Validating machine performance and making use of machine intelligence is often described as a simple and virtually "built-in" capability of many software tools available today. However, experienced machine builders know that this task is anything but simple. It requires mapping data to the control system, and because there is no consistent, standard format for coding control system data, software tools designed to capture metrics involve a time-intensive, manual coding process.
With increased pressures to improve plant efficiency, manufacturers rely on their machine builders to deliver equipment capable of collecting a wide range of performance and operating data. With accurate performance metrics, such as Overall Equipment Effectiveness (OEE), plant management can identify problems and establish proactive strategies for reducing machine or production line inefficiencies and drive improvements across the entire manufacturing process. For example, production may know that they made 100 units, but without quality data, you'd never know there was a 20 percent scrap rate.
A programmer's primary challenge when mapping performance data is that there is no standard way to define the vocabulary and definitions related to machine states or event data. If you were to ask two engineers working on the same machine to define and describe the various machine states, you would likely get two completely different definitions and different terminology. For example, where one engineer might describe the ABORT state as a transition to the IDLE state, another engineer might describe it as the execution of HOLDING state logic. If you multiply that scenario across the universe of machine builders, it's easy to see how machine programming language can rapidly evolve into a tangled nightmare of custom code, with every new project adding to the complexity and confusion.
Other data setup and configuration challenges come from the lack of clear performance data specifications. For most machine builders, the conversation with end users regarding the required performance metrics and how they are defined is often limited and vague. For example, the specification provided by the end user often will simply state, "We need a machine capable of X throughput, Y quality, and fits into floor space Z." Much of the performance and OEE-type data is left up to the assumptions and discretion of the machine builder.
|In any machine design project, the time and effort needed to program complete machine control can be very time-consuming, error-prone and expensive. To address this issue, many machine builders incorporate programming standards that allow an integrated, more modular approach to machine control design. These same standards, along with the help of new configuration tools, increase programming efficiency, reduce engineering time, and simplify the setup and implementation of machine metric reporting systems.|
Whether a machine builder approaches each application from scratch or attempts to reuse machine code from a previous project, without a common methodology, inefficiencies are abundant. If each machine design is started as a completely new program, time and effort is duplicated recreating and testing code. Likewise, if programmers attempt to reuse existing machine code from a dissimilar machine, they will run into setbacks re-working the code that require more time and effort than starting from scratch.
Map to a Model
Many machine builders realize the best way to bring order and efficiency to the data mapping process is by following industry standards. For example, what if programmers were given a pre-defined data structure based on industry standards and best practices for each machine, along with starter code that was already written and tested for them? And what if they could use a state model as their base specification and add fill-in-the-blank information provided by the end user? It would be like using building blocks to write a program, rather than a blank screen. Efficiency would grow exponentially.
That's the intent behind new modular programming guidelines recently incorporated into the ISA-88 standard. The guidelines provide programmers with a broadly recognized and predictable machine state model and standardized data model to help ensure they speak the same language and use the same terminology consistently during machine design.
With the ability to reuse definitions, structures and lines of code, programmers can better leverage prior work. This is particularly important as machine builders continue to assume more responsibility for designing information-enabled machines pre-configured and accurately coded with the operating and performance data end users need.
|Curt G. Joa Inc., a Wisconsin-based machine builder, designs web converting machinery for major manufacturers of disposable household and medical products. Although these products have made many lives easier, keeping up with changes sought by consumers can be cumbersome and costly for manufacturers. Just recently, a diaper manufacturer approached Curt G. Joa Inc. to help upgrade a converting machine that was outdated and increasingly expensive to fix. |
"Fifteen years ago when this converting system was made, it was considered top-of-the-line," said Kevin Zeinemann, electrical engineering manager at Curt G. Joa Inc. "But at that time, plant-floor information was not a priority as it is today when executives look for production data to help them make business decisions."
The old diaper machine employed several third-party systems including a different HMI, drive system and I/O system, none of which communicated effectively with each other or with plantwide information systems. When the time came to add a new piece of end-of-line packaging equipment, the old system no longer had the functionality or the flexibility to accommodate the new addition.
To ease the transition to the new control platform, engineers at Curt G. Joa Inc. used the Rockwell Automation Power Programming tool. This standardized programming methodology allows users to leverage pre-engineered, ISA-88-based modules of code across multiple machines and applications, helping to reduce design and engineering costs by an average of 45 percent.
"Our partnership with Rockwell Automation and the superior functions of the Integrated Architecture™ system help us to innovate without risking the flexibility our customers need or slowing our design process," Zeinemann said.
For the manufacturers of disposable household and medical products, the flexibility and convenience Curt G. Joa Inc. provides means they continue to innovate in ways that make consumers' lives safer and easier.
Just as the machine builder finds value in having standardized tools and programming concepts, so does the end user. Having the same program structure, from concepts to coding, helps simplify maintenance and troubleshooting. Information is consistent, and exactly where operators expect it to be. Moreover, with a common understanding of the definition of each state, training is much more intuitive.
Another benefit of standard programming methodology is that it helps simplify the setup and integration of the end user's performance measurement software with the machine control system. With OEE and other performance data already defined in the machine controller using standard data models, configuring all of the individual performance parameter definitions in the analysis software is significantly easier. End users eliminate the need to add custom code, which reduces complexity and accelerates startup.
Even with machines built using a standard state model and data model, all of the performance metrics still needs to flow from the machine controllers to a database and reporting system. To establish this connection, the end user must define a plant model, set each event and state definition, and then manually navigate to and select each data point for uploading. With potentially several hundred data points per machine needing collection, configuring each performance parameter in the analysis software is a complex and time-consuming task.
Configuration tools are available to help solve this problem. One that was designed by Rockwell Automation for use with its FactoryTalk® Metrics performance measurement system allows engineers to locate the user-defined data types programmed in the controller and easily configure the desired performance metrics in the reporting software. The tool can read machine data directly from a single machine, or from a supervisory controller that contains data from multiple machines.
As more and more machine designs incorporate standardized code, the configuration tool meets a critical and growing market need for greater speed and simplicity in reporting system setup and installation. The tool provides several key functions:
- Identifies all of the machines that will have data collected by the software.
- Extracts all the data from the machine controller that need to be configured.
- Maps the selected communication topics to the work cells found in the software's plant model.
The configuration tool searches and finds the standardized data points within the machine control system and tells the software their location, definition and role. The tool can cut reporting system configuration time in half, and allow end users to set up a model of the machine's data inside the software and reuse that model for each new reporting system installation. Should the end user decide to implement a six-sigma/lean program leveraging the OEE data from its machines, configuring and integrating the reporting software would be greatly simplified since all of the programming work in the controllers is already in place.
The ability to quickly configure the analysis software provides important benefits to the machine builder as well, giving them fast access to a wide range of performance data needed to validate machine performance. With access to detailed performance data provided by the software, machine builders can benchmark the machine's performance in-house as part of their factory acceptance testing (FAT) prior to shipping to the end customer. Once installed at the customer's site, any perceived performance issues can now be quantified through the analysis software and compared to the benchmarks during FAT. The machine builder can do all of this remotely, provided it the appropriate network infrastructure and security access has been established.
Standardized programming helps machine builders leverage their investment in complex equipment while providing end users with programs that have the same look and feel, which translates into easier training. The machine builder and end user both gain -- where one finds value in increased design efficiency and faster time to market, the other benefits from simplified maintenance, faster troubleshooting and more seamless integration with performance measurement and supervisory systems. Together, they can reap the rewards of an information-enabled machine.
FactoryTalk, Integrated Architecture and Rockwell Software are trademarks of Rockwell Automation, Inc.