During the 2026 Yabuli Forum Summer Annual Meeting, held in Chengdu from September 4th to 6th under the theme "Enterprise Innovation and Cycle Crossing," Xingdong Epoch Co-founder Xi Yue highlighted that the industry is at a pivotal juncture. He stated that this year represents the inaugural phase for the practical application of humanoid robots and embodied intelligence across various sectors.
Xi Yue noted that the most significant shift compared to last year is the initiation of pilot programs for humanoid robots across multiple fields. In industrial manufacturing, pilot projects have already been launched in automotive, consumer electronics (3C), and logistics. Meanwhile, in commercial service settings, trials and small-scale deployments are underway in supermarkets and unmanned pharmacies.
Robots have also begun undergoing functional tests and validation at familiar venues such as stages, sports arenas, and marathon events. Additionally, in numerous specialized operational environments, various types of robots have already engaged in related applications and deployments. Despite this progress, both domestically and internationally, the home environment is viewed as the ultimate goal for embodied intelligence. While this sector has yet to reach the mass-deployment phase, many companies are already investing in technological reserves and research and development.
He explained that previously, task-specific robots had completed validation for individual functions and tasks. The industry is now transitioning from specialized robots to general-purpose robots. The primary difference lies in their operational foundations: specialized robots largely depend on traditional solutions for single scenarios, relying primarily on pre-programming to execute fixed and repetitive tasks. In contrast, today's embodied intelligence and physical AI are distinguished by their use of large models and world models for reinforcement learning, with real-world data serving as the training basis. This enables robotic platforms to perceive and understand their physical environment, execute complex tasks, and achieve cross-environment capability generalization.
Comments