Video intelligent analysis system based on deep learning has powerful background analysis and computing ability, which can realize the recognition and early warning of abnormal events in the scene.BackgroundIn the steel industry, the finished steel coil is stored and logistics according to different grade standards. At present, the storage and logistics of domestic steel coil mills are basically carried out manually, and there are problems such as artificial wrong transportation and wrong placement, which have caused great losses to steel coil mills.SolutionsUsing artificial intelligence algorithm and real-time image analysis technology, Northern Lights Technology Development provides enterprise with AI video monitoring system.The AI video monitoring system can frame the moving area of workers or transport devices, and can monitor whether the steel coil has been delivered to the correct position in real time. If the steel coil in the warehouse exceeds the limited activity area, the monitoring end will be alerted by a red warning device.Advantagedentification and tracking accuracy：more than 90% Single frame monitoring coil quantity：20 / frameSingle frame recogni􀆟on speed：<=200 millisecondSingle frame tracking results return 􀆟me：<=200 millisecondThe on-line use of the system has greatly reduced the loss of human beings and brought real benefits to the steel coil mill.AI video monitoring system is a set of stable and compatible system, which can be widely used in aerospace industry, machinery manufacturing industry, 3C electronics industry, automobile industry and other fields.

SituationTechnology used was obsolete and required high maintenance personnel.Official websites had incomplete functions and poor user experience.SolutionUI update : Optimized web page access layout. Integrated ticket preferential information into Core Position.Optimize online ticketing process : Optimized the main functions of air ticket booking, on-line check-in, etc. Added new service modules.Front and back end separation : Sped up the overall response time of their website. Reduced the server concurrent load pressure.BenefitsAdjustment of website structure and other aspects to bring users a faster ticket-buying experience.Intelligent online services that were both active and personalized.

BackgroundInternal existing basic data platform, including companies, involving ten business data from a variety of related systems, but there are differences between the various data format standard, data quality is uneven needs a large number of data cleaning, cause of ETL scripts without version control, operation resources coordination difficulties, and execution of the script to human monitoring and a series of practical problemsThrough analysis, based on the existing big data platform, we integrate the data collection scheduling management, metadata management system and data quality management system in the governance system products of our data center to achieve business objectives.ResultsThrough the joint application of the three data center system products, script creation, control and warning are all automated, so that part of human resources can be released, and the overall degree of automation is estimated to be 52%In addition, the data quality of various data sources has also been greatly improved, saving a lot of labor costs. The overall cost savings is estimated to be 50%Business data relationships can also be tracked from untraceable to traceable, resulting in a significant improvement in work efficiency, with the overall efficiency improvement rate estimated to be 61%