From August 31st to September 5th, the 2022 China International Fair for Trade in Service, jointly hosted by the Ministry of Commerce and the Beijing Municipal People's Government, was held in Beijing. 507 Fortune Global 500 and industry leading enterprises from around the world participated offline and achieved 1339 various results. The Fair is an important platform for China to expand opening up, deepen cooperation, and lead innovation, injecting a "booster" into promoting China's high-level opening up and promoting global economic recovery under the pandemic.

As one of the special forums, the World Artificial Intelligence Integration Development and Security Application Forum, with the theme of "Smart Integration Security Escort", will be held simultaneously. Wang Zhonglin, Vice President of Minivision Technology, delivered a keynote speech titled "Exploring the Value of Smart Vision in the Field of Coal Coking".

AI+coal coking, ensuring safe production

Industrial digitization and digital industrialization are important goals during the 14th Five Year Plan period of the country. As an important component of industrial production, the digital upgrading of coal coking enterprises plays an important role in ensuring safe production and accelerating the process of industrial modernization in China.

All along, coal coking enterprises have faced the following safety production challenges: the safety production awareness of frontline workers needs to be improved; The safety supervision in production operations needs to be strengthened; The intelligent application of video surveillance, especially the application of intelligent identification in safety production risk prevention and safety review, needs to be improved.

The current artificial intelligence technology can not only adapt to coal coking production scenarios, with complete infrastructure and relatively sufficient computing resources, but also greatly improve research and development efficiency through small sample weak supervision technology, laying the foundation for AI implementation in coal coking scenarios.

Wang Zhonglin stated that from the perspectives of industry background, production demand, regulatory demand, and feasibility of technology implementation, the application of artificial intelligence technology in coal coking scenarios has high feasibility.

Exploring the "2+1" Industrial Value of Minivision

Minivision is value oriented and proposes a "2+1" model for industrial landing, which is "two frontiers+one transformation". By putting data annotation and algorithm training in front, the artificial intelligence expert team is put forward, focusing on the user value of the scene, moving data annotation and algorithm training from the laboratory to the production environment, and moving the artificial intelligence expert team to the production line, allowing technology and industry rules to be deeply integrated with the needs of the scene. By transforming "selling AI products to users" into "providing AI services to users", we aim to enhance the practical effectiveness of AI technology in scenarios.

Guided by the "2+1" model, Minivision attempts to apply the artificial intelligence visual center platform to safety production. The platform has functions such as intelligent recognition of video objects and their behaviors, active push of alarms for special events, etc., transforming the past "finding people to find things" into "finding people to find things", awakening a large number of dormant video monitoring resources. By linking scenarios with technology, strengthening platforms, improving applications, implementing safety standards, and supervising production operations, enterprises can achieve essential safety in production, reducing costs and increasing efficiency.

Focusing on the two core elements of "human" and "environment" in the production operations of coal coking enterprises, the artificial intelligence visual center platform can achieve the management of the identity and production behavior of special employees, such as personnel monitoring, key personnel deployment, intelligent point inspection, trajectory analysis and judgment, and identification of various violations.

At the same time, the platform can also manage special operations, hazardous chemicals areas, and mechanical equipment, such as compliance supervision of special operations, identification of illegal operations and fireworks in major hazard source areas, detection of foreign objects on conveyor belts, tear detection of conveyor belts, and coal loading detection. Wang Zhonglin believes that in the coal production environment, intelligent visual technology makes the essential safety of production manageable and controllable, fully leveraging the user value of AI landing scenarios, and playing a role in improving efficiency and reducing risks.

Based on intrinsic safety, Minivision utilizes intelligent vision to enhance the efficiency of humans and machines in scenarios, explore their value in the production environment, and generate quantifiable social and economic benefits in the upgrading process of the digital and chemical industry in the coal coking industry.

Unconsciously, AI is quietly promoting a change in production methods.