The 2026 government work report proposes to cultivate and develop future industries, including embodied intelligence. Will robots' going to work 'be realized in the near future? What obstacles do humanoid robots need to overcome when transitioning from the "runway" to the "workplace"? At the National People's Congress and the Chinese People's Political Consultative Conference, representatives and members gave their own opinions and thoughts. Liu Qingfeng, a National People's Congress representative and chairman of iFlytek, analyzed the necessary path for robots to "go on duty" from the perspective of technological evolution. He pointed out that behind the performance of the Spring Festival Gala robots is the improvement of their motion control ability, but for robots to truly enter home and industrial scenes, they still need three steps: first, to achieve environmental perception through models; Secondly, after perception, it is necessary to be able to independently plan the movement route, rather than a fixed route that has already been set; The third is to fully understand human gestures, expressions, and most importantly, voice control. By linking these together - the new generation of human-computer interaction, the new generation of scene understanding, the new generation of overall motion planning and design, combined with motion capabilities, robots can enter various application scenarios in homes and industries. This is a systematic innovation that requires linkage. ”Liu Qingfeng said. From 'active' to 'capable of working', interactive data has become a key factor determining the level of robot intelligence. He Han, a member of the National Committee of the Chinese People's Political Consultative Conference and chairman of Tianyu Digital Technology, frankly stated that although the current Chinese humanoid robot industry is hot, it is facing the embarrassment of "lively exhibition halls and deserted workstations". One of the reasons lies in the data bottleneck: unlike the general big model, which can use Internet data, the embodied intelligence requires a large amount of "task level, process level" interactive data, such as capture, assembly, transportation, etc., which is costly to obtain and difficult to label. Each research institution has its own data collection platform, sensor interface, and data format, forming a 'data island'. To this end, He Han suggests building several "national level embodied intelligent data collection and pre training centers" to form reusable datasets for typical tasks, unify data standards, clarify data ownership and compliance boundaries, and promote cross enterprise and cross platform reuse. In terms of scenario implementation, He Han emphasized that there is a huge real demand in the fields of manufacturing, logistics, commercial services, and elderly care. However, the fragmentation of scenarios, inconsistent acceptance criteria, and insufficient budget and iteration mechanisms have led enterprises to repeatedly oscillate between "project-based delivery" and "product based reuse". He suggested establishing a national level "list of humanoid robots on duty", prioritizing scenarios with quantifiable benefits and relatively structured environments, such as 3C assembly, warehousing and handling, hazardous chemical inspection, etc., and first forming a scale. The period from 2027 to 2029 will be a critical window for the development of humanoid robots worldwide. Most humanoid robots will become smarter, while a few will become less and less human like. ”National People's Congress representative and Chairman of Xiaopeng Motors, He Xiaopeng, gave a prediction. He also stated that in the future, some of the true humanoid robot factories will be self manufactured. "Today, everyone is exploring how to efficiently produce humanoid robots, but it has not yet reached a large enough scale. Embodied intelligence is becoming a new industry main track after big models, bringing algorithm capabilities from the screen world to the physical world, forming a new quality productivity of 'perception, decision-making, and hands-on'. ”He Han said. (New Society)