Industry News, Trends and Technology, and Standards Updates

Anticipating and promoting the increased adoption of SEMI’s Equipment Data Acquisition (EDA / aka Interface A) standards, we’ve posted a number of blogs over the past 12 months to address questions that potential stakeholders have repeatedly asked across the value chain. These postings have dealt with everything from the factory applications enabled by EDA to the best practices for OEM implementation of these standards to the development of robust equipment purchasing specifications.

Since the adoption process has now clearly reached critical mass, we must seriously address the question “How are we going to test the equipment and systems that incorporate these standards?” in a way that supports the entire industry. It’s an excellent question, and one that has a multi-part answer.

Given the structure and expected use of the EDA standards, the acceptance testing process for a unit of semiconductor manufacturing equipment will include at least three components, each of which addresses a different aspect of the standards. Note that we’re explaining this from the perspective of the end customer in a semiconductor factory, since this is the most common use case, but most of the same principles apply when testing EDA client infrastructure/application components as well.

• Compliance testing – does the equipment adhere to the specifications described in the SEMI Standards, and were these specifications interpreted correctly? Will it cleanly connect to the EDA client infrastructure without modification or extensive configuration?

• Performance and stability testing – does the equipment meet the end users’ performance and availability specifications in terms of data sampling intervals, overall data volume transmitted, size and number of DCPs (data collection plans) supported, demands on the computing/network resources, up-time, etc.? Will it support the range of application clients expected in a production environment?

• Equipment metadata model conformance testing – does the equipment model delivered with the interface represent the tool structure and content anticipated by the end customer? If the customer has requested that SEMI E164 (EDA Common Metadata) be fully supported, does the metadata model meet these specifications?

Of course, in addition to the requirements dictated by the standards themselves, most advanced semiconductor manufacturers will have a number of factory-specific requirements that must also be supported by the EDA interface. These may include special events and data for particular automation schemes, vectors of process parameters to support fault detection applications or other feature extraction algorithms, synchronization signals for external sensor integration, and the like. To address these requirements efficiently, an EDA test system should be extensible by its users.

You can see how interesting and vital this topic becomes when you consider the range of requirements outlined above. We’ll explore each of these in more detail in the next few postings, so stay tuned!

 In a previous blog we mentioned that two new SEMI standards, E172 and E173, demonstrated that the GEM standard was alive and well and even gaining new momentum by evolving to adopt new technology. The earlier blog focused on E172 with its SEDD files that use an XML schema to describe what is in a GEM interface. Today’s blog is about the E173 Specification for XML SECS-II Message Notation: a new way to log and document GEM/SECS messages, again using an XML schema.

A few years ago Cimetrix was involved in a project prototyping Wait Time Waste concepts and implementation alternatives. This work required Cimetrix engineers to review and extract data from many different SECS-II message log files from a variety of sources, and in the process, exposed a serious weakness in the industry. Because there was no standardized notation for logging SECS-II messages, everyone represented them differently, using different nuances and variations in their notation based loosely on SML (SECS Message Language, which is mentioned in the GEM standard). Additionally, SML itself was designed primarily for human readability, and certainly not for consumption by software programs; moreover, you can’t analyze a long message log without software to do the parsing for you. As a consequence, writing software to review the log files and to extract meaningful data from the log files was far more difficult than it should have been – SML and SML-like notations are simply not suitable for today’s needs. But now there is a suitable, industry-standard alternative. 

At Cimetrix we have utilized various notations for logging SECS-II messages for many years. In order for any notation to be useful it must meet certain criteria. First of all, it has to be easy for software to write (serialize). Secondly, it also needs to be easy for software to read (deserialize). And finally, it should be easy for humans to read and understand.

The original technique we used many years ago was based on the scripting language Tcl (pronounced “tickle”), which uses curly braces as structural delimiters. When programming within the Tcl language, this works very well. In other programming languages, however, it is easy to serialize, but not so easy to deserialize. Another technique Cimetrix had used for a few years was based on XML, which is well supported by all modern programming languages and an integral part of most internet activity. It is very easy to serialize and deserialize. And when formatted with carriage returns and indentation, it is quite easy to read for most humans (at least the ones who are software programmers or web page gurus).

Here is a subjective comparison between the notation alternatives using a scale of 1 to 5 where 5 is excellent and 1 is very poor or difficult.

