Collaborative Tracking Control Strategy for Autonomous Excavation of a Hydraulic Excavator

A hydraulic excavator consists of multiple electrohydraulic actuators (EHA). Due to uncertainties and nonlinearities in EHAs, it is challenging to devise a proper control strategy. To tackle this issue, a major goal of our study is to provide an efficient control strategy to minimize tracking errors of the bucket tip position for autonomous excavation. To accomplish the goal, the study offers a collaboration of PID and fuzzy controllers that are used to compensate for contour errors and achieve accurate actuator position control, respectively. Co-simulation models including control algorithms and hydraulic components were created using Matlab and Amesim to validate the performance of the designed controllers. Simulations indicate that the proposed method enables achieving accurate tracking control for autonomous excavation with small tracking errors despite the nonlinear characteristics of the hydraulic excavator system.

Surface Reconstruction for Ground Map Generation in Autonomous Excavation

Excavator’s main tasks include digging, trenching, and ground leveling at construction sites, as well as work efficiency and safety can be improved by using an autonomous excavator. A prerequisite step to achieving an autonomous excavation is to obtain a sound perception of the surrounding ground. For this, a LiDAR sensor has been widely used to scan the environment. However, the point cloud generated by the LiDAR is not ideal for surface reconstruction to generate a ground map, as it suffers from flaws such as noise and outlier points. To tackle this issue, our paper proposes advanced methodologies for surface reconstruction algorithms.

Robotic embankment

Automating earth-moving tasks has the potential to resolve labour-shortage, allow for unseen designs and foster sustainability through using on-site materials. In this interdisciplinary project involving robotics and landscape architecture, we combine our previous work on autonomous excavation of free-form shapes, dynamic landscape design and terrain modelling tools into a robotic landscape system. It tightly connects survey, design and fabrication to exchange information in real-time during fabrication. We purposely built a LiDAR survey drone for tight integration. The design environment contains terrain modelling tools to balance cut and fill volumes for material-neutral, on-site construction. Its parametric nature allows it to adapt the geometry to changing site conditions during fabrication. Our autonomous walking excavator is used to create these free-form shapes in natural granular material. We propose an excavation planner for free-form embankments that computes the next excavation location and subsequently the location where the excavated soil should be dumped. This robotic excavation system achieves the world’s first autonomous completion of free-form embankments with high accuracy. A $$20\hbox { m}$$ 20 m long S-shaped and a two-faced embankment with a corner with roughly 0.03–0.05 m average error were created.