The Value of Data in the Digital Factory is Much More Than Just Analysis


Michael Ford, Sr. Director Emerging Industry Strategy, Aegis Software

Advanced Analytics illustration with mobile devices and data dashboards
Advanced Analytics illustration with mobile devices and data dashboards

Today too many companies are focused on the acquisition of data in the hopes that the data will create value. Industry 4.0, on the other hand, is about optimizing the factory that today produces small lot-sizes and a larger mix of goods. Continuous optimization is all about searching for, and eliminating the waste, as well as eradicating all types of down-time. Industry 4.0 examines more than the simple internal machine stoppages. Industry 4.0 is focused on the analysis of time intervals in which there is zero data because the equipment or process is not actually working. In what way do we analyze “no data”? In actuality, we need to divide the scope of the digital factory into three levels, each of which must have their own transformation.

Level 1: The Factory-Floor

The IPC Connected Factory Exchange (CFX) has taken the digital manufacturing world by storm. With the substantial expense, resource-drain, risk, and deployment-related line downtime out of the equation, the rush is on in regards to the amount of information from machines, transactional operations, and people around the shop-floor. The CFX-driven Level 1 transformation, though published quite recently, is well underway in many factories already, with nearly every major machine vendor beginning to issue roadmaps and announcing CFX. Numerous “Proof of Concept” (POC) projects have begun, with many more soon to follow. The concepts behind IPC CFX, including true “plug and play,” the “last interface you will ever need,” and “machine connections for free” have motivated the entire industry towards full adoption within the next few months.

Level 2: Smart Manufacturing Software (MES etc.)

For generations, manufacturing software applications have been limited in functionality and capability in terms of machine connectivity. Solution vendors are now undoubtedly set apart from the rest. Those who were vital in the design and creation of CFX, who are already providing CFX support as standard, have current mature solutions based on IIoT architecture. They are intended to provide next-generation digital manufacturing benefits. Others are trying to get caught up and build their CFX support in a similar manner as any other interface, a remarkable advancement, but only to the degree that the original design of their solutions, having been architected decades ago, allows. The truth of the matter is that every software system, no matter if it is home-grown or created commercially, has an opportunity for improvement with the CFX being introduced.

Looking beyond the simple interfaces, and putting more concentration on the IIoT-based MES platform, there is critical value being added to the CFX data, by way of real-time contextualization. Let’s consider an everyday sequence of events, comparing an “analog factory” process against one with that has a digital MES. The process kicks off with a simple event. An SMT machine comes to a halt during production. The objective is to detect the stoppage, find the cause, resume production, and optimize the factory operation to prevent it from ever occurring again.

The Analog Factory:

The production worker notices a red light on the machine’s tower, walks over to the machine, and requests the reason for the stop. The machine report indicates that the stoppage was due to materials not being available for processing. The report only shows one piece of the problem and not the entire cause of the breakdown. This is the situation in about 80% of machine-based reports, as the results often pertain to issues stemming from problems not related to that particular machine. The operator then goes to the previous machine and runs another report. The machine indicates that it has undergone a material “pickup error.” The operator checks the machine and realizes that the feeder causing the problem is empty, but there are no replacement materials available. This is pretty straight forward, material logistics has messed up. After searching for a logistics support person, and describing the situation, the reply is that the issue will be investigated. The kit of materials was prepped and ready the previous day, so everything should be there. After looking through the kit, and emptying out the used reels from the trash, it was concluded that one of the reels of materials had in fact been missing from the kit. After checking the kit preparation records, all reels had been allocated and included, so it was assumed that the missing material must have been stolen from the kit – well, strictly speaking, not stolen, but was taken for use on another line. It can happen overnight that another line runs out of materials due to excess spoilage or an inaccurate reel count. Because the warehouse does not operate during the night shift, materials are pulled from other kits since that is the quickest and easiest way to keep the line -moving. Instead of correcting the issue, the unforeseen lack of materials is simply moved around. With two machines not working, the operator is dealing with the same pressure to get the line running again. It would take a great deal of time to issue another reel from the warehouse, accounting everything somehow with ERP, so it was suggested that it might be faster to look in the part-used material area, to see if there were materials available there that would get the line back up and running. After looking, a reel with a sufficient amount of parts was there, with the right part number. It is placed on the machine, and production begins. During a management meeting a few weeks after, the main concern is that a customer has discovered that unauthorized materials were used in their products, from a manufacturer who had not been qualified. This was discovered because a customer found a defect in their product. The dimensions of the unauthorized part had a slightly different geometry than the approved parts from other vendors, resulting in a wider variation of positional accuracy, which weakened joints and caused defects. The production manager vaguely recalls a decline in the yield at test, and a busy time at repair during the period in which the batch containing the defect was made. Nothing was done at the time since it was a busy period. Not anymore, however, as due to this issue, the customer is now looking for a new EMS partner.

