Paper presented at ISTVS 2025 | 55th Conference of the International Society for Terrain-Vehicle Systems https://doi.org/10.56884/PQZP98ES
Authors: Mattias Linde, Daniel Lindmark, Sandra Ålstig, Martin Servin
Keywords: Simulation; lunar constriction equipment; autonomous earthmoving
Abstract:
Planning of lunar construction projects, such as the Artemis Base Camp, is challenging. It includes deciding the design and numbers of different construction machines and how to coordinate them to reach certain time and energy budgets. The machines need to be unmanned to avoid human exposure to the harsh lunar environment. They also need to be highly autonomous due to the poor and delayed communication. Simulation has a central role for finding the necessary solutions for all this.
We present a simulation framework for lunar construction work involving multiple autonomous machines. The framework supports modelling of construction scenarios and autonomy solutions, execution of the scenarios in simulation, and analysis of work time and energy consumption throughout the construction project. A behavior tree manages the operational logic and error handling. High-level decision-making is separated from lower-level control algorithms, with the two connected via ROS2. Machine movements are planned using inverse kinematics. The simulations are based on the physics engine AGX Dynamics that integrates contacting multibody dynamics with deformable terrain.
We present results from a test case that involves an excavator and a dump truck that jointly remove lunar soil to construct a ground shape with a specified 3D profile. The excavator switches between the tasks of digging material from the target area and dumping it on the dumptruck, which has the tasks of receiving material, drive back and forth between the excavation area and dump area, dumping material, and doing levelling at the excavation area usings a blade. All these tasks are in turn be broken down into subtasks, that rely on sensors and controllers. The whole system is represented and managed by a behaviour tree.
Last updated
