MES Technology Enhancements that Mitigate the Big Data Challenge An Industry White Paper By Marty Moran, MES Product Marketing Manager, Aspen Technology, Inc.
Big Data has become an important issue in the IT world during recent years and for good reason. Data seems to be everywhere exploding faster than we have time to absorb it all. And, if anything, the problem is only going to get worse. In fact, Aberdeen estimates that the amount of data is increasing at 36% per year. 1 The process industries have dealt with many problems relating to "Big Data". Even in a small refinery, there are literally thousands of instruments that a MES system collects data for in a plant, roughly once a minute, over the span of many years. Of course, there are data averaging techniques that are used to condense the total amount of data. But, no matter how you look at it, that s a mountain of information. In a certain sense, having lots of data is a good thing because we potentially have more information to use to solve real problems. But on the other hand, trying to succinctly collate that data into a proper format to make it actionable represents a real challenge that we consistently struggle with. So, how are technology companies responding to the Big Data challenge? Some people want to dedicate multiple servers to analyze data and develop sophisticated correlations, as in the Hadoop technology solution, an open-source framework used by technology giants such as Yahoo, LinkedIn, and Facebook. In some circumstances, that may be the right answer. However, oftentimes the answer to a complicated problem is to break the problem into manageable chunks, then use technology to make incremental improvements on different aspects of that problem capturing significant benefits in the process. Those efforts, in total, can add up to significant improvements without the overhead of very complex, technological solutions. Intelligent Search For example, take the seemingly simple problem of finding tag information in a typical process plant. Engineers that routinely work with a given process unit will, of course, be familiar with many of the tagnames for the temperatures, pressures, and flows in their respective areas. But, they almost certainly will not be familiar with all the instruments throughout the entire plant. So, if they are given a new technical assignment that requires tags they are unfamiliar with, how do they quickly find all the information that they will need? The traditional approach is to start with process graphics and try to locate the tagnames in question based on points in the graphic. That s a very reasonable solution if the tag happens to be on the process graphics. But, what happens when the tag isn t routinely used by operators and doesn t appear on a graphic? The user will likely try to locate the tag by using different wildcard searches in the MES system. After a few minutes, they will usually find the tag that they were looking for. But if that doesn t work, they will ultimately need to drag out the PID diagram to locate it. While this may be only one tag for one individual, when one multiplies the extra 5-10 minutes to find the tag across the typical size of a process plant, it s easy to see that all this searching takes time. Are there more efficient alternatives? The good news is that with recent technology improvements, some MES vendors are now providing Google like intelligent search capabilities within their MES systems to improve a users ability to find information. Most people use Google to quickly find websites they want to visit and can t imagine using the antiquated search techniques from the late 90s that were not nearly as effective as the search techniques that Google pioneered. Intelligent search capability within a MES system operates in a similar fashion as Google. For example, if you want to find the tagname for the bottoms temperature in the depropanizer, you would type something close to Temperature Deprop. As you type, it instantly starts picking out the best fit same as Google. At a more technical level, this type of search uses a faceted search functionality, also referred to as guided navigation. The facets are really discrete attributes. The beauty of faceted search lies in its ability to allow users to create their own custom navigation by combining various perspectives, rather than forcing them through a specific path. 1
Intelligent search also uses hit highlighting, which backlights what occurrences have been found as a result of the search. Backlighting found values makes it easier for the user to quickly see the matches and find what they are looking for. Figure 1: The intelligent search functionality in Aspen InfoPlus.21 organizes results by relevance and allows users to quickly visualize relevant information. While there are certainly time efficiencies using intelligent search for experienced engineers who use the system on a daily basis, some of the real efficiencies come from the many more casual users who don t know many of the tags. If casual users are more likely to find the information in a system, they are more likely to use the system to solve real plant problems, extending the investment in your MES system to more users. This is just one simple way that today s technology can be used to mine for data within a large MES database, but it s hardly the only way that technology is having an impact on deriving value from large amounts of data. Trending is another way to capitalize on technology. Enhanced Trending Capability Using process trending to solve plant problems has been a part of the plant engineer s tool kit ever since digital