1. Equipment data accounts for only a small percentage of factory data! ?

In order to realize a smart factory, it is an essential requirement to widely collect factory production site data.
The first target for data collection is facility data such as PLCs and signal lights. In fact, many production departments have experience collecting and utilizing data from PLCs and signal lights.

However, what percentage of the data that should be collected in the entire factory is the data that can be collected from equipment such as PLCs and signal lights?
At production sites, there are production lines centered on equipment and production lines centered on workers. And even though it's basically a facility, there are many old facilities that are difficult to collect data in the first place, such as those that do not support Ethernet communication. In addition, there may be some production lines that are being considered for automation using robots and conveyors for worker-centered production lines. However, in recent years, as there is a need to respond to a wide variety of consumer demands, maintaining highly flexible worker-oriented lines may lead to a source of strength in the manufacturing industry.

With such a wide variety of production sites, it is extremely important to take a bird's eye view of the factory and consider the reality of data collection.
In this article, we will identify the characteristics of factory data and summarize the points that we would like you to keep in mind when collecting it.

2. Classification of factory data

What we want to achieve with a smart factory is to optimize the entire factory from the perspective of productivity and profitability while understanding a wide variety of production sites.
​For that reason, it is important to establish a common way of thinking and collect data according to that standard, even if the subject (equipment / worker) in manufacturing and the type of equipment (new / old) are different. . By standardizing and accumulating data using that data as input, it becomes possible to utilize data for highly accurate analysis and prediction that captures a bird's-eye view of the entire factory.

Here, I would like to classify the factory data according to its characteristics.
As mentioned above, it is possible to classify production lines based on equipment and workers.
In the case of facilities, data can be collected from PLCs and signal lights for new facilities. In the case of old equipment, methods such as collecting machine signals by modifying the equipment or attaching retrofitted sensors to collect data are conceivable, but data collection is more difficult.
In the case of worker-centered data collection, there is no data source in the first place, so the worker will need to do something such as input the situation, and the burden on the manufacturing site due to data collection will increase.

Relationship between ease of factory data collection and work load

3. Data collection approach

What should we be aware of when collecting data from production sites?
First, equipment data collection such as PLCs and signal lights can be realized for multiple production lines for several hundred thousand yen by using tools such as master PLCs, loggers, and IoT gateways.
Then, how should data be collected for old equipment and worker-oriented production lines?

Let's summarize the data collection approach for old equipment and worker-driven manufacturing lines.

• Refurbish equipment and collect analog signal data from old machines
• Install new sensors and collect measurement data​
• Arrange switches and cards at the production site and collect data by operator's operation​
• Collect data from data entered on a tablet (electronic report software, etc.)
• Work movement is captured with a camera, and ​AI analysis of the video is used to collect data for work situation judgment.
• Have workers hold a beacon (communication terminal) and collect data for work status judgment from position information​

4. Reliability and costs to be aware of in data collection

As mentioned above, among the various approaches, the points to be aware of when collecting factory data are "highly credible data" and "data collection costs."
Accurate analysis requires highly accurate input data. Relatively new equipment data is the data of the production machine itself, so the possibility of including false data is low and it can be said that the data is highly credible. On the other hand, collection of old equipment and worker-oriented production lines requires ingenuity. As a result, it often results in uncertain data that conforms to forcedly determined criteria, and there is no point in collecting it. Collecting highly credible data is a point that you should always be aware of when starting data collection.
The next issue is the cost of data collection. The cases where large budgets are allowed for data collection are often in the pilot phase, which also serves as a survey of the latest technology. After passing the test, when it comes time to introduce it to the entire factory and collect factory data, return on investment is required. Investments exceeding several hundred thousand yen for one line may not be accepted, and many people stop there.
When considering the collection of factory data, it is important to proceed with consideration realistically while keeping in mind the credibility and cost of the data.

Supporting quick start & quick win of smart factory

"DSF Cyclone" connects manufacturing results and production plans with structured data to realize a "highly productive factory".

If you are interested, please refer here.

Related article

What is the role of the production department in the smart factory promotion project?

What is the role of the information system department in the smart factory promotion project?

Reasons why DX in the manufacturing industry is not progressing ~Characteristics and specific examples of organizations where smart factories succeed~

What is factory KPI management? -Understanding the degree of achievement of goals-

Sophistication of improvement activities at production sites and future DX ~Things to be aware of when using digital technology~