Doosan Bobcat announced on November 16 that it has successfully demonstrated autonomous operation technology for construction equipment in collaboration with AI solution company Maum AI.
On June 2, Doosan Bobcat signed a memorandum of understanding (MOU) with Maum AI at the Doosan Tower in Bundang, Seongnam, to develop autonomous operation technology for construction equipment. At that time, Doosan Bobcat agreed to provide data related to small construction equipment, while Maum AI would apply its vision, language, and action (VLA)-based AI technology to validate the potential for autonomous operations.
On June 19, the companies solidified their collaboration by signing a contract worth approximately 1.12 billion won for a proof of concept (PoC) of the VLA-based loader automation solution. Doosan Bobcat's North American subsidiary also participated in this contract, which is set to last until December 9. The recent field demonstration is the result of about five months of technology development following the MOU.
The demonstration utilized VLA-based AI technology, which recognizes the surrounding environment through sensors such as cameras and understands the work situation, linking this understanding to the actual operation of the equipment. Unlike traditional methods that rely on fixed rules or human commands, this technology allows the equipment to make judgments and operate based on the work environment and conditions.
Doosan Bobcat selected manure handling in livestock facilities as the demonstration task. A company representative explained, "Livestock farms are facing chronic labor shortages due to a declining rural population and aging workforce. Among these tasks, manure handling is particularly avoided due to its strong odor and heavy workload."
The skid loader, a small piece of equipment from Doosan Bobcat, demonstrated manure transport and cleaning operations without driver intervention at a small livestock facility in Pyeongtaek, Gyeonggi Province. The loader repeatedly scooped manure with its bucket and transported it to a designated location, loading it onto a trailer 2.4 meters high. It then recognized scattered manure on the floor, autonomously generated a cleaning path, and swept the manure into the bucket while moving forward, continuously performing the task.
Park Hyun-cheol, Chief Strategy Officer of Doosan Bobcat, stated, "We will enhance autonomous operations and unmanned technology through collaboration among global R&D departments and various domestic and international partners, developing practical solutions that improve customer productivity and work efficiency."
Choi Hong-seob, Chief Technology Officer of Maum AI, emphasized, "The significance of this demonstration lies in the advancement of our self-developed physical AI model tailored to the working environment of livestock facilities, allowing for precise control of construction equipment and improved autonomous operation performance. By automating tasks that people tend to avoid, we have shown that physical AI can address real issues in industrial settings."
Doosan Bobcat plans to develop and expand the application of autonomous operation scenarios focused on simple repetitive tasks or those difficult for humans in various industries, including livestock, agriculture, landscaping, and construction.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.