At Cimetrix we decided to leverage our experience with XML, SECS/GEM standards and the SEMI Standards organization and related communities to develop a notation that everyone in the industry could benefit from. The result was this new standard: SMN. It is comprised of two parts: an XML schema defined specifically for GEM/SECS messaging; and a specification document describing how to use it (although many details of the specification are embedded as annotations within the XML schema file itself). It looks like this:

The schema is found on the SEMI website: http://dom.semi.org/web/wstandards.nsf/complementaryfiles

SMN brings the representation of SECS messages into the Internet era by defining an open, standard, XML-based notation for these messages. So what can you do with this? Here are some ideas:

• Document individual SECS/GEM messages (the SEMI E172 SEDD file uses SMN for this). You can also document entire message scenarios.

• Log individual SECS/GEM messages or scenarios in XML format. These can include only the messages, or might also include protocol messages (like the HSMS separate message).

• Share message logs with others. If their software supports SMN, they can immediately make use of it. This should increase collaboration in the manufacturing community, particularly between equipment suppliers and their customers.

• Embellish log files with comments and meaningful metadata, like data item names, variable names, collection event names, etc.

• Analyze and extract information from log files offline for projects like Wait Time Waste, where you don’t need to process a live data stream.

• Log messages in a raw binary format to save disk space, yet encapsulated in XML for convenience.

• Many of the numerous XML tools in the software development community can now be used by SECS/GEM software developers. This opens up a world of opportunities.

• Products like our CIMConnect and CIM300 can make use of SMN to make it easier to implement GEM and GEM300 interfaces on the equipment by using the SECSData element from SMN to pass data from the equipment supplier’s software into our product.

It is exciting to see the GEM standard evolve and embrace new technologies like XML to make integrating manufacturing equipment into the factories easier and easier.

For more information about these latest standards, and how you can incorporate them into your interface implementation, please contact us.♦

Interest in SEMI standard E30, known as the GEM standard has grown in recent years. That interest has increased in various manufacturing industries has matured in which factories are seeking to increase automation and carefully monitor equipment activity in order to increase production and product quality. Initially, the GEM standard targeted just the semiconductor industry, but then expanded to include any industry that used manufacturing equipment. In fact, a number of years ago the name of the GEM Standard changed from the “Generic Model for Communication and Control of Semiconductor Equipment” to “Generic Model for Communication and Control of Manufacturing Equipment.” Adoption by other industries is possible because the GEM standard defines generic features to control and monitor any manufacturing equipment. The GEM standard technology is not limited to semiconductor manufacturing. Over time, other industries have taken notice, especially as they try to develop increased control over their equipment.

CIMConnect™ is the best software product on the market for implementing the GEM standard. Of course CIMConnect supports all of the required GEM features as well as additional capabilities. This even includes excellent support of the Spooling feature, which saves messages that are otherwise dropped during a loss of communication. Early implementations of the GEM standard by others gave the GEM standard’s Spooling feature a bad reputation. This reputation is undeserved when Spooling is implemented by a robust product like CIMConnect. In CIMConnect, not only does Spooling work; it works well. It has been proven by customers that CIMConnect’s Spooling implementation does not lose any messages—even while under high-stress conditions. This means that when using CIMConnect, the Spooling feature can be used to effectively preserve critical data from the equipment.

Another feature that makes CIMConnect the best GEM software product is the CIMConnect Control Panel. In the new CIMConnect release, this application was completely rewritten, redesigned in .NET,  giving it a modern look and feel while adding lots of new convenient functionality. With other GEM products, the GEM interface is essentially a black box. With CIMConnect, however, the Control Panel application gives full visibility into the GEM interface. And you can run it at any time during GEM interface development and also during production. This means that you can see what reports and traces are defined, the link between reports and events, the status of all state machines, the state of each alarm, the enable status of every event, the history of occurring collection events, the history of alarm state changes and the current values of all data variables, and status variables and equipment constants. You can also view and capture the SECS-II message logging at any time for scenario diagnostics. Additionally, the CIMConnect Control Panel provides features to simulate the occurrence of collection events, collection event data, alarms, and variable data; thereby making it a built-in simulator included requiring no additional effort. And when you are ready to update your GEM documentation appendix with the list of defined collection events, alarms, status variables, data variables, and equipment constants, use the documentation builder feature.

