IoT/AI, the New Era of Technical Innovation

Nov. 15, 2019 [No.67-2019]

Mr. Yasuo Sando
Senior Chief Researcher
Nippon Steel Research Institute Corporation

IoT/AI, the New Era of Technical Innovation

For those who live in the current industry, it is no doubt that very few people have never seen or heard of the phrases "IoT" and "AI (Artificial Intelligence)." What are their main points? What are the background factors? They are both questions that are difficult to immediately answer.

Why digital?

The greatest background factor is found in technological innovation that is related to "digital." By means of a mechanism called IoT, people, things, and experiences can mutually, and easily connect to the net. By means of IoT, we can instantaneously see a clear picture of factory work sites in foreign countries. In the same way, we can know the timing of the change of parts from the machine operation status, even if the machines are in foreign countries. At the stage before a machine fails, we can even catch those signs. We can also decrease the backup stock of spare parts. Recently, instead of parts being obtained from far away, there have been examples of 3D printers being used and creating spare parts on the site. Instead of keeping physical stocks, we can just store CAD data in a virtual space that is somewhere on the net. Physical stock will be replaced with digital data that does not take up space. The burden of stock management will greatly decrease.

What do you do when there are requests for an expedited quote or delivery? You must urgently contact to other departments. You would be tremendously busy to consider and check the availability of production lines, the deadlines of work in progresses, the supplies of materials and parts, and the workforce.

What if we could instantaneously grasp this kind of situation with a list on a dashboard by means of IoT? One could immediately respond to quote and product requests. Speed that can instantaneously reach remote places is the greatest characteristic of "digital." Suppose a company consumes a lot of time inquiring with each related department. On the other hand, another can implement IoT and respond immediately. It's clear which corporation will be more competitive.

IoT and AI

IoT in a narrow sense means mechanisms and technologies to collects and stores necessary data by sensing. According to the types of issues corporations are having, it is decided what kind of data should be taken and how they can be taken. The sensor that is fit to collect the data you want is decided. Sensors are set in mechanical devices, equipment, tools, jigs, conveyance equipment, and more. The sensing result is converted to digital data and sent to a data storage (a data pool) and there stored and analyzed.

As for data produced in the mechanical industry and elsewhere, a significant amount of data is created each second. Noise is also included among that. Even types of data, when compared to data relating to SNS, are heterogeneous. The frequency of occurrence and data capacity are incomparable. These kinds of particular properties are found in IoT manufacturing.

A large amount of data that is collected and stored by IoT is called big data (abbreviated BD hereinafter). It is also sometimes called BD by including the analysis of BD itself. Either way, when it comes to analyzing various kinds of diversified, large data, suitable analysis technology is necessary. That kind of analysis technology is processed by means of computer software. One could also say that AI is the latest edition of such computer software.

Conventionally, computers couldn't calculate anything if people didn't write the program. However, computers are now finding answers on their own by learning themselves. The data they handle isn't just structured data, it can also process images or sounds. It's said that when AI studies a famous artist's picture, it can produce a similar picture. Machines that have eyes, ears, and even cognitive ability. Such an era is now coming upon us.

Unauthorized reproduction prohibited.