solutions were first introduced in the process industries some 30 years ago. It was standard fare for a typical engineer to create numerous sets of preconfigured plant trends that had the potential to view a small number of process variables on one trend. As process issues arose, the engineer would go back and forth between the pre-defined trends they had created ahead of time in order to be able to pinpoint the particular issue causing the plant problem. However, depending on the issue, it was quite common for the engineer to have to create new trends for the particular problem at hand, since it required a different group of variables to be plotted relative to one another. It was not conceivable, given the state of technology at that time, to be able to even remotely think about plotting all the necessary variables on a single plot at one time. 2
However, all that has changed. MES vendors now offer significantly improved trending capabilities that can almost instantly bring up over 80 variables over a long historical time period. This gives engineers the ability to put the problem into better context. If they have all of the variables that they need to peruse at their disposal, engineers are quite capable of spotting the necessary relationships between the variables in order to ascertain the sources of the plant problem. It certainly makes solving the problem much easier and faster than trying to switch back and forth between a series of much smaller, pre-defined plots or having to create new trends on the fly to get the necessary view to solve the problem. What s also great about these new trending capabilities is that they can be rendered on whatever device one wants web, smart phone, or tablet since they are based on HTML5 format. Figure 2: High Performance Trends display spark lines and trend previews, providing better context to data and reducing analysis time. Data for Specific Roles Another solution to the Big Data quandary is Role-Based Visualization. This allows users in pre-defined roles the ability to securely login to a particular workspace designed specifically for their role. Graphics, trends, and dashboards can then be customized specifically to the problems that are unique to that role. Once again, the idea is that no matter how much data is in the database, users will only see information that is relative to their role. The amount of data in the database no longer becomes a constraint since the user only views relevant information. Users can personalize their workspace by utilizing content relevant to their role with Aspen Role-Based Visualization. Figure 3: Users can personalize their workspace by utilizing content relevant to their role with Aspen Role-Based Visualization. 3
Mobile Solutions In combining the best of all worlds, some MES vendors are now offering mobile solutions available on smart phones, tablets, etc. that brings all of this functionality together. This includes the ability to quickly use intelligent search capability, high-performance trends, role-based visualization, and receive e-mail or text notification alerts. Figure 4: Aspen InfoPlus.21 Mobile displays critical, up-to-date information in various views for faster decision making. Conclusion While "Big Data" is a big problem if considered at a high level, there are many incremental approaches to being able to extract knowledge from the phenomenal amount of data in our MES systems making data actionable. The bottom line is that we want our engineers using their time to solve problems and make improvements not spending their time sifting through mountains of data. In summary, best-in-class companies are mitigating the Big Data problem by dissecting data into small manageable pieces and using technology to make incremental improvements. Often, this means that the solution to the problem does not necessarily need to be complicated in order to be effective. Intelligent search, high-performance trending, and Role-Based Visualization technology are some of those approaches. When multiplied across an entire organization, these approaches can have a significant impact on profitability. 1 AberdeenGroup, The Little Elephant in the Big Data World: Hadoop Goes Live, March 2012, Nathaniel Rowe, pg. 1. 4
About AspenTech AspenTech is a leading supplier of software that optimizes process manufacturing for energy, chemicals, pharmaceuticals, engineering and construction, power & utilities, mining, pulp and paper, and other mill products industries that manufacture and produce products from a chemical process. With integrated aspenone solutions, process manufacturers can implement best practices for optimizing their engineering, manufacturing, and supply chain operations. As a result, AspenTech customers are better able to increase capacity, improve margins, reduce costs, and become more energy efficient. To see how the world s leading process manufacturers rely on AspenTech to achieve their operational excellence goals, visit www.aspentech.com. Worldwide Headquarters Aspen Technology, Inc. 200 Wheeler Road Burlington, MA 01803 United States phone: +1 781 221 6400 fax: +1 781 221 6410 info@aspentech.com 2012 Aspen Technology, Inc. AspenTech, aspenone, the aspenone logo, the Aspen leaf logo, and OPTIMIZE are trademarks of Aspen Technology, Inc. All rights reserved. All other Regional Headquarters Houston, TX USA phone: +1 281 584 1000 São Paulo Brazil phone: +55 11 3443 6261 Reading United Kingdom phone: +44 (0) 1189 226400 Singapore Republic of Singapore phone: +65 6395 3900 Manama Bahrain phone: +973 17 50 3000 For a complete list of offices, please visit www.aspentech.com/locations