As the consumer side moves toward self-contained consumption and the manufacturing side faces a shortage of labor problems, the manufacturing industry must have the ability to adapt to the rapid and diverse and diverse environment, and the manufacturing system becomes more complicated than in the past. Thanks to the mature development of new technologies, the manufacturing industry can now improve the visualization of information and the controllability of the system by deploying advanced sensing technology and combining AI algorithms and introducing robots to further enhance the development of smart manufacturing in Industry 4.0. .
According to estimates by TrendForce's Tuoba Industrial Research Institute, the market for smart manufacturing in the world will reach 370 billion US dollars in 2022, with a compound annual growth rate of 10.7%. Based on the foundation of virtual and real integration, smart manufacturing is quite diverse in the application side, from large-scale smart factories, smart supply chains, on-site disaster recovery, to automated logistics vehicles and simple robots.
Looking at the industrial dynamics of this year (2019) and the index activities of Hannover Messe, the current smart manufacturing is Cobot, Digital Twin, Predictive Maintenance (PdM), drones, Manufacturing Execution Systems (MES), AI applications, etc. are the focus of development, and Universal Robots, Siemens, STMicroelectronics, Xilinx, GE and other vendors continue to introduce new and enhanced layouts. Edge AI reduces computing resources and becomes an important foundation for predictive maintenance. Tuoba pointed out that in view of the huge amount of data brought by smart manufacturing, the delay and bandwidth costs have gradually shifted the manufacturing industry from cloud technology to edge computing.
The three driving forces of data quantification, analysis precision and hardware efficiency also make AI move from the cloud to the terminal equipment, pushing the trend of the edge combined with AI. Tuo pointed out that Edge edge processing is a geographically related AI operation. It is collected and processed close to the data generation source, and combined with parameter learning and other AI techniques, the device can achieve the purpose of defect detection and use status prediction. So that the machine does not need to be connected from time to time, reduce computing resources, still have some decision-making power and instant response, and become an important basis for predictive maintenance. It can also strengthen the real-time collaboration of industrial robots; The cloud is also more able to meet the needs of the manufacturing industry to improve data security and privacy.
The chip will be the best focus for the Taiwanese company to cut into Edge AI, and the flexibility of SMEs will become a competitive advantage. Tuoba pointed out that the connection between smart manufacturing and Edge AI brings immediate advantages such as immediate decision-making, cost reduction, reliable operation and improved security to the manufacturing industry. It also turns precision machinery into a veritable intelligent system. Currently, from the chip manufacturers NVIDIA, Intel, Qualcomm, NXP, and even the cloud leader AWS, Google, Microsoft, etc. are actively involved in this field. If Taiwanese companies want to cut into the Edge AI market, considering the industrial advantages and government resources, the chip is still the best place to play, and become the engine of the upstream and downstream manufacturers.
From automation to intelligence, TrendForce pointed out that the industry 4.0 wave continues to promote the digital transformation of enterprises, and technologies such as the Internet of Things, big data, and robots have become important nodes for building smart manufacturing. However, whether it is the construction of industrial Internet of Things, the introduction of smart manufacturing, or the establishment of smart factories, it is not a long-term solution for enterprises, because it is time-consuming and costly. It can be seen during deployment and implementation. Tools developed by organizations such as the Industrial Internet of Things (IIC) to assess their maturity, and then adjust the pace and direction, such as the choice of passive maintenance, preventive maintenance, and predictive maintenance based on the degree of infrastructure completion.
In addition, TrendForce further pointed out that because many non-digital native manufacturing industries implement smart manufacturing through digital tool import and cross-industry integration, if the company has the advantages of cross-domain integration, more agile and flexible, and the ecosystem. It will be easier to cut into or cooperate with the big supply chain. Taiwanese SMEs have sufficient industrial knowledge and adaptability. When the digital age unearths more customers' pain points, the big companies are reluctant to do so and the small factories can't do it.