IoT for Legacy Equipment at Manufacturing Sites, High Cost Performance – Kyoto Semiconductor
In the wake of Industry 4.0 and other initiatives, successful cases of IoT application are being reported around the world.
In the past, Japanese workplaces often relied on highly skilled human resources, but with the aging of skilled engineers, small and medium-sized companies are now introducing new systems that are expected to produce significant results at low cost.
In our case study, we see the significant effects of the system, which is even compliant with Toyota Production System (TPS), such as significant reduction of downtime, effective use of equipment, and effective use of workers.
This has resulted in more than 1 billion yen benefit in 5 years, while the cost is only 2 million yen per year, which is very cost effective and a step towards the future development of AI.
Converting Legacy Equipment to IoT on a Small Budget
Kyoto Semiconductor (Kyoto, Japan) announced that it has started operating a “Smart FAB” as part of its digital transformation (DX) initiative on December 4, 2020. Kyoto Semiconductor is a manufacturer of optical semiconductors from wafer (front-end process) to packaging (back-end process).
The company is a small company with annual sales of 3.473 billion yen (fiscal year ending March 2019).
Collect operational data in near real time
The elemental technologies used in the manufacture of products determine the characteristics of the product and contain a great deal of know-how. For this reason, equipment that uses these core elemental technologies sometimes may be used for 20 to 30 years.
This time, they have started to convert some of these facilities to IoT by combining Raspberry Pi and various sensors, and monitor the operation status on the network.
By automating the sending of e-mails when an abnormality is detected, the system allows workers to check the operating status of the equipment and identify abnormalities at an early stage, even if they are not there.
Linking with information from the core infrastructure
In this initiative, sensors have also been installed in pure water purification and various gas lines to start monitoring.
In addition, with the introduction of Siemens’ industrial gateway and the open IoT operating system MindSphere, sensor data is automatically recorded to a server on the network. As a result, when an abnormality occurs or the condition of the equipment changes before or after maintenance, the operational status data of the equipment can be reviewed, and the “awareness” that was previously based on the experience and knowledge of skilled workers in identifying and addressing problems can now be shared even with less experienced workers.
Reduce downtime by 70%, Extend equipment life by 5 years
They aim to reduce downtime by 70% through early detection of defects and analysis of operating status data.
By analyzing and integrating these data and information from each facility, they will be able to analyze the causes of problems that were previously invisible, improve the speed of improvements, and accumulate data for failure prediction.
These efforts can extend the life of old equipment by five years or more, leading to reduced capital investment, and also to consideration of environmental conservation and business safety, which are important items in their sustainability policy.
Cost is 2 million yen per year, and benefit is over 1 billion yen in 5 years.
The investment for the entire system will be about 2 million yen (19K USD)per year, and the system will be able to monitor the operating status in almost real time from the cloud, which will improve the efficiency of equipment maintenance and product quality.
By reducing capital investment, they expect to achieve an effect equivalent to 1 billion yen (9.3M USD) in 5 years, and by further reducing costs and automating the failure prediction function and analysis through the use of AI, etc., they expect to achieve an effect equivalent to an increase in sales of 500 million (4.7M USD) yen in 5 years.
Conclusion
The above is a case study of Kyoto Semiconductor, which is expecting great results at low cost.
Here we are utilizing the inexpensive single board computer Raspberry Pi and various sensors.
The sensors being monitored include gas flow rate, vacuum level, and high frequency output.
As a system that achieves the same kind of thing, Amazon Web Services (AWS) provides a system that enables predictive maintenance using AI at a low cost, with a set of sensors, devices, and services. The sensors are designed to acquire vibration and temperature, and the monitoring targets are assumed to be rotating equipment such as motors, fans, and compressors.
Amazon Monitron includes not only these sensors, but also a gateway device to securely transmit the data to AWS, the Amazon Monitron service to analyze the data for abnormal machine patterns using machine learning, and a companion mobile app to set up the device and receive reports on machine behavior and alerts about potential machine failures.
With no development work or ML experience required, you can start monitoring the health of your equipment in minutes, enabling predictive maintenance using the same technology used to monitor equipment in Amazon fulfillment centers.
Amazon Monitron
Detect abnormal machine behavior and enable predictive maintenance
In Japan, Murata Manufacturing Co., Ltd. and ACCESS CO., LTD. jointly developed JIGlet, a smart manufacturing support tool which will be launched through their respective sales channels in Japan beginning on February 19.
JIGlet has three types of sensor devices with built-in data communication SIMs:
- an “illumination device” that detects whether a lamp is on or off
- a “button device” to count numbers by pressing a button
- a “dice device” that records time by operating it at any working time
It consists of single-finger data collection, a visualization screen, and a chat app for notifications.
The same trend will continue around the world as low-cost solutions begin to become available.
Company Announcement: Kyosemi Smart FAB starts operation (Japanese)
