advance. This study focuses on the development of a task planning strategy for autonomous excavators. A complete coverage path planning (CCPP) algorithm was developed by considering the characteristics of earthwork and environmental constraints on the autonomous excavator. The algorithm's cost function considers the accessibility of the dump truck and the external condition of the work environment. This enables maximized collaboration with the dump truck to reflect practical solutions, while the majority of the CCPP algorithms consider only the moving distance and internal work environment. In addition, to compare the performance of this algorithm with a conventional task plan generated by a skilled excavator operator, an evaluation scheme that can generate a quantitative result of the path similarity is proposed. Using this evaluation scheme, five cases at distinct construction sites were analyzed to compare the performance of the proposed CCPP algorithm. The results indicate that the CCPP algorithm trends resemble manually determined paths in terms of path similarity. These findings suggest that this algorithm could contribute to the development of autonomous construction equipment—particularly in dynamic and collaborative environments.