Smart Factory Case Study
The sensor-based “predictive maintenance” implementation at Murata's Fukui Manufacturing facility(Part 2 of 2)
In Part 1, we learned all about the IoT integration that took place at Fukui Murata Manufacturing, as well as the predictive maintenance system and achievements that the integration made possible. In Part 2, we look into how they proceeded with IoT integration at their manufacturing site. “It’s difficult to make progress in IoT integration, even if we understand the need for it.” Many companies voice similar opinions. In Fukui Murata Manufacturing’s case, what were the things you focused on as you worked to incorporate IoT technology? We asked Mr. Kubodera of production technology department at Fukui Murata Manufacturing about the rest of his story.
How Fukui Murata Manufacturing envisioned their ideal dream ideal towards paving the way for IoT integration
――The importance of IoT integration is frequently talked about, but I’m sure there are manufacturing sites that have trouble making progress with the incorporation process, too. With your story as an example, what do you think are the key points to moving IoT integration forward?
I’m sure there is more than one, but probably the most important is whether you can “paint a picture” of your outcome goal. IoT integration is not the goal in and of itself. Rather, IoT technology is what you are using to get to what you want to achieve as a company. I think that is an important point.
You don’t work to incorporate IoT technology just because it is becoming a thing in the industry, just like how you wouldn’t use a highly regarded device just because there are people using it. In other words, I would say that the real need isn’t there even if your reason for the integration is to solve immediate problems. Even in our own case with predictive maintenance, we had immediate problems such as the difficulties we faced in responding to sudden machine failures and the frequency of long wait times occurring before maintenance could be carried out. Still, these problems were not the only things we were looking to solve. Once you start articulating where you want your company to be in ten years and come to the conclusion that IoT integration is quintessential in achieving these goals, this is where you should start involving your entire company. If you can articulate all of this clearly, then you may be able to convince those at the top and managerial positions, and in turn drive the integration forward.
――What was the outcome goal in your case?
First, it was improving productivity per person. As Japan’s population declines, sooner or later, we would have to either continue with business as usual with a smaller workforce or achieve a higher level of productivity. I believe the incorporation of IoT technology is one way we could achieve this. For example, if we could use IoT technology to automate a part of our operational work, our operators would be able to in turn use that time to work on other tasks and projects as well.
Based on this premise, if we could use IoT to obtain the data for the performance and conditions of each machine, this might make it possible for even employees with little experience and technical knowledge would be able to perform simple inspections and checks. This would be possible because these employees wouldn’t even have to touch the machines to be able to use the data to determine the possibility of machine failure. We talked about IoT technology making it possible for us to visualize the conditions of the machines. If we could make the data so that it is simple enough for anyone to be able to understand, and make manuals that explain how to perform repairs accessible digitally, more of our employees would be able to perform that work.
With some of their work taken on by other employees, those operators I mentioned earlier might become able to take on some of the more “skilled” work. In the same way, IoT integration is directing skilled employees to different lines of work, while employees in other areas are pointed to others. This allows the boundaries of each and every employee’s capabilities to expand to include more and more of other types of tasks. I believe you can call that improving the productivity of each individual employee.
――So it’s all about establishing a concrete outcome goal and paving the way to that result.
That’s right. On top of that, it is also very important that you put together and provide data that is as detailed as possible so that you can convince others of the amount of loss occurring with the current methods, and what you must do to change it. With the predictive maintenance we incorporated in our facilities, we examined the amount of machine downtime and calculated the percentage of time that these were cases of corrective maintenance. We then went on to calculate and lay out in detail the amount we could increase in productivity each month if we were able to reduce the amount of corrective maintenance downtime, and used this to convince the management of the problem. We put together our outcome goal and then went on to thoroughly investigate our current situation. It was in this way that we were able to gain the management’s understanding and convince them that the situation necessitated the implementation of IoT technology.
――More specifically, were there any difficulties in putting all of this into practice?
There of course were difficulties. From the perspective of those at the manufacturing sites, even if you understand the work in your head, it’s hard to take those first steps. Your job description has done a 360 after all. That’s where we came up with making the integration a “full-time responsibility.”
This kind of reorganization is hard to work with just on the side. In the case of predictive maintenance, for example, if there is machine trouble occurring right in front of you, you wouldn’t be able to get to the actual predictive maintenance for machines that haven’t experienced trouble yet. This is exactly the reason why we created positions where employees work only with predictive maintenance.
First, we set up a production line in the factory that later become our model case for predictive maintenance. We then assigned 7 specialized members to these positions, and made them a team that will only be implementing predictive maintenance. Their responsibilities included everything from picking out the sensors to set up on each machine, to collecting data, setting thresholds that trigger alerts for each set of data, and performing maintenance when the alerts sound.
