On June 10, at the “Third New Business Civilization Forum – the 22nd Anniversary Annual Meeting of Business School Magazine and the Business Co-Creation Day of the National School of Development at Peking University”, Qin Yusen, vice president of the Digua Robot Software Platform, analyzed the changes taking place in the robot industry.
When large language models reconstruct the underlying logic of artificial intelligence, when sensor costs plummet, and when the employment gap in industry, warehousing, homes and other scenes continues to expand, the robot industry has finally emerged from its long dormant period.
As a veteran of the robotics industry for more than ten years, Qin Yusen reviewed the lag in industrial development in the past ten years: domestic artificial intelligence-related majors have been running for more than ten years, while the robotics engineering major has been established for less than three years, and the subject talent supply system seriously lags behind the speed of industrial expansion.
The long talent training cycle and extremely high cost of interdisciplinary learning in the industry have given rise to a consensus within the industry. After experiencing inefficient growth, Digua Robot decided to devote itself to the construction of general development infrastructure and create standardized development tools and modular technology components, which is what the industry calls “scaffolding” and general technology bases.
This system encapsulates and reuses the underlying technology modules that have been repeatedly constructed in the past R&D, so that the new generation of R&D personnel does not need to build a basic framework from scratch and reinvent the wheel, and reduces ineffective R&D internal consumption from the root. This is also the core reason for the birth and popularization of the industrial underlying tool platform.
Qin Yusen pointed out in his speech that we are currently facing three major disruptive changes.
First, the large language model reconstructs the development paradigm of general artificial intelligence, bringing a fundamental leap to the intelligence capabilities of the entire industry; second, the cost of various sensors has plummeted, and this industry dividend comes from the large-scale development of autonomous driving and new energy vehicle tracks. The combined demand for implementation scenarios in various industries has exploded, jointly reshaping the development foundation of the robot industry; third, the demand for scenarios has exploded significantly.
Qin Yusen believes that the explosion of scene demand can be compared to the development history of the home appliance industry.
Whenever new automated tools become popular, the public will always assume that the output quality of manual work is better than that of standardized machine work based on their inherent life experience.
Today, similar cognitive biases still exist, which also proves that the popularization of automated and standardized tools often requires a process of gradual adaptation of public cognition.
When the installed scale of the robot industry moves from the 10,000-unit level to the 100,000-unit level, the real gap in the industry is not hardware production capacity, but complete industrial infrastructure.
Qin Yusen said that the development of the real economy has always followed the underlying logic of “if you want to get rich, build roads first.” From the early years when the Internet was popularized and connected to thousands of homes, to the interconnection and systematic industrial infrastructure of massive robots, it is the support for the current large-scale development of robots.
Based on this understanding, Digua Robots focuses on two core directions to develop industrial infrastructure: first, to build a global intelligent computing power system for robots, covering all levels of needs from lightweight basic computing power to high-end high-performance computing power, giving each robot an independent intelligent decision-making “brain”; second, to improve the ecosystem of exclusive research and development tools for the robot industry.
In this way, Digua Robot makes up for the shortcomings of industry tools, builds a standardized and lightweight R&D support system, and empowers practitioners to carry out R&D work efficiently.
In order to lower the threshold for industry innovation and activate public R&D vitality, Digua Robot has launched a number of inclusive industry initiatives.
At the service ecological level, Digua Robot has built a hierarchical and full-coverage cloud infrastructure service system.
At the same time, Qin Yusen pointed out that the company is clearly aware of the objective laws of industrial development: every time the industry breaks through a core bottleneck, it will face new challenges that are more diverse and more segmented.
The current robot industry still faces multiple development bottlenecks such as imperfect data systems, limited model iteration speed, lack of industry standards, and high implementation costs.
But from an engineer’s perspective, the existence of bottlenecks is the core signal that the industry is maturing, and clear problem boundaries are precisely the source of motivation for technology iteration and industrial upgrading.
Compared with blind development without goals, facing up to existing shortcomings and accurately overcoming industry problems is the core path to sustainable development of the industry.
What is clear is that the perfection of industrial infrastructure will directly determine the development pattern and competition level of the robot industry in the second half.
Currently, China’s robot industry is in a stage of explosive growth. There are a large number of structural gaps that have not yet been filled in the industrial system, which requires the collaboration of various market entities and the division of labor.
Qin Yusen said: “One of our dreams is to allow a few people to make robots in the future. Build the industry base and let the entire industry grow on our ecology.”
