Pitfalls in the Conversion of Regular Factories into Smart Factories Seen in Data

do, 21/09/2023 - 21:05

Door: Redactie WINMAG Pro

A smart factory is a factory that achieves operational process improvements, higher quality, and lower costs based on digital data by connecting processes from design to development, manufacturing, and maintenance in a network. In the manufacturing industry, it refers to automating various processes in a factory to realize product quality and productivity improvements, a reduction in manufacturing costs, a shortening of the commercialization, mass production and delivery periods, and other areas of improvement. In other words, a smart factory is a factory working on digital transformation (DX) that aims to transform the entire business process from design to manufacturing, maintenance, and logistics in addition to improving productivity by automating factory equipment.

We introduce here the results of a questionnaire survey* we conducted with people involved in promoting the conversion of regular factories into smart factories in the manufacturing industry in Japan.

About the Questionnaire Survey

Organization conducting the survey: Murata Manufacturing Co., Ltd.
Survey targets: 11,084 workers in the manufacturing industry in Japan (screening survey) and 500 people working in the manufacturing industry who are involved in converting regular factories into smart factories in their companies (main survey)

Survey method: Internet survey
Survey period: Three days from January 25 (Wed.) to 27 (Fri.), 2023
*Total figures may not match due to rounding.

State of Progress in Converting Regular Factories into Smart Factories

We learned that over 38% of factories are in the process of converting into smart factories or considering converting into a smart factory in the questionnaire we conducted this time (Fig. 1). We can see a trend for even more companies to convert their regular factories into smart factories in the future.

Fig. 6: Initiatives in promoting the conversion of regular factories into smart factories

Clues to Reducing Short Stoppages

We need data on the operating status of equipment to reduce short stoppages. For that, we must be able to accurately and quickly record the equipment stoppage times due to the frequent short stoppages and their causes. We introduce here the m-FLIP solution tool that helps to resolve such issues.

What Is m-FLIP?

m-FLIP is a solution that combines Murata Manufacturing’s worksite improvement know-how and IT technologies to maximize the operating rate of manufacturing devices. This solution makes it possible to quantify factors such as operating rate, actual working rate, and equipment stoppages to increase production and achieve optimal maintenance. This greatly contributes to the visualization of problems that are difficult to reveal such as short stoppages.

Furthermore, m-FLIP calculates and outputs the operating rate and actual working rate based on data obtained from PLCs made by multiple manufacturers. In addition, it collects workers’ work information through touch-panel PCs. As a result, we can clarify the duration and number of equipment stoppages, which was previously done by guesswork. We are also able to clearly see the main causes of equipment stoppages and the issues that must be resolved (Fig. 7).

Fig. 7: Services provided by m-FLIP

m-FLIP Utilization Examples

m-FLIP can support the resolution of various issues at manufacturing worksites. We introduce here some typical utilization examples.

Equipment maintenance:

With checks by humans, it is only possible to monitor the operating status temporarily. However, m-FLIP can continuously monitor the operating status of equipment. It is possible to continuously collect data over several months or several years. This allows us to grasp the aging and changes over time of equipment. As a result, it enables maintenance at the optimal timing while minimizing the impact on the operating rate to the greatest possible extent.

Quality control:

For example, it is possible to monitor static electricity generation status based on data from temperature sensors and humidity sensors around the equipment and to output an alert when the set threshold is exceeded. We can utilize this in quality control such as to prevent the outflow of defective products. Moreover, we can also collect data including information on vibrations, current, and voltage. We can then integrate the data we have obtained from multiple sensors and output it as a single piece of information.

Work procedure review, equipment improvement, and personnel allocation optimization:

We can grasp the movements of workers and the operating status of equipment as quantified information. Therefore, we can also utilize it to consider work procedure reviews, equipment improvements, and personnel allocation optimization.

Preparation of reports:

m-FLIP comes with a function to prepare reports that can be used for reporting (Fig. 8). You can display information in charts and tables. It is also possible to input and save comments. You can share the reports you have prepared.

Fig. 8: Monthly report screen

After-sales support by professionals:

After introducing m-FLIP, we can identify issues from the data we have collected and receive suggestions for improvements and other advice from professionals familiar with production technologies, factory technologies, and other technologies.

m-FLIP Introduction Case Study (Example of Ogaki Murata Manufacturing)

We now introduce a case study from Ogaki Murata Manufacturing in which they resolved frequently occurring problems including short stoppages due to tray replacement in the product-washing process and short stoppages due to component adhesion errors with the m-FLIP operating rate improvement solution tool.

Pursuit of the Causes

We intuitively knew that short stoppages were occurring in the tray/rack refilling machine at the manufacturing worksite. However, we were not able to quantify the stoppage time or the number of stops. Accordingly, we collected data on the tray/rack refilling machine and the component adhesion device with m-FLIP. We uncovered the following problems as a result.
・Non-operation due to waiting for tray replacement: 12%
・Non-operation due to adhesion errors: 8%

Improvement 1: Reduction in the Tray Replacement Time

We now display on ANDON the number of trays remaining until the completion of the tray replacement (Fig. 9). We display in ANDON the timing of the tray replacement before the work of refilling the product into the tray is complete. This has meant we have been able to greatly reduce the time spent waiting for tray replacement.

Fig. 9: Countdown ANDON that predicts the equipment work completion

Improvement 2: Resolution of Short Stoppages

We analyzed in detail the contents of the short stoppages. As a result, we ascertained that 80% or more of the short stoppages were component adhesion errors (Fig. 10 and Fig. 11).

Fig. 10: Breakdown of the equipment operating rate and non-operating rate items

Fig. 11: Operating status Gantt chart

We found that the cause of the component adhesion errors was changes in the clearance between the product adhesion nozzle and product due to changes over time in the washing rack jig. As a countermeasure, we shortened the gap between the adhesion nozzle and the product. This allowed us to significantly reduce short stoppages due to adhesion errors.

In this way, it is first necessary to digitize the operating status of equipment and the behavior of workers to improve short stoppages. We can pursue the causes of short stoppages by analyzing and parsing the data we collect.

Summary

This questionnaire revealed issues in the conversion of regular factories into smart factories and challenges in promoting that in numerical data. We can see that quantifying and visualizing the various losses lurking in production lines is essential for improving productivity from these results. We focused on the theme of short stoppages this time. However, quantified data is also indispensable to resolve other issues in the conversion of regular factories into smart factories. The questionnaire results show that this requires high density and quality of the data to be collected as well as analytical skills.