In semiconductor manufacturing, innovation and operations must be tied to value. The tools that connect technology to value, operational models, mimic the operation (i.e., a factory, a business, a process), in whole or in part, to assess whether the path the organization is on (roadmaps, developments, etc.) will achieve the desired results. The concepts are tested in software to determine the payback, allowing smart choices to be made about the next steps for the business. At the core of operational modeling is one simple principle: Every decision, even a decision that appears technical, is a business decision.
The tools — cost of ownership (CoO), cost and resource modeling, and discrete-event simulation — have resulted in massive improvements in cost and performance. The underlying standards developed by SEMI for the semiconductor industry have also been ported over to the flat panel display (FPD) world and they also port over to the photovoltaics (PV) world. The methods that allow information to be analyzed are common to semiconductors, display, magnetic heads, crystal growing, solar cells, solar modules, and thin-film panels. These are solved problems.
Looking at common methods employed today, intuition does not work well when more than a few variables are in play. Most analyses are laden with subtleties. Single-product factories quickly become multi-product factories with embedded development lines. Companies are constantly ramping products up and down. One company might adopt cash flow as the most critical metric, another company might choose internal rate of return (IRR). Quick and dirty spreadsheets used for simplified “greenfield” situations have proven woefully inadequate in real-world situations.
Using a simple concept to illustrate the way forward, let’s look at changing a material at a specific process step. How does this change the cost? There are several possibilities. The material itself has a cost. The material may impact the tool it is used on in several ways, such as reliability, preventive maintenance, or throughput. It may impact other materials or waste disposal. It may impact multiple processes, not merely the one process where the material is used. Some of these impacts, such as a change in tool productivity, can change the factory physics. It may impact yield, the value of the finished unit, and even the probability of a failure in the field. One modest change can have a large number of impacts. In practice, while it is possible to model everything, most decisions require far less rigor.
When a potential change is identified, the next step is to determine the areas of impact. There is a tendency to limit the analysis to areas where it is easy to gather data, while avoiding what might be more challenging or time-consuming areas of data collection/analysis. This is where we find out if management is serious about optimizing the business, since several things become apparent rather quickly. Does management provide adequate time to do a proper analysis? Are the necessary resources, tools and/or experts available to do the analysis? Are data sets readily available that allow the analysis to be performed more efficiently?
The proper use of operational models involves determining the right set of items to consider, not merely what is easiest. This is done by using a list of questions to guide the user in setting up the analysis. In the previous example, one question is, “What other process steps might change based on the original change?”
Once the set of questions has been developed, the next step is to determine the appropriate analysis tool. To analyze a process step, where the issues are essentially self-contained within the step, CoO is a very effective tool. To look at sequences that are self-contained, a different form of CoO can be used. If the entire factory — or a significant portion of it — is impacted, then factory-level modeling tools are required.
SEMI and Wright Williams & Kelly Inc. (WWK) will present their annual “Understanding and Using Cost of Ownership” course Thursday starting at 9 a.m. at the San Francisco Marriott Marquis. The one-day workshop will provide semiconductor and solar industry attendees with an expert understanding of CoO and overall equipment efficiency (OEE). From fundamentals and industry standards (E10, E35, E79), to applications and misuses, attendees will gain practical knowledge they can immediately employ. Use the latest generation of software developed for Sematech to examine your own equipment, materials and factory conditions.
To sign up for the workshop, visit the SEMI booth just outside the entrance to Moscone North.
— David Jimenez, President, Wright Williams & Kelly Inc.