CIMConnect has also already adopted use of new SEMI standard E173 Specification for XML SECS-II Message Notation (SMN). CIMConnect allows software applications to use SMN notation both when providing variable values to CIMConnect, as well as when getting variable values from CIMConnect. This means that you can pass data around in XML, retaining the data type and data structure information; bringing the convenience of XML into the SECS/GEM technology. You can log the GEM communication messages using SMN format making log messages much more useful, and they are able to be easily deserialized by any software applications that has XML libraries.

For additional detailed information about CIMConnect or to request a product demonstration, please contact us.♦

As the SEMI GEM standard celebrates its 25^th birthday, you may have thought its evolution had just about run its course — but you’d be wrong. Last year, the Information and Control Committee of SEMI Standards passed two new standards that enhance the usability of the entire SECS/GEM suite of standards for equipment suppliers and semiconductor manufacturers alike: E172 SEDD and E173 SMN.

Let us talk about the first of these, the E172 Specification for the SECS Equipment Data Dictionary (SEDD) and postpone E173 Specification for XML SECS-II Message Notation (SMN) discussion for another blog. SEDD standardizes the approach for documenting an equipment’s GEM interface in a way that is both human- and computer-readable. All factories in every industry that use GEM require their equipment suppliers to provide GEM interface documentation in some electronic form for each type of equipment. This is because the GEM interface on every equipment type is unique, supporting unique features and publishing a unique set of data. Of course, the GEM standard itself requires documentation and what has to be in the documentation but does not specify how this is to be accomplished. Until now there has been no common approach or format. This has always left the equipment suppliers to come up with their own format. At best this might be in a multiple-tabbed Excel spreadsheet or a PDF file; and at worst a text document that might or might not have been accurate or even complete. And every equipment supplier completes the documentation in a different structure and style so that no two GEM documents look the same. In summary, everyone is trying to complete this GEM and factory requirement by providing documentation, but in the end what factories are receiving has to be consumed and digested differently based on the equipment supplier, and sometimes even based on the specific equipment type from the same equipment supplier. It is a lot of work for the factory just to understand exactly what is in each GEM interface.

SEDD was created to solve this problem by defining a standard XML schema for documenting a GEM interface. Equipment suppliers create an XML file that complies with the SEDD XML schema to document the GEM interface and then deliver this XML file (called an SEDD file) to the factory.

Why XML? Because XML is the perfect technology for organizing data into a uniform structure that is well supported by modern programming languages. This means that equipment suppliers can use a software program to generate the SEDD file. It also means that factories can write software to read and view the SEDD file. Moreover, they can create intelligent host applications that automatically configure themselves and adapt to a specific GEM interface.

So what’s in an SEDD file? Below is a visual representation of the SEDD file schema, identifying the major elements of the SEDD file.

So essentially the SEDD file includes a list of the data available for collection by a host, some general information about the equipment (in the header), and the format of the data variables, status variables and equipment constants. As an example of what details are included, here are the details for collection events.

As an example, for a collection event, the SEDD file includes a list of all collection events available, and the ID, name, description, related SEMI standard, and the list of related data variables and other variables for that collection event. This is everything you need to use a collection event.

So far this is a summary of what is available today in a SEDD file. Cimetrix is leading the GEM300 task to extend the SEDD file to include additional information. This work is in SEMI ballot 5872 that proposes to extend the SEDD file to also include:

• A list of supported SECS-II messages and the acceptable format for each message (using E172 SMN)

• A list of support remote commands and available parameters for each remote command

• A list compliance tables for supported SEMI standards

• The list of predefined event reports

This is all work that was postponed from the original SEDD standard development. Hopefully ballot 5872 will pass and make SEDD files even more useful. With this additional information an SEDD file would empower GEM host software to configure itself to fully communicate with a GEM interface and make all of the features in the GEM interface available.

This is one example of how GEM technology just keeps getting better. It is not surprising that GEM is getting used in more and more industries.

