The drive for increased automation and significantly escalating the productivity of an SMT line requires several critical factors coming together to create a truly smart line. First, innovative individual machines should maximize the potential performance for the respective processes. Second, there must be interconnectivity between the machines so that they can communicate with the upstream and downstream processes. Third, you should have the ability to automatically adjust to products coming down the line with minimized downtime caused by any required changeover.
What people are ultimately seeking is all of the above with the capability to use the line data to statistically analyze how to make improvements to production processes autonomously. I will explore why this is the ultimate goal, how far we have come, and what is on the horizon.
Software: Influencing Capability
The capabilities of SMT equipment and subsequent improvement from systems are usually a gradual process. However, we’ve entered a period of software influences, including continuous update releases and shorter R&D cycles, which have led to a steady stream of new capabilities for the connected manufacturer. Because the goal of a smart factory is ambiguous, the route to achieving the golden state of a fully automated process also has nuances, but these are being solved from many angles (Figure 1).
Figure 1: A fully automated process is the goal of improved manufacturing. Innovative equipment suppliers are helping companies achieve this ideal.
Let’s say that you already have top-quality machines installed for individual processes. Sure, they can continue to be improved, but you may have gotten to a point where the increases in capability will remain similar. You want interconnectivity on an SMT line so that the individual parts can communicate freely and you can easily see an overview of data for the entire line’s performance. Traditionally, the industry has used information from the Surface Mount Equipment Manufacturers Association (SMEMA) as the standard of communication between machines, which has worked well. You can even use extended SMEMA for the communication of machines further downstream or upstream.
But SMEMA has its limitations. Due to the method of communication using pins and basic signal transfer, you are limited to the amount of information that can be sent and the number of machines it can be sent to with the two-way transfer of data. For example, a screen printer can tell your magazine loader that it is ready for a new board. However, SMEMA can’t communicate the product type, required program, or convey any of these details with the rest of the line unless it uses extensive barcode readers.
Utilizing New Standards
Next, you might think about the line’s capability automatically adjusting to a product changeover. This is important for manufacturers because they want to reduce changeover time and increase productivity. Ultimately, increased productivity leads to increased profitability. Usually, the industry has used barcode scanners and automatic width adjust board handling, but the cost of all of those barcode readers is something that sites are now looking to avoid. These various factors have given rise to smart board handling, which utilizes the new communication standard Hermes rather than the SMEMA standard. This change further introduces the full capabilities of Industry 4.0 technologies to the assembly line.
Imagine you have a magazine loader feeding a laser marking system that marks a customer and product barcode after communicating with your manufacturing execution software (MES) on what needs to be produced. Then, the product information is distributed across the line via an Ethernet connection so that all of the systems are aware of what program needs to be selected on each machine to complete the build. Programs change, board handling widths adjust, and the line reacts to that initial master laser.
The machines can also compensate for each other in the line by constantly examining progress, errors, and faults to reduce bottlenecks (Figure 2). That is exactly what the Hermes standard allows for, and it is available currently from all reputable suppliers. The other element of this communication comes in with IPC Connected Factory Exchange (CFX) and the ability for each machine to communicate with the MES central system. This is an important topic for most software companies and a significant step towards becoming a smart factory.
Figure 2: Smart lines communicate to help solve the puzzle of reducing bottlenecks by examining progress, errors, and faults.
Imagine two teams of 10 people both trying to solve a puzzle. On one team, each person can only speak and hear the people on their immediate right and left. Meanwhile, the other team has a captain (MES), but they can all communicate freely and share relevant information that could affect each other. Which team do you think would be best equipped to solve the puzzle?
Using the Smartest Software
What about automating system feedback to improve performance without a human operator required? To break it down a bit further, an automatic process improvement guided by a machine learning algorithm utilizing statistical analysis of real-time production data. This might sound too good to be true, but the best suppliers in 2019 have already been working to incorporate forms of AI for years. A real example would be between a screen printer and solder paste inspection machine. It’s a little-known fact that 84% of usual production errors are due to solder.
So, how could artificial intelligence (AI) make that better and be the cherry on top? Today, the KSMART Process Optimizer (KPO) module from leading smart factory solutions provider Koh Young uses real-time production data from the inspection of solder paste (Figure 3). The KPO statistically analyzes the outputs of the printer using data points from the solder paste inspection system and pre- and post-reflow AOI to understand any underlying improvement it can make to the process. This increases performance even if the results are already within the set tolerances.
Figure 3: AI is the cherry on top of smart manufacturing.
The solder paste inspection system will then instruct the jetter or printer to adjust its program without human interference, improving production results. Just like the human body uses a similar feedback loop method to control its internal body temperature, this line of machines can take real-time data, analyze the outputs, and adjust accordingly.
The Ultimate Smart Line
The key buzzwords you should look for include Hermes, CFX, AI, smart conveyors, and machine learning—all of which will be the pillar of the ultimate smart line (Figure 4). The key driver for moving the project forward will be the companies that show innovation within technology—specifically, software companies that create products to interface machines easily and react to the relevant data produced by one another.
Figure 4: A smart SMT line is the future, and getting the equipment correct is essential.
Richard Booth is the managing director to the Altus Group Companies including both Altus Group Ltd. and Danutek Kft/Srl, an electronics production equipment distribution partner in seven territories across Europe.