The development of a spectacular excavator automation system has been completed that integrates sensing, planning and control capabilities to allow the excavator to load material by itself without human intervention, over a long period of time and achieving a performance almost equivalent to that of an experienced human operator.
This technological advance is the work of researchers from the University of Maryland in the United States and the Robotics and Self-Driving Laboratory of the Baidu Research company.
This unique excavator robotization system, created by Liangjun Zhang’s team, is called AES (Autonomous Excavator System).
The AES is one of the world’s first unmanned excavation systems that has been deployed in real scenarios and has operated continuously for more than 24 hours.
Excavators are vital for various kinds of activity, such as infrastructure construction, mining, and some rescue operations. The size of the global excavator market was $ 44.12 billion in 2018 and is projected to grow to $ 63.14 billion by 2026.
Given this forecasted market growth, construction companies in quite a few countries face an increasing risk of not being able to find enough qualified heavy machinery operators to hire, especially excavators. Additionally, the COVID-19 pandemic has exacerbated the labor shortage crisis.
Another factor that contributes to the need for robotic excavators is occupational safety. Some of the places an excavator must work are dangerous in themselves. For example, a site where toxic substances that can harm human health have been dumped, or a semi-collapsed building that can collapse at any moment.
Excavator operators suffer accidents in scenarios such as those described. In the United States alone, these accidents cause approximately 200 casualties a year in the United States alone.
For this and other reasons, the development of systems such as AES is a growing trend, along with the introduction of other robots, such as self-driving cars.
Basic diagram of the main systems of a robotic excavator with the AES system, which allow it to work alone. From left to right and from top to bottom: inclination sensor, pressure sensor, camera, LiDAR-type laser tracking system (so-called “Light Detection And Ranging”) that allows three-dimensional mapping with one level. Huge detail the scrutinized environment, and RTK positioning device. (Image: Baidu Research)
While most robots in the industry are comparatively smaller and operate in more predictable environments, robotic excavators must operate in a wide range of hazardous environmental conditions. They must be able to identify materials, avoid obstacles, and continue to operate in adverse weather conditions.
AES uses precise, real-time algorithms for perception, planning and control, along with a new architecture that incorporates these capabilities for autonomous operation. The perception module integrates multiple sensors to perceive the 3D environment and identify target materials, along with advanced algorithms for core job-related tasks.
With this modular design, the AES architecture can be used effectively by excavators of all sizes (including 6.5 and 7.5 ton compact excavators, 33.5 ton standard excavators and 49 metric ton large excavators) and is suitable for various applications.
To assess the effectiveness and robustness of the AES, the researchers, with the help of a specialized company, robotized an excavator and put it to work in a landfill, a real and pernicious scenario in which automation is in high demand. Despite the difficult task, the excavator with the AES system was able to operate continuously for more than 24 hours without any human intervention. (Source: NCYT from Amazings)