AGPNet -- Autonomous Grading Policy Network
This addresses the challenge of automating grading operations in construction or similar domains, representing a domain-specific incremental advancement.
The paper tackled the problem of autonomous control for a dozer grading uneven terrain with sand piles, achieving human-level performance and outperforming state-of-the-art machine learning methods in this task.
In this work, we establish heuristics and learning strategies for the autonomous control of a dozer grading an uneven area studded with sand piles. We formalize the problem as a Markov Decision Process, design a simulation which demonstrates agent-environment interactions and finally compare our simulator to a real dozer prototype. We use methods from reinforcement learning, behavior cloning and contrastive learning to train a hybrid policy. Our trained agent, AGPNet, reaches human-level performance and outperforms current state-of-the-art machine learning methods for the autonomous grading task. In addition, our agent is capable of generalizing from random scenarios to unseen real world problems.