Open three advantages, AI quality inspection market share first for four consecutive years

“AI+ Industrial Internet” is the biggest feature of Baidu intelligent cloud open matter. After years of accumulation, Kaiwu platform focuses on core enterprise scenarios such as quality control, production safety, energy consumption management, process optimization, and intelligent scheduling, accumulating more than 40,000 industrial models and intelligent applications, realizing the standardization of intelligent solutions, and accumulating many well-known customers in the industry. For example, in view of the problem that the quality inspection false spot rate is too high in PCB printed circuit boards, Baidu Intelligent Cloud’s intelligent quality inspection solution can filter 80% of the false spot, which can probably save more than 2 million yuan in enterprise labor costs a year. The solution has now been used in a number of local enterprises such as mutual Electronics, Yangxuan Electronics and Shengyi Technology in Changshu, Suzhou. In the workshop of Hengyi Group, a chemical fiber enterprise, the AI quality inspection equipment of Kaiwu inspects a silk ingot in just 2.5 seconds, and the efficiency is 70% higher than that of manual inspection. At present, Baidu’s intelligent cloud industrial AI quality inspection solutions are deeply cultivated in 15 industries, ranking first in China’s market share for four consecutive years.

Baidu Intelligent Cloud ranked first in China’s AI quality inspection market share for four consecutive years

Baidu Intelligent Cloud also insists that the operation and services of the open object platform should be around the park and enterprises, so it has built an engineer team that can respond to customer needs in the first time to provide “personal” services for enterprises. For example, in the year of deep cultivation in Suzhou, Baidu intelligent cloud open material service 300 enterprises, help 20 enterprises to promote comprehensive digital transformation, and provide consulting and diagnostic services for 183 enterprises. At present, Baidu intelligent cloud Kaifu has been deeply landed in 18 regions or parks such as Suzhou, Guangzhou and Chongqing, and has built an “AI+ industrial Internet” industrial base in the local area.

In terms of ecological construction, Baidu Intelligent Cloud believes that the ecosystem in the era of large models will have three major characteristics: the location of the application ecology will be more core; Companies with industry expertise face great opportunities and challenges; Start-ups and developers will drive the adoption of generative AI. To this end, Baidu Intelligent Cloud launched a series of new large-scale model ecological partner programs in early September.

First, the development of 10 core comprehensive partners, to provide business opportunities worth more than 1 billion yuan, 10 million yuan of special fund support.

The second is to develop 100 core application software development partners.

Third, it plans to support 10,000 start-ups and land the “AI accelerator” project.

The fourth is to extensively support the transformation and upgrading of the agent partner system, and touch the large model service to hundreds of cities and thousands of counties.

First based on the reconstruction and upgrading of large models, the new object has three new functions

Baidu Intelligent Cloud believes that the large model will bring a new paradigm of intelligent production to enterprises. Employees no longer need to devote a lot of time to memorizing routine content, and they can quickly get high-quality answers by asking large models for questions. Big models will revolutionize the relationship between humans and machines. Employees only need to say a word, and the large model can automatically schedule and direct multiple systems to complete a large number of transactional tasks. More importantly, the powerful understanding and generation ability of large models can integrate knowledge from different fields to create unprecedented “new species of AI.” With a large model, innovation is no longer the preserve of geniuses, and every employee has the opportunity to achieve disruptive innovation.

For example, in the field of safety production, Baidu Intelligent Cloud’s safety production intelligent assistant “Du ‘an” helps enterprises upgrade their safety management from “perception intelligence” to “decision intelligence” with a large model. In addition to helping workers answer safety problems in the first time, it can also classify and summarize various safety warnings, so that managers can quickly know the overall safety situation and quickly deal with it. Daxie Petrochemical, a large refining and chemical enterprise under CNOOC, has introduced “Duan” to achieve second-level warning of safety hazards and increase accident response speed by 50%.

Baidu Intelligent Cloud is based on the one-stop enterprise platform “Qianfan” to provide large-scale model services and a full set of tool chains to all walks of life. Up to now, there are nearly 10,000 monthly active enterprises on Qianfan platform, covering more than 400 scenarios in finance, education, manufacturing, energy, government affairs, transportation and other industries.

In order to help industrial enterprises to use more quickly and have the ability to use a good large model, on September 5, 2023 Baidu Cloud Intelligence Conference, Baidu Intelligent Cloud announced the reconstruction of open objects based on the Wenxin large model. The new product will achieve three functions: first, it can use less data samples and less time to build intelligent applications such as quality control, safety production, and production scheduling; Second, it can help managers quickly obtain production and operation conditions, make decisions and implement them through natural language interaction; Third, it can assist local governments to make analysis and decision-making of “strong chain and complementary chain”, and promote the global optimization of the industrial chain.

This means that the new development will realize the leap from “production line intelligence” to “enterprise intelligence” and then to “industrial chain intelligence”.