People who are involved in production control and production technology in the manufacturing industry may be hearing the terms “smart factory” more and more these days. The concept of this smart factory is quite broad, and there are cases where the meanings of each position are different. In addition, there is a movement by the government to take the lead in making the industry smarter. Here, we will introduce the general definition of smart factories, the problems facing the current manufacturing industry, the government’s roadmap for smartening, specific examples of roadmaps, and examples of on-site introduction.
What is a smart factory?
A smart factory is an advanced factory that implements Industry 4.0 advocated by the German government. Specifically, we will introduce robots, artificial intelligence (AI), IoT (Internet of Things), etc. to unmanned, and improve productivity and quality. In other words, the aim is to further promote the automation of existing factories and build more sophisticated production systems.
Current manufacturing challenges and smart factory roadmap
This smart factory is considered to have a great impact on Japanese industry, and the Ministry of Economy, Trade, and Industry also compiled a survey report called “Manufacturing Smart Roadmap Survey” in 2017. In this roadmap, we show current issues such as quality improvement, cost reduction, productivity improvement, shortening of the time period, shortage of human resources, improvement of added value / provided value, risk management/traceability, and their solutions. It describes how manufacturing should be 20 to 30 years from now and the issues that lead to it. In addition, the following three-step solutions to these challenges are presented.
Data collection / accumulation
By extracting and visualizing useful information, it is to turn the obtained awareness into know-how and knowledge. For example, the idea is to collect a large amount of data by automating the inspection and organizing it with graphs and diagrams.
By analyzing and learning a huge amount of information, it is possible to identify the factors of the phenomenon, model the phenomenon, and make future predictions. This includes analyzing data by factor analysis, statistical analysis or machine learning by artificial intelligence.
Data control / optimization
It is to make the best judgment and execution based on the analysis result and the prediction result. One example is to improve the defects found by analysis.
In this way, we can see that “data” is very important for the realization of smart factories. In the previous “automation”, only numerical data such as dimensions and weight were used, but “data” here is becoming broader. It can be said that it is a quantification of “things occurring in the production system” such as images, sounds, and movements of people, as well as numerical data. It is thanks to the tremendous development of IT equipment in recent years that it has become possible to handle such data. It can be said that smart factories can be realized by the breakthrough of IT technology.
A specific example of a roadmap
Next, let’s take a closer look at the above three steps. For example, if the issue of the production system is “improvement of quality”, “reduction of defect rate” is necessary to improve quality. To do this, first, we collect data by sensing operator mistakes and processing defects. This is the stage of “data collection/accumulation”. Next, in the stage of “data analysis/prediction”, we analyze past mistakes and processing defects and identify processes where mistakes and processing defects are likely to occur. Next, we will educate workers who have made mistakes and make design changes to avoid processes where processing defects have occurred. This is the stage of “data control/optimization”.
Specific examples of efforts at the field level
What kind of efforts is being made in the field to realize such a smart factory?
At a machining factory, machine tools were connected via a network to provide a function to check the machining status and notify the end of machining. This is the so-called IoT. This has made it possible for one person to be in charge of multiple machine tools.
In addition, the processing time can be acquired and stored as data. This corresponds to “collecting and accumulating data”. Then, by analyzing the acquired data, it became possible to optimize the machining time. You can see that “data analysis/prediction” and “data control/optimization” are performed. By repeating these three steps, we were able to optimize processing conditions and processing time on a daily basis, improving productivity and reducing manufacturing costs.
Similarly, a sensor was attached to the machine tool (here, the grinding machine that performs polishing) so that the polishing force applied during polishing can be measured, acquired, and stored as data. Again, by analyzing the accumulated data, it became possible to optimize the processing conditions. You can see that smartening is successful by including the same steps as we saw earlier.
These field-level efforts are an extension of existing field-level cost reduction activities, but the ultimate aim of smart factories is to promote intelligence and automation in the entire product supply chain. There is in that. When the entire product supply chain becomes smarter, the manufacturing floor will become much more efficient and data-utilized. However, it should be noted here that the Japanese manufacturing industry is supported not only by large companies with many factories and large supply chains but also by SMEs and town factories. Therefore, for the development of the Japanese manufacturing industry, it is essential to involve these small and medium-sized enterprises and town factories.
Smart factory and edge computing
As mentioned above, data is very important in smart factories. It can be said that data collection is one of the keys to smartening. However, when it comes to realizing a smart factory that covers the entire supply chain, the amount of data is enormous. In addition, it is also necessary to process these speedily. Against this background, cloud computing is indispensable for data collection, analysis, and storage, but it may be difficult to connect the factory floor to the cloud. In such cases, introduce edge computing where you can analyze and process the data smoothly.
The concept of a smart factory is wide and deep
This time, I introduced a small part of the concept of the smart factory. However, the concept of smart factories does not stop there. As already mentioned, the concept of smart factories is the idea of smart factories, which aim to solve and improve the problems mentioned in the text. So to speak, all the technical efforts to evolve the factory can be said to be a smart factory.