I believe it is all because we had them working only with predictive maintenance that we were able to come all this way. The 7 members of this team were also each highly knowledgeable regarding the manufacturing processes they were working with. We chose these individuals because we wanted people who were influential and had the power to really relay the results of our trial. People who would be able to comprehensively communicate the results of the trial to others in the company and convince them of the effectiveness of the process. These were people who would be able to get others to think, “If they say so, maybe we can give it a try.”
――What were these seven individuals consideringwhile working with this trial? Were they convinced of the processes when they started?
Right. We discussed what I mentioned earlier--our outcome goal, significance of the implementation of IoT technology, and the challenges we were facing without IoT technology--over and over again within the team. It was like we were working to steadily increase our number of believers. Once we’d set all seven members’ sights on the same vision, we moved on to the execution of the plans.
Conducting trials to prove the concept on one production line, then using these results to gain “fans”.
――Were you carrying out predictive maintenance on only one line in the factory when you first started out?
Yes. We started out small, on only one of our lines. On that line we put predictive maintenance into practice and generated our results: a decrease in corrective maintenance and improvements in the operating ratio and production. We were basically performing a demonstration to prove our point. We used our results to convince others, and expanded predictive maintenance techniques to other lines as well. We used “Predictive maintenance sure is nice” as our motto, and steadily gained “fans.”
IoT integration would probably be pretty difficult without any backup from the company. I think that to gain this kind of cooperation, there needs to a clear vision set for the outcome goal and communication with those at the top about the necessity of IoT integration.
――On the contrary, was there anything you’ve learned through the process of implementing predictive maintenance using IoT technology?
That would be balancing the cost side of things. In the case of predictive maintenance, the more detailed your sensor setup, the more detailed your data. The only problem is that this raises costs.
When we first started integrating IoT technology, we would put forward the problem we were having the most trouble with in a certain facility. Low number of finished products, overheating in the machinery, or something like that. We chose one problem, and while thinking about why it occurs, we also thought about the type of sensor to install for it. It is here that we run into the problem of finding a balance with the costs.
If I were to use cars as an example, let’s say there is the tendency of a decrease in passenger comfort while the car is in motion. In order to figure out the cause of that problem, it would be ideal to attach sensors to each of the suspension parts or tires to collect data.
However, if we did this, costs would go up. If that’s the case, then we can consider the option of attaching a sensor to the steering wheel only, to get alerts to sound from data collected from the position of the steering wheel, such as when the car moves to one side even when the steering wheel is straight, or when there are irregularities with the steering wheel. Of course, with this method, we would not be able to find out whether the cause of the problem is in the tires or the suspension system, or in an entirely different part of the car all together. However, once we detect a problem with the steering wheel, we could have specialists such as the maintenance staff come in for an inspection. With this method, we save on costs while being able to detect problems in the machinery.
――So there is no need for sensors to detect the exact cause to each problem, but rather the key lies in where to install the sensor while finding a good balance with the cost.
That’s right. The key lies not only in the costs but also in the training of our employees. While it is true that the more detailed the sensor setup, the easier it is to identify the exact cause of a machine failure, it may also be necessary to consider a few other factors, since it is also possible to train the maintenance staff to be the ones to identify the causes. While IoT integration should help simplify tasks to improve the work productivity of each employee, developing the skills of your employees can also lead to an increase in productivity. I think this is something that should be tailored to suit the policies and goals of your company.
――Finally, please share with us what you expect out of IoT integration in the future.
The goal is to increase the work productivity per employee. To that end, we are currently working mainly on IoT integration in our facilities, production lines, and production processes, but I believe it would also be possible to implement IoT technology that focuses even more on the individual employee. For example, there might be prospects in using wearable devices that can check employees’ health and collect data to help employees with health management as well as help them perform better.
When discussing the improvement of efficiency through the implementation of IoT technology, there is often mention of the notion of a “reduction of manpower,” but I believe it is also important to note that it also allows us to make the most efficient use of our workforce. In other words, to consider how much we could actually increase productivity for each and every employee. I hope to work towards that objective as I continue with the implementation of IoT technology.
Mr. Kubodera, Production Technology, Fukui Murata Manufacturing
Started working at Fukui Murata Manufacturing in 1994. After working in manufacturing and maintenance and then working to streamline production, he now plays a key role in promoting process innovation in facility maintenance. Introduced “predictive maintenance” in order to create a shift from maintenance performed after a machine failure occurs, to detecting impending problems and preventing them before they occur. He rationalized the predictive maintenance process, set up predictive monitoring technology, carried out the circulation of its implementation, and more. He was also involved in the development of the maintenance support systems. 3 years after starting the project, corrective maintenance time has gone down by half of what it was before the project.
Related articles
- How to Fill the Gap in the Proficiency of Workers? Iwate Murata Manufacturing Edition (Part 2)
- How to Fill the Gap in the Proficiency of Workers? Iwate Murata Manufacturing Edition (Part 1)
- What Is the First Step to Converting a Regular Factory into a Smart Factory? Komoro Murata Manufacturing (Part 2)