For more information about this latest standard, and how you can incorporate it into your interface implementation, please contact us.♦

Since the concept was first articulated in 2011 by a German government-supported program promoting deeper integration of manufacturing software and hardware across the production value chain, the term “Industry 4.0” has gained recognition and momentum as the rallying cry for the 4^th industrial revolution (see left Image by Christoph Roser at AllAboutLean.com). Wikipedia  summarizes it like this: “Industry 4.0 facilitates the vision and execution of a ‘Smart Factory.’ Within the modular structured Smart Factories of Industry 4.0, cyber-physical systems monitor physical processes, create a virtual copy of the physical world, and make decentralized decisions. Over the Internet of Things, cyber-physical systems communicate and cooperate with each other and with humans in real-time…” 

This definition may lead you to ask “What aspects of Industry 4.0 are truly revolutionary, and what technologies and tools are available today that would enable me to start building “Smart[er] Factories?” In this blog, I offer some potential answers to these questions that put the vision of Industry 4.0 within reach for automation practitioners familiar with the latest generation of SEMI Standards.  

Semiconductor manufacturers have been collecting and using data from the equipment in their factories for decades. Throughout this period, device sizes and process windows have shrunk continuously according to Moore’s Law, and the SEMI Standards have evolved by necessity to support the insatiable demand for data exhibited by the process analysis and control applications that keep a modern fab running profitably (see left). The newest of these standards, the Equipment Data Acquisition suite (EDA, also known as “Interface A”), provides the power and flexibility to support a wide range of critical manufacturing applications and human users with ever-changing requirements; moreover, these standards can be deployed in a variety of system architectures without disturbing the “command and control” capabilities of existing factory systems.

“What does all this have to do with Industry 4.0?” To understand this, let’s look at the foundation of a “Smart Factory,” the collection of the many thousands of devices that might need to communicate over the so-called “Internet of Things.” 

We already see evidence that the availability of low-cost, low-power, networkable computing hardware will likely result in an explosion of “smart sensors” and other intelligent devices on the factory floor. However, as social scientists have observed over the millennia, groups of smart individuals don’t necessarily exhibit smart behavior in the aggregate, so what additional attributes must these devices possess to be good citizens of a collaborative, Industry 4.0 environment? How will these devices communicate effectively with one another? And what oversight will be required to ensure this communication achieves the ultimate manufacturing objectives?

As a starting point, I propose that each device, or manufacturing “thing,” at a minimum should be discoverable, autonomous, model-based, self-aware, communicative, and well-behaved. Depending on the role the device must play, it might also be self-monitoring, capable of defending itself (secure), and a consumer of data from other devices/systems as well as a provider. So defined, these devices would need a minimum of external monitoring and supervision (read “management overhead”) to perform their basic functions, but would rely on higher-level systems to provide specific objectives, instructions, and constraints (read “configuration, recipes, and limits”) for their operation in a given context and timeframe.

I realize that’s a lot to absorb at once, but now imagine that each of these devices could implement a subset of the services called for in the EDA standards, especially those defined in E120/E125/E164 (equipment modeling and standard metadata modeling), E132 (session management), and E134 (data collection management). Consider the collaboration among independent devices and systems this would enable…and ask yourself, how much closer to the vision of Industry 4.0 can you possibly get?

I hope the ideas above were useful…or at least thought-provoking. We’ll be developing this theme further in the coming months, but I wanted to use this blog as a conversation starter. We’d love to hear your feedback, so give us a call, or feel free to reach out to us.♦

Cimetrix participated in the recent European Advanced Process Control and Manufacturing (apc|m) Conference, along with more than 130 control professionals across the European and global semiconductor manufacturing industry. The conference was held in Reutlingen, Germany, a picturesque city of stone and half-timber buildings just south of Stuttgart.

This conference, now in its 16^th year, is one of only a few global events dedicated to the domain of semiconductor process control and directly supporting technologies. The conference’s attendance this year was comparable in numbers and demographics to that of the previous two years, a clear indication that this area continues to hold keen interest for the European high-tech manufacturing community. Another highlight this year was the sponsorship of Bosch, a relative newcomer to the conference but a pillar of the German manufacturing industry. Reutlingen is home to Bosch’s automotive electronics division and its related semiconductor manufacturing facilities, so they were very well represented in the conference and excellent hosts!

Cimetrix was privileged to make two presentations at this year's conference. The first was entitled “Data Fusion at the Source: Standards and Technologies for Seamless Sensor Integration,” authored and delivered by myself. The external sensor integration and related data unification topics have enjoyed increasing interest over the past year, and even though the techniques outlined in the presentation leverage the latest versions of the Equipment Data Acquisition (EDA)/Interface A standards, they apply equally well for the 200mm manufacturing nodes prevalent in European wafer fabs and assembly/test factories. The solution architecture is shown in the slide below, but for the background and rationale behind this approach, feel free to download a copy of the entire presentation from our website by clicking on the link below.

