Standard Practices

End users and machine builders can reap big benefits from the use of standards, but challenges remain.

Standards play an essential role in manufacturing and technology, as evidenced by the number of organizations dedicated to creating and maintaining such documents. Indeed, the development of standards is one of the hallmarks of a maturing technology. It makes sense, then, that standards have an increasingly large role to play in the application of motion control.

Motion control is an enabling technology across a variety of industries. As a result, the standards applied to it tend to be those created around specific functions like controls and networking, or around industries and applications like pharmaceutical processing or packaging. Examples of the first type include the International Engineering Consortium’s (IEC’s) 61131 controls standard or the American National Standards Institute (ANSI) B11 TR-3 safety risk-assessment standard. Examples of the second type include packaging standards from the Packaging Machinery Manufacturers Institute (PMMI) or the Organization for Machine Control and Automation (OMAC).

Standards can bring significant benefits. For end users or machine builders, using components or systems that offer consistent performance and/or programming can simplify sourcing, integration, staff training, and maintenance. Depending on the company and scope of operations, return on investment (ROI) can be quite quick. For system or component vendors, standards represent a way to add value to their products and satisfy customer demand. True, implementation requires a certain amount of upfront engineering, but once performed, that work can be applied in a modular fashion to future designs.

Of course, no engineering effort would be complete without challenges, and the standards effort faces plenty of them. One of the biggest is the sheer variety and number of standards available. With the exception of the International Standards Organization (ISO) and the IEC, very few of these are global. Instead, standards vary around the world, as do requirements for compliance. “There is no one standard that is agreed on by Asians, Europeans, and North Americans,” says Sunny Ainapure, Servo and Motion Product Marketing Manager at Mitsubishi Electric Automation (Vernon Hills, Illinois).

“Machine builders are complying with standards that are global in nature but they're going to be a lot slower to adopt standards that are more regional, meaning only dominant in Europe, for example,” says Sloan Zupan, Controller and Network Product Marketing Manager at Mitsubishi. “The investment needed to build machines to meet U.S., European and Asian standards can be significant.”

In theory, to remain compliant, a machine builder would have to design a different product tailored to the specific standards for each region. I say in theory, because in practice, your average machine builder is unlikely to develop a full product line of compliant machines. Instead, they would probably produce a standards-compliant machine based on the particular requirements of a specific customer. Once the work has been done, they might offer that machine as a product, but they probably wouldn’t be unlikely to design it without the original order.

In some ways, the number of standards in existence is simply an example of human nature in action. “Everybody has their own opinion and those opinions change from one company to the next, from one industry to the next, from one country, one state to the next,” says Eric Hollister, Product Sales Engineer at Pilz Automation Safety (Canton, Michigan). “Some places, some industries, or some markets are further ahead than others, which leads them to adopt newer technologies that other groups don't agree with or understand.” Consider safety, in which the concept of safe motion is both a new trend and a source of some controversy.

Industrial standards in the United States tend to be developed by consensus. With one or two exceptions, they are considered guidelines, not law. The primary reason that the U.S. Occupational Safety and Health Administration (OSHA) does not enforce safety standards by rule of law is purely pragmatic:  standards are living documents and need to be able to change without requiring legislation to do so. From the perspective of OSHA, liability devolves to the end user. Companies can choose to follow a U.S. standard or decide that one from Europe or Asia standard is more appropriate.

"Safe speed, or safe motion, has been adopted as safeguarding method in certain instances in Europe but not yet in the United States," says Hollister. "That leads to newer, more innovative products that allow operators to work more closely with the machine in a safe manner, but we cannot yet apply them in the U.S., at least not as a standard, anyway.” A standard that defines Safe Torque Off may provide a method to remove power from the motor in a jammed packaging line, for example, but it is up to the operator to manually clear up the jam. A standard that defines a safe-speed condition with a hold-to-run device such as a foot pedal or pendant would allow the operator to use the power of the machine to back product out of a jam. It can also simplify hardware design. In the case of press brakes, for example, which are difficult to guard without compromising operator access to the pedal, safe speed can improve operations while maintaining safety.

Another emerging trend in this area is a focus on "lean” safety. The idea is to produce a machine that is safe for operators, product, and tooling alike, but to do it as economically and efficiently as possible. ANSI has two relevant documents: B11 TR-3, the aforementioned risk assessment standard, and B11 TR-7, which discusses techniques for incorporating lean engineering into safety design. “If you can lean out some of the hardware that's used in the safety solution and keep it as safe as it needs to be, you have lower costs,” says Hollister. “You have a more available machine that allows the operator to do his job without anything getting in his way.”

Standard Challenges
Launched by a group of end users, machine builders and technology providers, OMAC focuses on developing controls and communications standards to simplify the process of integrating disparate packaging machines together. The result of their efforts is a group of documents loosely referred to as Connect and Pack.

“There’s a common look and feel for both the operator interface and control logic that is running each machine, so it's easier for the end user to operate and maintain” says Jack Faett, Industry Manager, OEM at GE Intelligent Platforms (Charlottesville, Virginia). “It’s easier to move personnel around the factory floor because the machine interfaces always look the same. It's easier to train operators and maintenance personnel because of the consistency from machine to machine, across the line and across the plant.” It all saves time and engineering efforts, which add up to cost savings for the end user.

Machine builders and OEMs can benefit from standards, as well. Certainly, there are costs involved in implementing the standard, but once performed, that work pays dividends in the form of a set of modules that can be duplicated from machine to machine. Everything from code to templates to drawings for a given technology platform can be reused because they are consistent from machine to machine and customer to customer,.

