ExACT: An End-to-End Autonomous Excavator System Using Action Chunking With Transformers

TL;DR

This paper presents ExACT, an end-to-end autonomous excavator system that uses action chunking with transformers and imitation learning to control excavator operations directly from raw sensor data, demonstrating effective task completion in simulation.

Contribution

Introduces ExACT, the first end-to-end autonomous excavator system using action chunking with transformers and minimal human demonstrations for direct control from raw sensor inputs.

Findings

Successfully completes excavation tasks in simulation.

Uses minimal human demonstration data for training.

Employs multi-modal sensor data for control decisions.

Abstract

Excavators are crucial for diverse tasks such as construction and mining, while autonomous excavator systems enhance safety and efficiency, address labor shortages, and improve human working conditions. Different from the existing modularized approaches, this paper introduces ExACT, an end-to-end autonomous excavator system that processes raw LiDAR, camera data, and joint positions to control excavator valves directly. Utilizing the Action Chunking with Transformers (ACT) architecture, ExACT employs imitation learning to take observations from multi-modal sensors as inputs and generate actionable sequences. In our experiment, we build a simulator based on the captured real-world data to model the relations between excavator valve states and joint velocities. With a few human-operated demonstration data trajectories, ExACT demonstrates the capability of completing different excavation tasks, including reaching, digging and dumping through imitation learning in validations with the simulator. To the best of our knowledge, ExACT represents the first instance towards building an end-to-end autonomous excavator system via imitation learning methods with a minimal set of human demonstrations. The video about this work can be accessed at https://youtu.be/NmzR_Rf-aEk.