Go deep into the core scenario and let the enterprise see the digital effect first
For industrial enterprises to carry out digital transformation, it is emphasized that “input-output account” should be calculated first. Baidu intelligent cloud also emphasizes that it focuses on the core scene, so that the effectiveness of digitalization can be seen and counted by enterprises.
In the PCB printed circuit board industry, quality inspection has always been a difficulty. The traditional visual inspection method is inefficient, the accuracy of the machine recognition method is not high enough, and the false point rate is as high as 70%, that is to say, 70% of the products that are judged unqualified may be qualified, and it needs to be manually rejudged. At present, the intelligent quality inspection of Baidu intelligent cloud can filter out 80% of false points, greatly improve the efficiency of inspection, and probably save more than 2 million yuan in labor costs a year.
Today, Baidu intelligent cloud intelligent quality inspection solution, which allows enterprises to experience the value of AI, has been used in Changshu local mutual electronics, Yangxuan Electronics, Shengyi technology and many other enterprises.
Based on customer feedback, the products provided by Baidu Intelligent Cloud continue to upgrade, expanding from solving a single scene pain point to more key scenarios in the system. Evolve from intelligent solutions to system-level complete solutions that include intelligent platforms, cloud bases, and AI infrastructure.
Born in 1930, Hailida is an old auto parts manufacturing enterprise located in Changshu, ranking the top three in the industry in many fields. But it has its own troubles, “Our production line is producing a lot of data every day, but these data are ‘sleeping’, which is a big loss for the enterprise.” We need to find partners to help put the data to use.” Now, the industrial data intelligent platform of Baidu Intelligent Cloud can track the raw materials, production and test data of each product after landing in Hailida, and then through data analysis and knowledge extraction, turn the experience of “old master” into a model and apply it on the production line, greatly improving the yield. Knowledge and data from individual products to the entire production process can be intelligently managed.
Lower the threshold for enterprises to acquire AI capabilities and accelerate replication
In many fields such as quality inspection, safety, energy consumption optimization, Baidu intelligent cloud has landed a series of benchmark cases, and the next more critical step is to achieve faster and better replication of cases, so that the effectiveness of the scene continues to expand, and build a ‘jump level’ digital capability.
For how to achieve faster and better replication? Baidu vice president Li Shuo said, “First of all, relying on Baidu’s AI large base, you can expand from the intelligence of a single scene to multiple scenes. Secondly, the development of large model technology enables Baidu Intelligent Cloud to develop models with better generalization with less data, less computing power and a shorter cycle to cover new scenarios and new fields at a lower cost; Finally, end-to-end closed-loop optimization improves the value space, through algorithm upgrading, computing power pooling, chip optimization, converging into a better user experience and richer industry value.”