X Square Robot, a technology company backed by industry giants Alibaba, ByteDance, Xiaomi, and Meituan, has introduced a new embodied AI foundation model designed for domestic use.
The first set of deployments is scheduled to begin within the month, the company announced at its “Born to Bot, Bot to Family” launch event.
“Robots in factories and robots in homes are fundamentally different,” Qian Wang, founder and CEO of X Square Robot. “In factories, they repeat the same action 10,000 times. In a home, they may need to perform 10,000 different actions, each in a different context. The real challenge is not repetition, but whether a robot can execute new, untrained actions in an unstructured environment.”
The company also revealed its World Unified Model (WUM) Architecture at the event – a system designed to integrate vision, language, action, and physical prediction into a single training framework from the start.
X Square says its first full implementation of the WUM architecture, named Wall-B, departs from traditional modular systems that train perception and control separately. By training these capabilities jointly, the model is equipped with an inherent understanding of physical dynamics, such as force, friction, and collision, rather than being layered on afterwards.
“We train vision, language, action and prediction in the same network from day one,” said Wang Hao, chief technology officer of X Square. “Human infants do not learn to see, move and communicate in isolated stages. They learn by integrating perception and action at the same time, with constant feedback from the physical world. That is the principle behind our architecture.”
X Square Robot’s strategy relies on two primary foundations: a data collection process focused on non-staged, real-world home environments and a physics-aware predictive mechanism. The goal is to prepare robots for “long tail” scenarios, such as misplaced objects or unexpected human activity, by allowing the machine to anticipate physical outcomes before they occur rather than reacting after the fact.
During the event, the company showcased the system’s capabilities through live demonstrations. In one experience zone, a robot demonstrated flower arrangement while adjusting its grip and motion in real time to adapt to visual occlusion and shifting variables.
Acknowledging that the technology is still in early stages, Wang said the systems are prone to errors, such as misplacing household items or requiring pauses to process complex actions.
The upcoming deployment to residential households marks the next phase of X Square’s efforts to establish a permanent presence in the home robotics sector.
