CJ Logistics Takes First Step Toward Commercializing AI Humanoid Robots in Logistics Industry

Two Humanoids Deployed for Packaging Operations
Plans to Expand Humanoid Applications to Picking, Inspection, and More

CJ Logistics is accelerating physical AI-based logistics automation by moving its logistics humanoid robots beyond proof-of-concept and deploying them directly into the operational processes of actual logistics centers. After verifying the potential for humanoid robot applications in logistics field sites last year, the company has expanded the scope to operational sites handling real customer logistics.

On September 3, CJ Logistics announced that it recently deployed two bimanual humanoid robots at the Yangji Olive Young logistics center in Yongin, Gyeonggi Province, assigning them to product packaging processes. The humanoid robots are tasked with placing cushioning materials inside boxes on the packaging line.

CJ Logistics' humanoid robot is performing product packaging processes at Yangji Olive Young Logistics Center. CJ Logistics

This marks the first time that humanoid robots, previously field-tested at the Gunpo Fulfillment Center last year, have been integrated into actual logistics operating processes. CJ Logistics plans to continuously improve the robots' task accuracy and responsiveness by leveraging work data accumulated through on-site operations for further AI learning.

Humanoids are considered a key technology to overcome the limitations of existing logistics automation equipment. Unlike manufacturing, which repetitively produces standardized products through fixed processes, logistics centers handle numerous types of goods that vary in shape, size, and material, with task conditions that constantly change based on the product and order specifications.

Traditional automation systems such as conveyors and fixed-position robots may offer high productivity for specific processes, but often require changes to both equipment and workspaces when new tasks are added. In contrast, humanoid robots, which replicate human form, can utilize existing workspaces to perform a variety of tasks, thus offering both greater flexibility and scalability in response to process changes. The ability to adapt flexibly to various tasks within human work environments is why humanoid robots are emerging as a new alternative for logistics automation.

CJ Logistics is enhancing the potential of humanoid applications by combining data collected from actual logistics field operations with AI technologies. Particularly, the company is continuously advancing the Robot Foundation Model (RFM), which serves as the robot's 'brain.' RFM is a core component of physical AI, enabling perception of surrounding environments through vision and sensors, and allowing judgment and execution of appropriate actions. By training the robots using both real-world data from logistics operations and simulated virtual data, CJ Logistics aims to strengthen the robots’ ability to respond to new products and work environments.

Beginning with this field operation, CJ Logistics plans to gradually expand the range of tasks assigned to humanoids. While starting with specific tasks such as inserting cushioning materials, the company seeks to broaden applications to more complex and sophisticated workloads including product picking, sorting, inspection, and packaging—drawing on the vast amounts of data accumulated across logistics sites. Ultimately, the goal is to evolve into a general-purpose logistics humanoid capable of performing multiple logistics processes in sequence.

To achieve this, CJ Logistics is also pursuing technological collaborations with external specialists and research institutions. The company is establishing strategic partnerships with a range of partners, such as Robotis (hardware), Aidin Robotics (robot hands), and RealWorldAI (RFM), thereby expanding the AI humanoid ecosystem. CJ Logistics is actively participating in government-led national R&D projects, including developing humanoid robot hands, and is working to transform logistics sites into AI-driven autonomous operations.

Junghee Kim, Head of the TES Logistics Technology Research Center at CJ Logistics, said, "The significance of this field deployment is that humanoids have moved beyond demonstrations and feasibility tests to actually performing roles in logistics operations. By combining the massive logistics datasets accumulated on-site with AI technology, we will enable humanoids to independently assess and carry out various tasks, and realize a future fulfillment model where humans and robots work together."

This content was produced with the assistance of AI translation services.