Publications / 2024 Proceedings of the 41st ISARC, Lille, France
In modern construction sites, tower cranes play a crucial role but often rely on multiple human operators. Despite the advancements of the Construction 4.0 era, a comprehensive framework for automated tower crane operations is currently lacking. This study proposes a framework that integrates a real-scaled construction site and tower crane into a physics-based simulation in ROS (robot operating system) framework to enable collision-free motion planning and control. Specifically, we develop time-varying linear quadratic regulators (LQR) for trolley and jib control while employing a proportion-integrated-derivative (PID) method for hoisting control. Additionally, we utilise 5th-order quintic spline trajectories to plan the desired pose of the payload, reducing acceleration discontinuities. The framework's effectiveness is validated through simulations of a real-scaled tower crane and construction site equipped with LiDAR sensors. The results demonstrate that higher-order trajectories effectively minimise oscillations in unactuated systems. Our scalable framework holds promise for real-scale operations in the field of tower crane automation.