Task planner design for an automated excavation system

Abstract

An automated excavation system, which is a robotic excavator with site modeling capability, is being developed by a Korean research consortium in order to improve the productivity, quality, and safety of conventional earthwork. This paper presents the excavation task planner devised to incorporate the intelligence of a construction planner and a skillful operator into the robotic control mechanism of the automated excavation system. The excavation task planner aims to generate an optimal excavation plan based on 3D models of the work environment and the excavator updated by various cognitive technologies. The structure of the task planner was designed in harmony with the sensing and the control schemes of the automated excavation system. The algorithms used to partition the work area and to generate the excavator path were developed as the critical components of the task planner. The suggested design of the excavation task planner focused on the functions required to utilize the automated excavator at actual construction sites. Case studies showed that the task planner was able to generate effective work plans that could be fed into the automated excavation system.