The second presentation, entitled “'Smart Manufacturing' solutions for high-mix manufacturing using Wait-Time-Waste improvement opportunities” was authored by Jan Driessen, a Principal Industrial Engineer with NXP Semiconductor in the Netherlands. It summarized the work of a project team from six companies and as many countries, and funded by the European Union's “integrate” program (cover page is on the left). Because of an unexpected work conflict during the conference, however, Jan was unable to attend, and, based on our companies’ shared interest in the Wait-Time-Waste technology and standards over the past several years, he thought that Cimetrix would be well qualified to give his presentation. I willingly agreed, worked with Jan to make sure I understood the latest material, and made the presentation. It essentially makes a compelling case for using equipment event data in a legacy 200mm fab to improve OEE, operational effectiveness, and factory capacity through a “chain of data operations” paradigm that he explains in some detail. The good news for 300mm fabs is that these same results can even more readily be achieved, because the availability and fidelity of the event data is much higher, especially if the fab has a full GEM300/EDA E164-compliant system infrastructure. For more information, request a copy of this presentation directly from Jan Driessen at jan.p.driessen@nxp.com.

Other themes that were evident at the conference included 1) applications of APC and supporting metrology techniques for structures found in smart sensors, MEMS devices, LEDs, and other semiconductor products outside the traditional processor and memory segments; 2) increasing emphasis on equipment data collection in the back end to support productivity monitoring and control applications; 3) unit process control for a number of equipment types; and 4) an entire session devoted to industrial engineering topics.

As with other similar conferences around the globe, the takeaway for Cimetrix is that “Smart Manufacturing,” Industrie 4.0, the Industrial Internet of Things (IIoT), advanced process control and fault detection applications, “big data” analytics, and a host of other high-tech manufacturing technologies all depend on the ability to get the right data at the right time from the right sources on the factory floor, and then make it available wherever and whenever needed… For more information about how Cimetrix’s product families that directly address this “sweet spot,” please contact us.♦

On March 30, 2016, Linkgenesis and Cimetrix hosted an Equipment Data Acquisition (EDA) standards seminar in Suwon City, South Korea to introduce equipment suppliers, sub fab component suppliers and system integrators to the concepts, history and best practices involving the SEMI EDA standards, also known as Interface A. The seminar was in response to Samsung’s EDA pilot targeted for Line 17 in Hwaseong with further rollout in Pyeoungtaek. While Interface A is being adopted in the US, Europe, Japan, and Taiwan; this is the first usage in South Korea. Recognizing this, Linkgenesis used their strong contact base in the South Korean semiconductor industry to bring together the major fab suppliers to share knowledge about EDA and demonstrate how Linkgenesis and Cimetrix can help them meet Samsung’s requirements.

The agenda was:

1. Introductions by Sungwoo Jung, CEO Linkgenesis and Eric Ko, Sales General Manager

2. EDA SEMI Standards Overview by Inhyeok Paek, Managing Director Research and Development Center

3. EDA Industry and Market Trends by Dave Faulkner, Cimetrix Executive Vice President

4. Factory Use Cases for EDA by Brian Rubow, Cimetrix Director of Client Training and Support (and co-chair of SEMI North American DDA Task Force)

5. Best Practices in EDA Implementations by Brian Rubow

6. CIMPortal Plus Feature Overview by Brian Rubow

7. Development Steps Using CIMPortal Plus by Mingyu Chung, Linkgenesis Principal Engineer

8. Question and Answer Period

With over 70 attendees, the Q&A session was lively with many good questions and exchange of ideas. Action items and next steps have been established with the attendees as everyone involved is working toward helping the Samsung pilot be successful.♦

If you would like to learn more about the application of the SEMI EDA/Interface A standards, click here for 10 in-depth videos on EDA technology. You can also download a white paper on the SEMI EDA/Interface A standards here.

