If there’s one area Taiwan is #1, it’s semiconductors. I was fortunate to be master of ceremonies for National Taiwan University, a host of SEMICON Taiwan — one of the industry’s largest conferences. Semiconductor = SEMICON, get it?
Taiwan used to be a world leader in toys. Now the toys are made in China, but the CPUs inside may be from Taiwan. For devices and electronics that are even smaller, higher tech, and proprietary, it’s more likely it’s made in Taiwan. You’ve heard of AirPods? AirPods are made in Taiwan by Inventec. Today, advanced manufacturing is one of Taiwan’s strengths, and semiconductors lead the way.
- In semiconductor materials, Taiwan is the largest market in the world
- Taiwan is the largest market for new semiconductor equipment
- TSMC is the world’s largest independent semiconductor foundry
Smart Manufacturing and National Taiwan University
There’s a bigger opportunity for semiconductors with industrial applications than consumer goods. New CPUs can make smart anything possible. This is what the president of SEMI Taiwan says about it.
“The four major trends of applications in the market… Internet of Things (IoT), Smart Manufacturing, Smart Transportation, and Smart Medtech.” – Terry Tsao, president of SEMI Taiwan
Smart everything, basically. These are all major categories of their own, but it all starts with smart manufacturing. Knowing what is being made and how, anywhere and when in your supply chain. And, optimizing it.
“Effective Ways to Make Facility Smart.” The High-Tech Facility International Forum by National Taiwan University’s College of Engineering and its High-Tech Facility Research Center focused on the path to developing smart manufacturing facilities. This also made it one of the more popular forums at SEMICON Taiwan.
Because everything in the manufacturing future incorporates principles of information technology businesses. Especially engineering, mechanical or electrical or civil. Building these facilities ties all of that together. Smart factories simply know more about what goes into smart products.
Get to Industry 4.0: The Hot Topic of SEMICON Taiwan 2017
Many countries have initiatives in this area. In the U.S., it’s smart manufacturing. Europe calls it Industry 4.0, which is what Taiwan generally goes with (it’s catchier). China put it in their China 2025 plan. This was also the major discussion at SEMICON Taiwan. How can Taiwan become a leader in Industry 4.0?
Well, each nation faces major roadblocks. One example. China relies on imports to meet 90% of its integrated circuit needs. This is because for now, China is doing mass manufacturing for the world. Meanwhile, Taiwan is hanging onto its lead in advanced applications. Each nation is traveling a different path.
- Taiwan is optimizing its supply chain
- Germany is focusing on green manufacturing
- The U.S. is re-industrializing (With help from Taiwan’s Foxconn)
Everyone is taking manageable steps because certain obstacles to ongoing transformation can only be taken apart over time. Big Data is one of the more general roadblocks, and one I find personally compelling.
Big Data vs. Business Intelligence. Know the Difference.
Big Data is not an enormous flat file.
Big Data can help optimize smart manufacturing, but people confuse Big Data with Business Intelligence. Here’s what it’s not. Big data is not an enormous flat file. “Big data is a term for data sets that are so large or complex that traditional data processing application software is inadequate to deal with them.”
There are three ways to work with data. Only the third is truly Big Data.
1. What has happened? This is Business Intelligence. Descriptive Analytics looks at past data to review what happened. “Sales were down because we produced less units than last quarter.”
2. What might happen? Predictive analytics. Here, we are finding patterns between different data sets that might lead you to make a different management decision. Great for forecasting demand. Still based on the past. “Retailers cut their orders, partly because of increased shipping costs, so we produced less units, and sales went down.”
3. What will happen? Prescriptive Analytics is modeling the effect of future management decisions on future outcomes. This is where Big Data comes in. To help you figure out what you could be doing before you know what you should be doing.
One barrier is there aren’t enough people in Taiwan capable of using prescriptive analytics. Ironically, Taiwan has the manufacturing but not the data scientists. The U.S. is in the exact opposite position. One estimate I’ve heard places Taiwan’s shortage at 5,000+ data scientists. If you’re about to graduate from MIT, there’s a job for you in Taiwan.
As a practical solution, some leaders like National Tsing Hua University professor Chen-Fu Chien are advocating Industry 3.5. Maybe Taiwan can’t enable, but it can enhance. Perhaps that’s good enough for right now.
Warning: This Isn’t Covered in a MBA Program
Advanced manufacturing and semiconductors aren’t discussed in my MBA program. I got it from the College of Engineering, where the experts are. An advantage of getting a MBA in Taiwan is being around all this high tech. But it isn’t going to be conveniently given to you. Even if you’re paying for it! So, I don’t want to give the impression this is what you can expect if you come here.
Management schools, by design, don’t do cutting-edge. They do Practical. Until then, it’s not coming into a MBA program. Or it is, but in “know-this-buzzword” form. Despite the marketing, a MBA probably isn’t raising your ceiling. It eliminates the liability of what you don’t know or aren’t connecting with.
This is also why managers get Big Data confused. Statistics for MBAs covers the most practical aspects of statistics — what has happened and anticipating what might happen. Many decision-makers tend to plug the term Big Data into their understanding of statistics. This is not correct. An easy way to remember this is Big Data is beyond the powers of Microsoft Excel.
This graphic from Wikipedia explains the difference between Industry 1.0 and 4.0. You can intuitively understand the different ways data was used for management decisions, and what some current and future needs are.