Guide dogs are rigorously trained to obey the owner's orders, but their training process includes what is called "Intelligent disobedience" behavior, which is to violate the owner's instructions for better results. This behavior is the key to the success of guide dogs, but also applies to the Internet of Things (IoT). According to reports, blind people who walk along the same road every day are familiar with the roads and can understand the traffic and the crowd with a better sense of smell and hearing, and can make appropriate decisions. But there may be accidents in the corner that he has never encountered. A guide dog who sees something the owner can't see will be aware of the danger and violate the command of advancement. This behavior may also have the same advantages in the IoT field. Intelligent systems should be designed so that if there is a situation that is not in the best interest of the current work, the instructions should be postponed, especially in the case of danger. The sensor is the IoT's nervous system, collecting continuous data streams for processing. They provide data to computers that process faster than anyone. The information obtained through the IoT system should be more comprehensive than any operator. It may see the danger faster than the operator and believes that the necessary precautions need to be taken. When the operator instructs to do something, the system knows that it is too dangerous to do so. This is the best time to take smart disobedience and warn the operator that their instructions will have negative consequences under current conditions. Of course, sometimes the instructions must be followed despite the system's opposition. For example, when the blind man walks to the corner, the guide dog finds two unfamiliar men in front and prevents the owner from moving on. But when the other party opened, the blind person recognized his voice, and the guide blind dog indicated that the other party was a relative. As a result, smart disobedience will be overthrown. Similarly, the operator may know what the IoT system cannot know. Therefore, an upgrade procedure should be established to allow staff to challenge system requirements and continue with the original instructions. Red Hat developers worked with Cloudera and Eurotech to create an Industry 4.0 demo that includes a predictive maintenance application and provides this program history and real-time data for analysis, modeling, and machine learning. New business rules can be built on this data, and machine learning models can be executed at the edge to solve problems and react to unpredictable events. Models like this can be used to statistically analyze and predict when a machine may fail and when it will be repaired. As the system becomes more intelligent, future developments should be integrated into the process to make decisions based on its recommendations. Consider this feedback before making a decision to ensure better decisions. One way is through the virtual factory. Eclipse Kapua has such a digital simulation (Digital Twin). When a system is unavailable, its digital avatar will display the last known state. In this case, the user can query the operational status and request a measurement, or even send the change to the device. Another concept is to create a virtual copy of an engine and let users test the changes. With AI and machine learning, what happens when you make changes before deploying to a physical machine. However, the second technology has few products on the market. According to the report, in order to make the technology more widely used in the industry, open standards should be established to share information. To create the complex models you need, you should use software to simulate the standards of these devices. Promoting the standardization of the Digital Twin model is one of the potential efforts. Pcb Board,Circuit Board,Pcb Electronics,Custom Pcb Cixi Zhongyi Electronics Factory , https://www.cx-zhongyi.com