Perhaps the key contribution of OMAC is PackML and the establishment of the state model. Designed to set up a standard method for hardware to communicate to software applications, PackML defines 17 possible common states for a machine: for example, executing, suspended, aborted, etc. (see figure 1). Users can access a single screen to check the status of the machine. More important, each machine in a line can have its own independent state model (see figure 2).

Consider a packaging line consisting of fillers, accumulators and palletizers. “They may all be in ‘execute’ mode at some point in time, then something happens on one of the machines " says Faett. “It will be in a ‘stop’ state but the others may still be running because there is enough buffer in between to allow them to execute while maintenance works on the stopped machine.” The PackML state model allows this type of multi-stage operation. The standard also supports an additional state model to represent the entire line as a whole, allowing an area manager, plant manager, or front office staffer to closely monitor the status of their assets. PackML is effective enough that it has been included as part of another standard, ISA-TR88.00.02: Machine and Unit State.

The next step is to move the standard [PackML/PackTags] up into the manufacturing execution systems (MES) layer, allowing it to feed into operational equipment efficiency (OEE) calculations. If all goes well, the plan is for the new standard to be released in the next year.

Standard Deviations
Perhaps the most difficult part of writing a standard is that the process requires input and cooperation from of a variety of people across all different backgrounds, ages, opinions, languages, and political persuasions. The process of road mapping, negotiating a draft standard, soliciting industry input, and iterating that loop enough times to develop a useful final document can be both time-consuming and expensive. Of course, the final document may be so detailed as to be opaque. “You get a group of engineers in the room to write a standard and a lot of times you will get into language that is so specific that people may lose track of the actual meaning behind it," says Hollister. "Interpretation of the standards is probably one of the biggest hurdles that people run into.”

Indeed, going from standard to implementation can be a significant enough challenge that the OMAC technology providers have developed application templates and quick-start implementation guides to assist machine builders in using the standard.. The idea is that the implementation guide/template package gives machine builders a 75 to 80% head start on the process. “The standards as they're written can seem a little daunting because of the depth of the definitions,” says Faett. “We’ve developed implementation guidelines that give a clear roadmap of what a machine builder or an end user needs to do implement a PackML/PackTags standard in their solution.” Because the solution of every technology is a bit different, each vendor needs to produce their own implementation guide to accompany the OMAC application template.

Which brings up the issue of variation from the standard?  Open up almost any motor catalog and you will find listings for NEMA-rated motors. Many elements of automation technology, particularly controls hardware and software, are not commoditized, however, which makes the task of developing standards even more challenging. You can buy a USB cable from any manufacturer across the globe and use it with no additional modifications. Not so with a motion controller or a programmable logic controller. Here, there is a significant amount of company-to-company variation, even when the product is designed to comply with a standard.

“The initial purpose of standards was to allow customers a certain level of interoperability, but in a sense, the implementation of a particular standard becomes almost proprietary because each automation supplier implements it in their own way to add additional value,” says Zupan. He points to the IEC 61131 standard as an example. If the code written by vendor A contains only IEC61131 compliant instructions and data types, a customer can easily use it with a component from vendor B. It’s rarely that easy, though. Suppliers add custom instructions to enhance performance, or allow the customer an opportunity to create their own instructions or data types (see figure 3). Once the compliant product departs from the standard, interoperability goes out the window.

For vendors, sticking rigorously to the standard means giving up the opportunity for a proprietary advantage. From the end-user standpoint, interoperability might seem desirable but in reality, it cuts both ways. On one hand, there is a benefit to being able to train once and engineer or operate all compliant solutions. On the other hand, forcing vendors into a single solution means less access to innovative solutions.

“It is very difficult for any one standards organization to regulate how a standard has been implemented, says Zupan.  "Our minds work differently, so the interpretation and the resulting implementation of the standard are going to be different.”

Looking Ahead
In order for standards to work, they need to pass some critical mass of industry adoption. That is the primary challenge for the groups involved. "Standards are important but adoption by the greater mass of end users is still not taking place," says Zupan. “More effective communication is needed to more clearly explain the benefits related to ROI and how end users can take the equipment they have today and implement the standards.”

This latter issue of noncompliant incumbent technology is a serious one. PackML, for example, can dramatically simplify the integration between multivendor hardware products and common software applications that need to know the real-time state of manufacturing assets. The value proposition is less clear for a single compliant machine being installed in a factory full of older equipment. “In order for an end user to really reap the benefit of some of these standards, they need to implement the standard across the whole manufacturing facility," says Zupan. “Having a single machine or two that meets one or two standards may produce less than the desired benefits.” The argument becomes even more difficult to make if the incumbent equipment still has a 10 or 20 year lifetime. There is a low business benefit to spending money to change something that is working already.  “At end of the day, how is that going to help customers increase revenues? What is the value to the end user? So far, that message has not been well articulated.”

Currently, the value of standards to the end user varies depending on the scope of their operation. A large end user with projects around the world needs efficient integration and fast communication with all assets in all locations.  For corporate engineering, a standard like PackML provides the tools to get the job done, which makes a compelling ROI argument.  A small end user who perhaps upgrades one or two machines a year will take much longer to achieve ROI, especially since a compliant machine will carry a price premium over a standard design.

Over time, in a form of industrial Darwinism, the most effective standards -- or those with the most influential sponsors -- will win out. As time passes and the proportion of noncompliant installed base shrinks, market adoption of the dominant standards will increase. The primary benefit will still fall to the end users, but a satisfied customer is a return customer, so the process will aid all participants.

Acknowledgements
Thanks go to Jim Wiley of Parker Hannifin for useful conversations.