Cimetrix and our new distribution partner, Linkgenesis, participated in a joint booth at SEMICON Korea last week in Seoul. With over four exhibit halls located in the COEX Convention and Exhibition Center, SEMICON Korea is the largest SEMICON in terms of number of visitors—about 40,000—and included over 1,870 booths in 36,000 sq. meters of show space. The theme of the show was “Connect to Future, Market, People and Technology” with keynote speakers from leaders at Synopsys, Texas Instruments, and Audi. The show was co-located with LED Korea. Included in the schedule was also a meeting for the Information and Control Technical Committee, Korea Chapter, to work on SEMI standards.

Korea remains the second largest equipment market for the second year in a row, and represents the largest region of installed 300mm fab capacity in the world. This show has more equipment manufacturers, in my opinion, than most SEMICON shows. So, because of the show’s location, it made it an ideal opportunity for us to meet with our current worldwide equipment manufacturing customers who seemed to be present in every row, and to meet the growing list of Korean equipment manufacturers as they build volume and increase their expertise. We already have several customers in Korea and expect this number to grow over the next few years.

Our new partnership with Linkgenesis was founded on providing EDA/Interface A solutions in Korea; currently Samsung and Hynix are both looking at incorporating this technology into their manufacturing systems. As the world leader in EDA/Interface A, Cimetrix can provide superior expertise to make the adoption process efficient and successful. Linkgenesis already has business with most Korean equipment manufacturers, so, by combining forces, we bring strong local engineering support together with our world-class EDA solution, CIMPortal™ Plus.

To accelerate our penetration into Korea, we have hired Mr. Hwal Song (+82 (0) 10-5058-0895) as our Korean General Manager to accelerate our customer introductions and partnership with Linkgenesis. Song has a long history in the semiconductor software industry.

We met with several new potential customers during the show setting a strong foundation for our continued growth in the Korean market. We left the show feeling very positive about Cimetrix’ entry into Korea.

At the Information and Control Technical Committee meeting, David Francis, Director of Product Management, represented Cimetrix. In the meeting there was an introduction to work being done by SEMI Japan related to Role Based Access Control (RBAC); however, this effort is still under development by the Japan Task Force. The North America DDA Task Force is starting discussions about Freeze 3 of the EDA Standards. Tom Salmon, Vice President Global Member Services and Standards, introduced information about the SEMI Automation Technology Committee that is working on the Smart Manufacturing initiate in support of efforts coming out of Industrie 4.0 and Industrial Internet Consortium. There was also an information share on the status of efforts around standards for Recipe Management Systems that address some of the concerns with the existing standards such as Recipe and Parameter Management (RaP).

SEMICON Korea 2016 as a whole was a success for both Cimetrix and the Korean semiconductor industry. Seoul was a great backdrop for the show and we look forward to returning next year with greater market penetration and the chance to catch-up with even more clients.♦

To be contacted about Cimetrix' CIMPortal Plus or any of our products and/or services, please click here.

In our November EDA-related blog, I covered highlights of the Factory System Infrastructure topic shown in the figure below, and emphasized the need to have a long-term architectural vision to guide the development of a scalable data collection and management environment. Today’s topic completes the picture by summarizing the kind of Manufacturing Applications that can leverage a factory-wide EDA implementation. Unlike infrastructure software alone, these applications are what really provide the ROI for the process engineers and other factory customers of the manufacturing IT department’s efforts, so it is important to understand the scope and requirements of these key applications early in the strategic planning process.

Even though Cimetrix is principally in the business of providing software products that enable equipment suppliers to provide data using EDA technology to the factory application developers that use the information in their production systems, we’ve been involved in this process for many years, and have a good idea of the dominant uses of this data to improve manufacturing Key Performance Indicators (KPIs). So in this blog, I’ll cover a little of the high-level picture of what applications fully leverage EDA data.

First and foremost, it is very easy to connect a basic EDA client to a piece of equipment, upload its metadata, and collect information about that tool’s behavior, so implementing a generic “quick-connect production monitor” independent from the fab-wide data collection system is a very common use for EDA. Moreover, if the model in the tool is compliant to the E164 (EDA Common Metadata) standard, you can make a lot of assumptions about the names of the modules, the wafers, the substrate locations, the process jobs, etc., since all of this information is standardized. As a result, you can quickly get an idea of what the equipment is doing, what recipes it is running, what wafers are being processed, and how well the tool is performing with no custom software whatsoever.

Once this is accomplished, the next step most process and equipment engineers take is to more fully characterize the tool’s behavior, so a very common use of EDA is simply improving equipment and process visibility. By inspecting the equipment model, you can see all the events and parameters that are available to be collected, plot them in Excel or on real-time strip charts, or pass them to other analysis applications.

