X Square Robot demonstrates high-speed logistics automation using embodied AI

X Square Robot has demonstrated a significant advancement in logistics automation through its embodied AI system. During a live demonstration, the company’s dual-arm robot, equipped with standard industrial grippers rather than complex humanoid hands, successfully processed 1,816 parcels per hour with an accuracy rate exceeding 98%.

The system utilizes the proprietary WALL-B AI foundation model, which employs a World Unified Model (WUM) architecture. This allows the robot to integrate vision, language, and action while predicting physical changes in the environment. Unlike traditional systems that rely on rigid programming, the WALL-B model enables the robot to make real-time decisions, such as adjusting its grip, flipping packages to orient labels toward scanners, or using dual-arm coordination to handle large boxes.

By prioritizing sophisticated software over expensive, high-degree-of-freedom hardware like humanoid legs or five-fingered hands, X Square Robot claims to have reduced implementation costs by approximately 70% compared to humanoid alternatives like Figure AI. The company’s approach focuses on practical industrial application, aiming to solve labor-intensive and unpredictable tasks in logistics and warehousing by using smarter decision-making to compensate for simpler mechanical hardware.