According to Omdia’s prediction, the global market size of robot artificial intelligence chipset will reach 866 million US dollars, which will contribute to the popularization of GenAI in the robot field.
Omdia pointed out that since Google introduced the transformer RT-1 for robot applications in 2022, many companies have been vigorously promoting the wide application of GenAI in robots. In addition to Google, companies such as Meta, OpenAI and Toyota are trying out or testing various basic models in their robot applications. China service robot suppliers such as CloudMinds and OrionStar have successfully developed their own basic models, and plan to integrate these models with client software systems.
However, GenAI is a resource-intensive technology. In most industries, GenAI is usually deployed in the cloud because the model needs large-scale GPU clusters for training and reasoning. On the contrary, robots prefer local processing, because the tasks and business-critical applications they participate in usually give priority to real-time control and ultra-low delay response.
Su Lianjie, chief analyst of Omdia application intelligence, commented: "Although NVIDIA’s GPU is still the preferred artificial intelligence chipset architecture for cloud infrastructure and robots, non-GPU vendors such as Qualcomm, Intel and AMD have launched artificial intelligence system on a chip (SoC) or dedicated artificial intelligence chipsets for device-side robot applications such as machine vision, navigation and mapping, and functional safety."
Encouragingly, the wide application of GenAI has also promoted the popularity of humanoid robots. Humanoid robot is the closest robot to human form, so many robotics experts believe that its combination with humanoid GenAI will be a natural fit. In this upsurge, companies such as Agility Robotics, Boston Dynamics, Figure, Tesla, Fourier Intelligent, Tatsu and Ubisoft have launched various humanoid robots for industrial and service fields. However, the technology is still in its infancy, and it is impossible to popularize it on a large scale in the next five years. Automated Guided Vehicle (AGV) and Autonomous Mobile Robot (AMR) are still mature application forms of GenAI.
Su Lianjie concluded: "The industry should not focus on hype, but should concentrate on consolidating the data and technical foundation. For robot suppliers, they need to expand the function of low-power GenAI in robots through various model optimization techniques, emphasize real-time control and performance, and be better at taking advantage of the integration of computing and connection. For robot users, the development of GenAI models in specific fields and strict examination of ethics, security, safety and performance will greatly promote the popularization of robots adopting GenAI. "
(Editor: Song Zheng HN002)