On November 23, the “2021 Artificial Intelligence and Robot Application Summit Forum” with the theme of “Intelligent Manufacturing Leading Innovative Development” opened in Guangzhou.
Dr. Li Nan, Gechuang Dongzhi, was invited to attend the conference and delivered a keynote speech, “Use of digital and artificial intelligence technology to help the transformation and upgrading of manufacturing industry”. He collided with academician Li Bacon, academician Luo Xiwen, and Dr. Huang Pei, chief editor of e-works, and other people on the same stage to provide new ideas, new ideas, and new models for manufacturing enterprises to use artificial intelligence technology to build smart factories.
AI helps enterprises to build intelligent
“Yes and No”
At present, in the industrial field, artificial intelligence cannot be 100% handed over to computers and deep learning models, so that it can output high-quality information about quality prediction, equipment management and other related information without professional knowledge or industrial model input. Artificial intelligence in the industrial field still needs to be based on production data, industry experience and “human-machine cooperation mechanism” to realize intelligent optimization decision-making of data-driven manufacturing process and bring data value into play.
The essence of enterprise intelligence is the gradual iterative upgrading of the six steps from recording, connecting, visualizing, analyzing, forecasting to adaptive, but with different scales and ranges (from equipment to workshop to factory). At present, the intelligent development of most companies still stays in the step of visualization, and very few companies can develop to data analysis and reverse prediction guidance.
The core reason for this phenomenon is that the application of AI technology in the manufacturing industry still faces many challenges, which can be summarized into four points: weak data base, low application sensitivity, complex data, lack of professional tools and talents.
Benchmarking case
Provide AI technology application template
Taking Gechuang Dongzhi as an example to build an intelligent factory for a high-tech manufacturing enterprise, according to the overall development thinking – three modernizations and four steps (three modernizations: automation, data, and intelligence; four steps: automation, IoT, big data, and intelligent manufacturing), gradually implement multiple key applications to promote the construction of intelligent factories.
01 Industrial Internet of Things platform to build a solid data base
The manufacturing enterprise has the characteristics of high automation, high tempo, almost reaching the limit of yield efficiency, and huge downtime loss, so it must effectively play the value of data. The industrial Internet of Things platform built by Gechuang Dongzhi can collect all the production data that could not be collected in the past into the platform, and use these data to do equipment predictive maintenance, production environment monitoring and other applications to improve the value of data utilization.
02 Data center support upper application
After the data collected by the industrial Internet of Things platform, the data is analyzed through the data center to support the upper applications such as production analysis platform, operation analysis platform, intelligent analysis platform, etc
03 Key applications are launched in succession
Gechuang Dongzhi has created multiple intelligent applications for the enterprise and gradually formed a 360-degree quality monitoring management system: industrial big data applications (such as Dongzhi’s multi-factor analysis and modeling tool MFA, etc.) improve the ultimate yield; Visual detection applications (such as Dongzhi visual detection system, ADC system, etc.) establish independent learning models to achieve uninterrupted and accurate AI image recognition function; The equipment predictive maintenance application (such as Dongzhi Equipment Health Management EHM) detects equipment abnormalities in advance to reduce the loss of equipment abnormal shutdown;
Energy management applications (such as Smart Energy Management EMS) provide a comprehensive energy service platform to reduce energy consumption and improve energy utilization
......
* Dr. Huang Pei, chief editor of E-works, interacted with Dr. Li Nan on the same stage
AI helps the construction of intelligent factories
“A brilliant plan”
How to make use of AI technology to better promote the construction of smart factories, Gechuang Dongzhi summarized the following four experiences:
01 Focus on core business scenarios
The “KPI” of manufacturing enterprises can be summarized into four key words of QCDI (Q – quality; C – cost; D – delivery time; I – innovation). Enterprises need to determine which link is the most important pain point in the construction of their own intelligent factory, and then “teach students according to their aptitude” to solve the pain point through AI technology.
02 Organizational and cultural change
As mentioned in the previous sharing, we suggest that enterprises must establish an intelligent manufacturing promotion department, whether virtual or physical, and select 1-2 core key users from all relevant departments such as IT, quality production, supply chain logistics and so on, and put them into the promotion department. These people will jointly formulate the enterprise’s intelligent manufacturing promotion strategy for the next 3-5 years. Such division of labor can better ensure the promotion and replication of intelligent manufacturing.
In terms of the way of TCL transformation, TCL has incubated an independent industrial Internet company, Gechuang Dongzhi, and formed an “iron triangle” with its internal digital committee and its industries, which is also the most important reason why enterprises can successfully carry out intelligent transformation and achieve great results.
03 Enable front-line production engineer
In the final analysis, the construction of smart factories is to empower front-line production engineers to understand data and equipment even if they do not have rich IT experience and knowledge. They can respond quickly or predict in advance to solve problems at the first time of production process or equipment failure, even before failure. This is where the construction of smart factories can most generate business value.
04 Professional technical platform support
By providing professional tools for front-line production engineers, data aggregation, data visualization and data value mining can be realized, and through this process, artificial experience can be internalized to finally form a model base, which is precipitated on the professional technology platform, and then the transformation from decision-making relying on human experience to intelligent decision-making can be realized.