Huawei’s “genius boy” startup company, humanoid robots are already doing housework
On March 11, Zhiyuan Robot released its first universal embodied base large model Zhiyuan Qiyuan large model (GO-1, Genie Operator-1), and simultaneously announced that its universal embodied robot has been mass-produced and 1,000 units have been rolled off the production line, covering industrial and commercial scenarios.
Zhiyuan Robot is headquartered in Zhangjiang Science City, Shanghai. Its co-founder and CTO (Chief Technology Officer) is Peng Zhihui, who was once called Huawei’s “genius boy”.
According to public information, Peng Zhihui was born in Ji’an, Jiangxi in 1993. After graduating from the School of Information and Communication of the University of Electronic Science and Technology of China in 2018, he later worked in the AI Laboratory of OPPO Research Institute. In 2020, Peng Zhihui joined the Huawei team with the highest annual salary of 2.01 million yuan in the “Huawei Genius Boy Program” to engage in research related to Ascend AI chips and AI algorithms.
Peng Zhihui resigned from Huawei at the end of 2022 and co-founded Zhiyuan Robot in February 2023. In August of the same year, the first embodied intelligent robot “Yuanzheng A1” developed by Peng Zhihui and his team made its first public appearance. In January 2024, Zhiyuan Robotics and Peking University established a joint laboratory to focus on solving key embodied intelligent technology problems.
Ren Guanghui, executive director of the Zhiyuan Embodied Research Center, mentioned that the current embodied model still faces four major difficulties in implementation. First, the generalization of the existing embodied model is relatively poor. For example, in the past, the model was trained in the laboratory, and its success rate would drop significantly when it was changed to a new scene; second, the expansion of new tasks is still facing challenges, that is, after training a new task, if you want to do other tasks, you need a lot of data, and the cost will be very high; third, different ontology data cannot be shared, which also leads to high data costs. For example, Zhiyuan Company has multiple ontologies, but when different ontologies are adapted to different scenarios, data is difficult to share, resulting in high costs; fourth, it is difficult to form a complete data reflux system, just like in the autonomous driving scenario, if the model cannot continue to evolve from the reflux data, it will lead to the model being unable to continuously improve performance.
Now, through open source, robots are moving towards universalization. They are no longer limited to single tasks, but can coordinate multiple tasks. They can also move from closed scenes to the open world, and are no longer limited to laboratories. Most of the previous generation of robots had some preset programs, but now robots are increasingly able to understand human instructions. The release of this large model can also accelerate the popularization of embodied intelligence.
The GO-1 large model released this time, with the help of human and various robot data, enables robots to acquire revolutionary learning capabilities, which can be generalized and applied to various environments and objects, quickly adapt to new tasks, and learn new skills. At the same time, it also supports deployment to different robot bodies, efficiently completes the landing, and continues to evolve rapidly in actual use.
The GO-1 large model will accelerate the popularization of embodied intelligence, and promote robots from tools that rely on specific tasks to autonomous entities with general intelligence, and play a greater role in multiple fields such as business, industry, and home.
Ren Guanghui said that it will take about 5 years for humanoid robots to enter the home market, and the price may be at the level of 50,000 yuan. However, for each scenario, each user and task, the capabilities required by the robot vary greatly, and the product forms are also different. But scenarios such as serving tea and making breakfast can currently be realized in home scenarios at the software system capability level with an investment of several thousand to ten thousand yuan.
Related topics:
- Ultenic T10 Lite Robot Vacuum Review: A Smart, Affordable Option for Cleaning and Mopping
- Gtech Offers £100 Discount on Popular AirRAM 3 Cordless Vacuum for Spring Cleaning
- Courtney Cox’s Shark Vacuum Tempted Me to Ditch Dyson for Good – Here’s Why