NetEase Fuxi launches smart robot beta version to help intelligent upgrade of the industry

According to the seventh national census data, my country's population aged 60 and above exceeds 260 million, accounting for 18.70%. The data shows that the degree of population aging will further deepen. As the demographic dividend disappears, all walks of life will continue to face pressures such as labor shortages and high labor costs. Artificial intelligence (AI) and robots are gradually changing our lives, but the in-depth implementation of the technology is still in its infancy. Letting AI and robots handle more repetitive and high-risk tasks and increase the value of human resources is the key to the intelligent development of today's society. The inevitable trend.

According to the "China Robot Industry Development Report (2022)", it is expected that by the end of 2022, the global robot market will reach US$51.3 billion, of which China will reach approximately US$17.4 billion. However, there is greater demand behind this estimated market size. However, due to the complexity of actual business, AI and robot products often cannot be effectively applied in relevant scenarios.

With the continuous development of the Internet of Things, industrial 5G and virtual-real integration technology, artificial intelligence technology will be integrated and upgraded with them, giving full play to the collaborative advantages of humans and machines, solving problems and integrating AI capabilities and innovative technologies applied in real-life scenarios such as robots, games, and the metaverse. This will improve application development efficiency and reduce enterprise labor costs.

From virtual to real, oriented to intelligent agent programming, to build a new solution for unified task modeling

From the current development stage, artificial intelligence It is already easy to solve local simple problems, but complex scenarios still require a large amount of data or high-quality simulation environment support. How to apply AI's cognitive intelligence and decision-making intelligence to a wider range of physical fields, rely on autonomous learning capabilities to collaborate with humans, and even imitate humans to engage in more intellectual and creative labor is a research topic in the field of artificial intelligence and the robotics industry. important direction. Based on the current situation that the degree of machine intelligence is generally low, by building a "human in the loop" (HITL) method, the concept of human-machine collaboration is used to help enterprises solve the problem of high AI thresholds, long R&D cycles, and application implementation. Difficulty and other practical problems have become a feasible way to solve the pain points of the above-mentioned AI industry.

NetEase Fuxi was founded in 2017 and is the top institution in China focusing on the research and application of AI in games and pan-entertainment. They have rich technical accumulation in AI fields such as digital twins, reinforcement learning, user portraits, and natural language processing. In addition, they are also committed to applying virtual world technology to real economic fields such as intelligent manufacturing, construction machinery and service industries. In this context, Fuxi Youling Robot came into being.

Different from traditional PaaS platforms, the decision-making of "humans" in intelligent robots is crucial. The platform issues tasks through the crowdsourcing model and quickly calls on "human capabilities." Crowdsourcing users complete tasks online through the platform, solving the problem of high costs in collecting and annotating data. The platform will also actively learn and implement closed-loop data, feeding real task manual operation data back to the AI ​​algorithm to realize task preprocessing, which greatly improves the efficiency of task completion; crowdsourcing users can also use the "online order taking and remote work" mode to solve construction problems in high-risk and harsh environment scenarios and achieve delivery without leaving home.

In addition to mining robot operation scenarios, the crowdsourcing platform also supports AI enterprise user demand scenarios such as game AI, AR/VR, and Metaverse. Users with data collection and labeling needs can publish tasks through the crowdsourcing platform. The platform covers 80% of mainstream data labeling scenarios in the AI ​​field, and continues to reduce the number of manual labeling through automatic learning, with automatic labeling accounting for up to 20%.

For business scenarios such as robots, the platform also provides end-side management capabilities, allowing users to write end-side business logic in a platform-based manner, combined with one-click device initialization, continuous integration release, OTA and other functions truly achieve integrated development of cloud, edge and device, allowing users to focus only on business.

Open ecosystem and build a new era of intelligent robots

In the next few decades, human-machine collaboration will be the key to the intelligent development of society The main theme is that machine intelligence requires human support, and human intelligence requires machine assistance. At the root node connecting the intelligent agent to the world, there will certainly be no shortage of humans, but the role may continue to change from task executor to organizer, manager, decision-maker and discoverer. On the basis of providing standardized task modeling tools, NetEase Youling Robot hopes to serve as a bridge between small and medium-sized enterprises and scattered workers in the society at large, providing platform support for task modeling, release and execution.

The implementation of NetEase Fuxi's excavating robot in the construction of key infrastructure projects in southwest China has also verified the strong implementation, sustainability, and replicability of the human-machine collaboration platform. On the one hand, the platform allows traditional excavators to achieve intelligent production through human-machine collaboration. On the other hand, through online distribution of crowdsourcing tasks, traditional excavator masters no longer engage in repetitive, boring and dangerous work, further promoting the digital economy. Deeply integrate with the real economy to promote the quality and efficiency improvement of the real economy. In addition to exploring robot application scenarios, the platform has also been practiced and verified in scenarios such as game character face pinching, image and text recognition detection, smart parks, and art crowdsourcing.

In the future, intelligent robots will explore more scenarios that integrate reality and reality, allowing intelligent agents to better meet people’s needs for production and labor, and fully open up opportunities for cooperation and co-construction, and integrate with the ecosystem. Partners create a new era of human-machine collaboration.

The above is the detailed content of NetEase Fuxi launches smart robot beta version to help intelligent upgrade of the industry. For more information, please follow other related articles on the PHP Chinese website!