Reading time ( words)
Nolan Johnson speaks with Mycronic’s Clemens Jargon and Göran Frank about data-driven production technology and strategies. With the introduction of new data-driven software, Mycronic has made monitoring the line easier to use, faster to learn, and more purposeful. They discuss the company’s plans to drive competition and improve customer engagement.
Nolan Johnson: Let’s start with a quick overview. How does Mycronic define data-driven production?
Clemens Jargon: We have recognized in the past couple of years that next to improving our product portfolio from the hardware side, there was a strong demand for adding software competence and continuously enhancing software capabilities. And we don’t do software just because of software. We do it because everything in the production line needs to be rationally analyzed based on data. Therefore, it is so important to collect good data, because then we can analyze it and continuously improve the performance of lines. This is a significant step forward compared to SMT in the past years, and a driving factor and motivator for customers, as well as for us as suppliers going forward.
Johnson: That does seem to be the challenge. It’s one thing to gather all the data, and there is plenty of it. But making use of that data, turning that into knowledge, into action or reaction, finding out what’s really happening both on the manufacturing floor and in the front office, is a critical outcome of all that data.
Jargon: You’re right. It’s not only about collecting data; it’s about collecting the right data. At productronica, we showed two software products, MYCenter Analysis and MYPro Link, that are both focused on analyzing specific data coming from targeted sections of the SMT production line. Goran, can you elaborate on what we are offering here to our customers?
Göran Frank: MYCenter Analysis and MYPro Link share the same philosophy and principles to serve the same goal, which is to maintain and to improve the overall production efficiency. Both are based on real-time data generated by production equipment: MYCenter Analysis analyzes data from the placement process, while MYPro Link focusses on data from inspection processes, SPI and/or AOI; both of these software have three main purposes.
First, to provide a dashboard for the customer to monitor on a high level the performance of a specific process, and its adjacent processes, when possible, in order to better understand the overall performance of the line. For example, MYCenter Analysis will look into pick-and-place cycle times, and KPIs of the process while MYPro Link will monitor the First Pass Yield or DPMO based on inspection data.
The second objective is to help staff maintain overall process efficiency by displaying real-time alerts and providing an in-depth understanding of the process behavior. For example, if the placement process features an unreasonably high reject rate, MYCenter Analysis enables a deep drill down analysis of the pick-and-place data to identify the problem’s root cause and to eliminate it. MYPro Link tells which product or component package generates most defects, or whether the solder paste printing process is slowly drifting off tolerances, and gives SMT production staff the ability to quickly and efficiently dig into inspection data to identify and treat the root causes of such issues.
Third, this software can be used as a management report, showing where to look into production data to see where all the improvement areas or the bottlenecks are right now.
Jargon: To complement our answer, Nolan, we can ask about the market driver behind this data revolution. MYPro I series, the new Mycronic 3D AOI, is a very good example because it combines the need to provide systems and solutions, hardware, and software, able to compensate for the lack of time or experience from production staff. Thus, all the required knowledge is put into the software, giving it the ability to exploit and analyze the data reliably and very fast. A trend we are seeing in the low-volume, high-mix production environment is the growing number of smaller batches, which adds more difficulties.
If you want to implement a high level of control over your production in a low volume and high mix environment with very high expectations over reliability, then you need to process an even higher amount of data in the same period. Therefore, you need the assistance of smarter, and more autonomous systems to help you maintain your level of reliability and profitability. This is why our new 3D AOI is powered by artificial intelligence: it allows to handle a larger amount of data, such as new program creations, in the same period of time and with the same production staff.
Johnson: Where’s the primary focus? Is it using this data to give actionable information analysis to the operators, to business operations, or to someone else?
Frank: We see three major stakeholders for this type of product. First is the operator at the line, providing them with actionable, timely information to be able to prevent, mitigate, and improve instantly. That’s the first user group. Second, on the production engineer side is keeping track of, monitoring, and improving production over time using a management report. It helps to see where we have the bottlenecks. How long is my changeover? How problematic is my changeover? How could I break the changeover process down? This is something that most of our customers are monitoring in weekly, biweekly, or monthly meetings to see how they can improve.
The third stakeholder using this type of product is, of course, the manager (or the owner, if it’s a smaller company), looking for high-level KPIs with, for example, the overall equipment efficiency (OEE). Data is needed in many areas for the customer to use, but also on the engineering and manager side.
Johnson: When we think about digital factories and data driving manufacturing, there is an immediate focus on manufacturing process windows. But it seems you’re suggesting process optimization at an even higher level. Is that on your mind as you’re doing this development work?
