Generative Excavation Path Planning using BIM: A Simulation-based Approach with Pure Pursuit and Differential Drive Robots

Abstract

Earthmoving is ubiquitous in construction. However, persistently low availability of skilled labour and conventional techniques, especially in excavation works, has resulted in a consistent decline in construction productivity. Excavation is not new to automation; however, the focus has been on the automation of digging operations. On the contrary, excavation requires that a machine cover the whole area under consideration for successful operation instead of point-to-point movement. This research, therefore, implements generative excavation coverage path planning strategies using BIM. The generative parameters used for creating paths are the area, the size of the excavator/robot, and the number of excavators/robots. The path assessment criteria are completeness of coverage and time. The coverage path planning algorithm devised in Revit Dynamo is exported into MATLAB via Microsoft Excel to simulate differential drive robots using the pure pursuit algorithm to provide a proof-of-concept for the idea. Simulations showed that the coverage of excavation was complete with an overlap of 20-30% for the area of slope stability. The sweep pattern requires 3-5% less time than the spiral pattern. The variation is due to the different sizes of the excavators. For multiple agents, this may also vary depending on the number of agents. This approach affords greater flexibility to the planners by devising excavation strategies beforehand. Moreover, it would also help improve excavators’ on site manoeuvrability and productivity that, for years, have relied on operators’ judgement and experience. Furthermore, these paths can be transmitted directly to robots to perform excavation leading to the automation of the whole process.