• Subscribe

The Road to Full Auto in Semiconductor Assembly Test Manufacturing – Emerging Challenges (Part 1 of 3)

Learn what matters most to experts as they address emerging challenges and complexities in semiconductor backend manufacturing.
Learn what matters most to experts as they address emerging challenges and complexities in semiconductor backend manufacturing. Key areas covered in this video series (part 1 of 3) include supply chain, technology and products, materials, equipment, and human resources.


Welcome to our first video in our on-demand webinar series on full auto and semiconductor assembly tests. In this video, we’ll talk about challenges in today’s assembly and test manufacturing. So what are the manufacturing challenges in assembly tests and what are the KPIs those challenges impact? First, supply chain complexity, which impacts cycle time, on-time delivery, and of course cost.

Next, technology and product complexity, which primarily impacts cycle time and the speed of technology development or the cycle of learning. Technology and product complexity is very related to our next challenge, materials complexity and traceability. Consider that in assembly test, we see a transition from simple single die lead frame packages to multiple chips in a single package or even multiple packages with 2.5D or 3D packaging advances as well.

These complexities impact OEE, ease of traceability analytics, and traceability granularity. Next, equipment and process complexity, which impacts OEE and yield. Finally, human resource complexities, which are measured in part by productivity, scrap, and on-time delivery.

Let’s dive a little deeper into supply chain complexity. In the back, we see the whole semiconductor supply chain from silicon to final product, starting from raw silicon to silicon wafers used in the fabs to assembly test and ultimately to a final product in the customer’s hands. What are the key challenges in the supply chain? Of course, diversity.

Vertical integration in the supply chain can vary anywhere from a wholly fabulous company that outsources every link in the chain to integrated device manufacturers trending to more and more internal manufacturing. Next, dynamic ordering patterns. I’m sure you are all familiar with the inventory challenges caused by COVID with parts on backorder for several months and some foundry capacity booked for the next couple years.

Multiple packaging technologies and emerging technologies, 2.5D, 3D, wafer-level packaging is causing an explosion in part numbers and thus additional inventory dependencies. What are the key manufacturing challenges related to the supply chain complexity? Short technology cycles. There’s always a new version of some widget to buy.

ROI requirements and their high capital investments. Tight lead times and on-time delivery. The cost of yield and scrap.

By the time an IC makes it to assembly test, a lot of time and money has already been invested. The further down the supply chain, the higher the potential losses. And of course, integration and interoperability between all of these facilities.

This can impact the ability to optimize supply chain planning, failure mode analysis, and much more. Now let’s look at materials complexity and traceability. As we mentioned earlier, there’s a transition from single chip packages to multiple chip 3D packages.

This transition directly causes a complexity of individual components as more and more different parts are needed for these more complex packages. So how do these complexities impact productivity and quality? Again, more materials means more critical dependencies. So we can potentially have more lined down situations due to short inventory.

There are more and more potential points of failure if we have poor quality source materials. And finally, the all-important yield. We need a constant supply of an ever-growing list of high quality source materials to get high yield.

Product and technology complexity. Here we see a couple of key technology trends. The internet of things and autonomous vehicles.

Each of these trends is causing an explosion in the demand for various electronics, including but certainly not limited to various sensors, processors, and displays. With the internet of things, we see more and more devices connecting to Wi-Fi from alarm clocks to bathroom scales to refrigerators. As you can imagine, the quality standards for electronics in the automotive industry is already high, and with the advents of autonomous vehicles, we should certainly expect these standards to get ever more stringent.

Next, equipment and process complexity. First, consider we have new emerging packaging technologies. This leads to, of course, an increasing high product mix and increasing counts of manufacturing processes.

These new packaging technologies also have more complex processes, and we see an increasing mix of process equipment used. In most assembly test facilities, we have legacy equipment, which may not conform to semi-standards like SECS/GEM and Interface A, which causes challenges in retrieving data from that equipment. This all leads to a big data challenge, considering the increasing volumes and variety of data.

And it’s not just about capturing the data, but also modeling and effectively using that data. Next, let’s discuss human resources complexity. Consider that most assembly test facilities are executing highly manual processes.

Without the benefit of full auto process enablers, people are making important decisions that directly impact and potentially harm productivity and quality. What lots should be processed next and on what equipment? Are there enough materials on hand? Is the equipment available or overdue for maintenance? Are there any restrictions based on process queue times or equipment certifications to consider? In our next video, we will show how Applied SmartFactory can mitigate these challenges. Now that we’ve had a deeper dive into the manufacturing challenges, let’s consider what we are hearing from the industry.

First, the number one priority is quality and, of course, addressing yield issues. Secondly, on-time delivery. Customers want to see prompt deliveries and committed dates held.

So the focus here is on mid-term, long-term planning, short-term scheduling, and real-time dispatch. Next, the transition to advanced packaging. By this, we mean fab-like processes like wafer-level packaging and panel-level packaging.

Next, data infrastructure. This is in part to provide direct access to customers, but also advanced analysis. More modern equipment can make gathering data straightforward if compliance with standard protocols, but old legacy equipment, as we mentioned, certainly can make data availability more difficult.

Productivity, both human and equipment. In part, this is referring to automated material handling, perhaps with AGVs or overhead transports, as well as minimizing human errors, both of which we cover in more depth in the next video. Finally, accelerated issue resolution.

The cycle of learning has to be quick and thus problem-solving needs to be very fast as well. Customers will continue to apply cost pressures and demand more traceability. These six points are the key items we hear from our assembly test customers.

Join us in our next video where we define full auto and how we can address these challenges. Thanks for tuning in.

About the Author

Picture of Joe Napiah
Joe Napiah
Connect with us on LinkedIn to begin your journey towards creating a world class factory today.