Frank: Our ambition is to take an overall view and give the tools needed for improvement. So, this is one tool that could be used. In some cases, if you need to do a deep dive in some process, then of course you need to look into other data as well. But on the high-level side, we can provide that type of overview for our customers. In some cases when even more data is needed, then we can provide from this type of product. That’s the reality.
Johnson: What’s the customer experience so far in using the Mycronic tools for data-driven manufacturing?
Frank: MYCenter Analysis has been very well received by our customers since it was released about two years ago. We’re giving them the right tools to act based on data. All our customer groups found something beneficial in this software. If they run larger batch production, we can look into cycle time improvements and line balancing topics. If they run smaller batch production, we can look into how to improve the reject rate from the pick-and-place machine.
Jargon: I think customers like it because it’s very easy to use. There is so much software in the market competing to be the most complex and pretending to be able to manage everything for everyone. It makes these software solutions expensive, complex, or not easy to use. We asked whether we should go for some standardized, off-the-shelf market solutions, or develop something which is specific for our customers’ needs. And that’s what we made; customers really like that the data we take and analyze is the right data, and these are performance drivers for them.
We want to make their day-to-day production more efficient. For example, now they can easily identify that a high component reject rate is correlated with some specific feeders. They can understand that these feeders that have been in used for a couple of years are worn out. So, why not exchange them? Then, they suddenly have a lower reject rate, they save money, and improve their profitability here.
We try to focus on what really matters and what makes sense for our customers, and this is why they like it. Sometimes, it’s better to have less but be focused and efficient rather than having an overwhelming software solution.
Johnson: Here in North America, and I think it’s somewhat global, it’s difficult to find staff. Hiring enough people to fill the manufacturing floor tends to be a problem right now. Basically, you’re training people, hoping to keep them for three years, maybe. That starts to put more pressure on the software, and on building knowledge into the systems.
Jargon: Equipment and software must be easy to use and to be trained on. Trainees should be able to easily remember certain relevant things. It’s one reason a software or a system needs to focus on what really matters. When you give the operator something where four weeks of training are needed, and then he or she still makes mistakes, then your solution is not really helpful. You’d rather have something that is focused or something that can be learned quickly in order to avoid human mistakes.
The software has to be safe and user-friendly. Some operators have told us that after only a few days of using MYCenter Analysis with the placement machine, they already know how to operate the system; they know what data is provided and what is expected of them to improve the reject rates, for example. It is so critical and difficult to get skilled people in these days, so you need to make it easy for them to get started and to enjoy working with the system. Typically, operators who work on Mycronic systems say they like working on it and tend to stay on longer.
Johnson: Let’s just do a thought experiment. A customer decides to start using this software. They’re in the setup phase. What’s the process? For the yields you’re mentioning, I would expect some detailed consulting analysis, set up programming, optimizing phases to get the Mycronic software set up, and running on my manufacturing floor. What’s the process?
Frank: With this new type of software, such as MYCenter Analysis, in many cases we can do remote installation. We don’t need to visit the customer onsite; we don’t need to take action on the machine. The machine needs to be stopped, but we don’t need to schedule a maintenance window to do this type of installation. Rather than in days, the install can be done in hours. We are putting in a lot of effort to make it easier for customers to implement these solutions.
We spent a lot of effort to make it intuitive: hovering with the mouse will display guided information. We have good, embedded help in the software, showing the most frequently asked questions. And as the new generation is stepping into the business, we had them in mind, making the software very intuitive and immediately usable. With a one-hour training, you should be able to start using it.
Johnson: That all sounds good, but what about understanding the company process?
Frank: Sure, while some of the alerts and where to troubleshoot is made easy for the operator to understand using the software, understanding the complete production process and where you have bottlenecks needs some additional skills. But the benefit of the software is to have the right people working on the right task in the company. The highly skilled production engineer and managers should look at what they do best, high-level KPIs, and break that down. The production engineer should look to, for example, the activities going on in production. For example, what’s the changeover time, and then they could break that down based on their knowledge.
Johnson: And on that analysis that you were just describing, it usually takes someone from outside the organization to lead that conversation?
Frank: Actually, we are already doing this in some regions. In the Nordics, for example, we are guiding our customers by looking into their data and giving them advice on where the possible bottlenecks could be. These discussions and services are already running in some areas.
Jargon: It’s why and how we train our sales, presales, and service organization: to first visit customers and evaluate their request, to analyze and to understand their application demand. Based on this, we propose a solution which fits their needs, usually hardware and software. It’s extremely important for them that we are not just selling a machine, but that we solve their problem when they have a need for batch size solutions. At the same time, they need to be flexible to offer a batch of whatever, 10,000 or 15,000 products. Then they know that it can be done with our systems and our analysis. It is an important part of our sales pitch, how they finally use our solutions.