After the equipment has been characterized, the first major production application most fabs will implement is multivariate fault detection (MVA FDC). This is actually the predominant application of EDA data in the industry to date, because in order to do well-architected fault detection applications, one must “frame” the trace data very carefully. High-speed data collection is usually only required in a small number of specific recipe steps after certain conditions have been established, so you can use EDA’s powerful event-based trace data collection to frame the precise data you want, and pass that on to the multivariate control and fault models.

Of course, once you understand a tool’s behavior and have good fault detection capability, you then start to use EDA data to compare tools across a fleet. You would normally want a set of similar equipment to behave in the same way, but perhaps you have one tool that performs exceptionally well, and you’re not quite sure why…In this case, you do what’s called a “golden run” analysis on that equipment, and compare the key trace variables in one with like variables in similar equipment to see where the differences are, and try to explain why those differences exist. Other names for this class of applications include chamber matching and tool matching.

Another key application that we’re starting to see significant interest in is external sensor integration. Factories are now starting to use EDA to present information collected from independent sensors alongside the information collected directly from the equipment. Sharing a common equipment model across these systems effectively “unifies” that data, so the downstream analysis applications believe the information was collected from a single, integrated source. The EDA metadata model offers an ideal way to accomplish this unification.  

Finally, in many advanced wafer fabs, it is important that substrates do not “sit around” after they’ve been processed. Minimizing inter-process wait times is especially important for some advanced processes, so knowing a priori—the precise moment that a lot is going to complete—is a critical capability so the material handling systems can be scheduled to pick up that material and take it to the next process. EDA provides an ideal way to make these predictions generically for multiple process types using the information that is required in the equipment model.

We’ll address these last two applications—external sensor integration and lot completion estimation—in more detail in later blog postings, but I wanted to get you thinking about these ideas early in the discussion of real EDA usage in semiconductor factories.

There are many more EDA application ideas and examples we could share at this point, from component fingerprinting to wait-time waste analysis to dynamic sampling for wafer-level feedback control to feature extraction for predictive maintenance…but these just scratch the surface of what factory customers will come up with once they experience firsthand the flexibility and power of EDA in their factories. More later as this creative process unfolds!

To schedule a time to discuss your EDA needs, click here to set-up a time to talk with one of our knowledgable experts.♦

In my October 27th blog, I wrote about the Equipment Automation topic shown in the figure below and stressed the importance of developing good equipment purchasing specifications from the outset to ensure the company’s manufacturing objectives can be met. Given the number of EDA pilot and production projects currently active across the industry, it’s likewise important to consider what kind of Factory System Infrastructure will be necessary to support a full-scale EDA deployment… so the purpose of this posting is to highlight this topic for the semiconductor manufacturing IT professionals who may face these challenges soon.

However, before diving into a detailed design process for an EDA factory system, you must decide what overall system architecture will govern that design. A number of factors go into this decision, including 1) the functional requirements that distinguish EDA-based data collection from other more traditional approaches, 2) technology constraints of the existing factory systems, 3) budget limitations, 4) schedule requirements, and especially 5) the non-functional requirements (scalability, performance, reliability, ease-of-use, etc.) that often make the difference between success and failure of a given system.

Each of these factors deserves a thorough treatment of its own, but since we were invited to address this topic at a recent seminar sponsored by SEMI Taiwan, we’ve assembled an overview presentation entitled “Factory Systems Architectures for EDA” that you can use as a starting point. It not only covers in more depth the requirements above which drive key architectural decisions, but also suggests what some of the major architectural components of a production system would need to be, based on the experience Cimetrix has gained working with the earliest adopters of EDA across the semiconductor device maker and equipment supplier communities. These include provisions for handling the scores of equipment metadata models that will exist in a production facility, for creating and managing the thousands of data collection plans that are resident at the equipment instances themselves, for monitoring and maintaining the overall performance of a system with such inherent flexibility, and for a number of other examples. Finally, the presentation describes some high-level examples of architectural “styles” that have been implemented in the industry thus far.  

We sincerely hope you will download this presentation and its companion “The Power of E164: EDA Common Metadata” that was also presented at the SEMI Taiwan event, and contact us when you want to know more about any of these topics.♦