The Digital Factory:

From the onset of the above scenario, in the digital factory, you would have anticipated that the machine stopping would have transmitted a CFX message to the II0T MES system to indicate that it had stopped because the product was not delivered. Having knowledge of the line configuration, MES would immediately identify the preceding machine in the line to check on the status. Although an improbable occurrence, as MES would have already reacted to the “pickup error” message from the previous machine. In fact, this, too, would probably not occur. MES is aware of the product detail, and how the job is divided up between the machines in the line, knowing where all of the material setup locations. MES also knows the specific materials that are loaded to the machine, the amount each reel contains, and the number taken per production unit, in addition to the rate the units were being produced and any spoilage. Prior to this situation occurring, MES would have requested an authorized replacement reel, way before the machine stopped working. However, MES cannot predict unplanned human behavior like in this situation where the material was physically taken out of the kit. The digital MES system, however, always allots and distributes material to machines on a “just in time” basis, so it does not require kits. There is never a reason to “steal” materials, since there is never a kit shortage, never even a kit. Even if an operator were to attempt to use a material in the wrong location, the system would prevent it. No possibility of consequential quality issues coming up in the future, no dissatisfied customer, no loss of business.

Was this example benefit used when determining the ROI for the digitalization of the shop-floor? Obviously, a digital smart factory using MES cannot be compared with the old analog factory, as things do not happen in the same way. Many down-stream problems are corrected by root-cause avoidance of bad practices. The value created in this example by the CFX data was created not by the analysis of the data itself, but by the operational improvements that it supports. This is just one simple example of many everyday issues that occur in analog manufacturing, that are removed in the digital factory revolution.

Level 3: Artificial Intelligence (AI)

Some operations are aware of the theories of factory process simulation. Not a lot, though, since the best software in this area is somewhat specialized and costly, and even that is now outdated when placed into the Industry 4.0 environment. Simulators are based on daily, weekly, or monthly periods of time of the factory operation. The industry 4.0 factory process changes daily, in conjunction with customer demand, and so any simulation is useless almost immediately after it is finished. The problem that simulation applications dealt with, however, is still pertinent to Industry 4.0 and is actually more important if the operation is to maintain a high level of output and also accommodate a large variety of products. The answer, however, has to be very different. Level 2 MES is very good at effectively managing schedules, as this is part of its foundation, and handles what is currently taking place on the lines and everything that relies on it, analyzing the data in real-time in order to automate and help in crucial decision-making. Level 3, conversely, is centered around gaining insight into areas that are void of data by fully understanding the entire operation when it comes to bottlenecks to provide better time management, for example, completing maintenance tasks when the line is not being utilized. No matter how knowledgeable or skilled a person might be, it is impossible to follow and monitor the thousands of variables that can be used to manage and constantly tweak the manufacturing operation. The ultimate goal of Industry 4.0 is to assist and improve the self-sufficiency of high-mix production, using AI in manufacturing.


The ability of Level 3 to evolve is dependent upon Levels 1 & 2 and is why it is difficult for most people to even imagine the possibilities of level 3, and the reason outdated and limited simulation systems still exist. The technology combining the CFX data capture in Level 1 with Level 2 IIoT-based modern MES platform available now, can provide amazing visibility and control on the factory floor and the entire manufacturing process. Let’s now anticipate the transformation in digital factory solutions at Level 3.

Sign up for our blog

Stay up-to-date on the latest in manufacturing trends, insights and best practices.