We look at the deficiencies in the overall process. It starts very basic, but we must make sure that they get the right sourcing in place, and that all material is ready on time when they need it for production. In the sourcing department, for example, they can get an early alert when the storage is running out of components, and they know what to do to reduce the deficiency of missing components. That’s so important in today’s supply chain; if you don’t control your material flow, and you recognize that certain components cost 20 times as much as they normally cost, you are thankful for having a supplier which helps you to do material supply control.
Then we discuss reject rates. In the old days, every component you reject might cost you €17; now the same component suddenly costs €200, so, you must be careful. It’s another area where we try to avoid deficiencies. When we visit customers, we ask questions, such as, “Where are your deficiencies in regard to a good quality output? How can our solution help you to overcome hurdles, avoid barriers, and solve your application demand?” And then we start installing.
And when we install, we don’t just install. We have the customer set up their production. We have had customers for 15 or 20 years because they appreciate how we get along. New customers see our flexibility, continuity, and a total solution offering. I think that differentiates us from the competition. There are some high-volume players with huge portfolios, but they typically focus on the high-volume markets. We feel Mycronic offers the same high quality as those high-volume customers get, but in the high flex area. And this is our sweet spot we have settled in. And again, I think this can only be done in the quality we are doing it, when you have a good offering in hardware and software.
Johnson: Great. With this software, are you hardware-agnostic? For example, if it’s based entirely on pulling information from Hermes, CFX, etc., can you use that software with any machinery on the line? Or are you anchored to at least one piece of Mycronic hardware in the line?
Frank: We provide full line solutions. But also understand that we may need to coexist in production, in lines with other vendors. When it comes to MYCenter Analysis, it could even benefit the software by having non-Mycronics products in the software. We could understand what’s happened before our process and off the run. Even though we only have one or two processes from us, we could still provide quite a good tool to monitor the complete line. With the takt time of the line, you still have some benefits, but for other customers with a larger MES installation, for example, we don’t intend to make an MES solution for the complete factory.
In that case, we are working with standards. You mentioned CFX. For us, CFX is the foundation to pulling out data from all our different production processes. For the customer who needs to integrate to our data, we will do both. We don’t intend to build this top layer in competition with the larger players on the pure software side like Valor or Cogiscan. In that case, we will collaborate with open APIs.
Johnson: I could easily imagine being in a facility where there are multiple lines, and maybe Mycronic equipment are on some lines and not on others, but the company choice is to standardize on your software. At which point, you may be using the software on lines that don’t actually have Mycronic equipment.
Frank: For those cases, it’s important that we have the APIs to be able to pull out the data.
Johnson: Let’s talk about what you see as the key market drivers. What are the pain points, really, for customers? We’ve already talked about supply chain and yields and reject rates on components when they’re so expensive, but are there other drivers besides that that are causing customers to look at the sort of solution hard?
Frank: Of course, we meet with customers for variety of reasons, but we also look to the data of our customers. We share what we’ve learned regarding full line solutions, jet printing, inspection, pick and place, and material handling solutions. We have a good understanding of the processes. Our DNA is our material handling process, and we know for a fact that the material handling process is a bottleneck, as Clemens also mentioned here, we’re starting with the logistics process and the complete planning process of the customer. It’s integrated in the complete flow.
And here, we know it’s not a hardware or a software thing. It’s a process thing. And on the topic of data-driven manufacturing, now we have different sensors to collect the data. With the reel, some are with unique IDs, we keep track of quantities. All this data we have been digitalizing or collecting over the years is important. I’m quite confident we have the best material handling solution in the industry, especially when it comes to the high mix, high flex segment.
Johnson: Do you have any closing thoughts?
Jargon: Today's electronics manufacturers, regardless of their size, are faced with an unprecedented level of complexity. This complexity stems from a need for competitiveness that requires ever finer control over the efficiency of production processes, using more sophisticated equipment and technologies. But this growing complexity is also the result of a changing and uncertain production environment—as is currently the case with disruptions in the component supplies—and the multiplication of electronics technologies and fields of application. All this translates into a mass of data to be processed that has never been so large, and that will continue to increase in the coming years, while the human resources in place will not be able to evolve at the same pace. For this reason, it is necessary for electronics manufacturers to follow this evolution by adopting tools that allow them to integrate, analyze, and interpret this data, which today represents a raw material that needs to be leveraged by all means. Once again, whatever the size of the electronics manufacturers, those who will be able to seize this opportunity will become tomorrow’s leaders.
Johnson: Well, gentlemen, this has been a great conversation. Thank you.