Wireless Control in the Factory: In Search of a Standard

Determinism is still an issue for high-speed wireless control, but some factory users are moving ahead.

Mike Read
Mike Read
As reported in the article,"Wireless Control in the Process Industries: Blasphemy or Common Sense?", some end-users in the process industries are taking the first steps toward use of wireless technology for control applications. But what about the future for wireless control in the discrete manufacturing industries?

“The biggest concern that we have right now is determinism,” says Mike Read, senior technical specialist, IT—Plant Floor Systems, at automaker Ford Motor Co., in Dearborn, Mich.

Ford has tested a number of off-the-shelf wireless products for potential use in automotive manufacturing control applications, says Read. “What we find is that we can set up a [wireless] network that does exactly what we want, and maybe it just squeaks by on performance characteristics.” But then, for unknown reasons, the system will periodically require much longer update rates than normal, Read relates. “And that’s just not acceptable, because it could cause a tool to crash.”

Read spoke with Automation World as a follow-up to a paper he co-authored for ISA Expo last October in Houston, titled “High Value Wireless Applications for Factory Automation.”

Most of the Ford tests to date have used devices based on the Institute of Electrical and Electronics Engineers’ IEEE 802.11 wireless standard, and on IEEE 802.15.4-based Bluetooth devices. To meet machine control requirements, the automaker needs deterministic response of 10 milliseconds or less, a goal that has yet to be consistently achieved in long-term tests, says Read. But he expects that goal soon to be met. “And at that point, I would not hesitate to use wireless for control,” he says, though likely not in safety-related applications.

There are various factory automation applications for which wireless control could produce major benefits for automakers and other discrete manufacturers, says Read. The list includes robot end-effectors, where wire breakage in frequently flexing joints can be a costly and time-consuming problem, and festooned cable and commutating rail equipment, which is likewise prone to flex-induced cable wear and breakdown.

Rotating equipment found in packaging machines and dial index tables are additional prime candidates for wireless control; in these machines, repeated circulation around the slip rings used to transmit power and control signals to devices on the rotating assemblies can lead to wear and electrical contact failure.

Besides the still nettlesome determinism issues, a major barrier to control without wires in discrete manufacturing is the lack of a wireless factory automation standard, says Read. The WirelessHart standard was developed for the process industries, and while the ISA100.11a industrial wireless standard does not exclude factory automation, it is focused on the requirements of process manufacturers.

Pitching in

That’s why Read has been active in the International Society of Automation’s ISA100 Factory Automation Working Group 16 (WG16), which is addressing the wireless standards issue. Read’s co-author for last fall’s ISA Expo paper was WG16 co-chair Cliff Whitehead, business development manager for Milwaukee-based automation supplier Rockwell Automation Inc. An eventual wireless factory automation standard to be developed by the group is expected to join ISA100.11a as one in a “family” of ISA100 industrial wireless standards.

WG16 activities to date have focused on developing a “requirements document” that will define wireless requirements for factory automation from a user perspective, and will also identify technologies available in the market that could satisfy those needs, according to Whitehead. That document is expected to be completed early this year, providing “an opportunity to stimulate innovation, which is, frankly, already ongoing,” says Whitehead. “People are not waiting for this document to actually go off and innovate in the area of discrete control with wireless,” he observes.

What’s needed

Factory automation shares some common requirements for wireless with the process world, including reliability, interoperability, security and the ability to coexist with other radio frequency (RF) sources. But discrete manufacturers also have some unique requirements. At Ford, Read mentions several qualities needed to make wireless control work in the automotive factory automation space.

The process-oriented ISA100.11a standard is designed for applications in which latencies of 100 milliseconds can be tolerated. While that works for many process industry applications involving longer time constants, the discrete industries need faster update rates for machine control. “We want an update from that wireless device, either reading a sensor input, or, we want an output device to react, every 10 milliseconds, if there’s a state change,” says Read. Unlike process industry tendencies to deal in analog measures such as temperature or pressure, wireless communication in the discrete industries can be simpler, involving single-bit “on” or “off” signals.

Another major difference between factory and process automation involves need for wireless mesh, or multi-hop networks, says Read. While mesh networking is called out in process-oriented wireless standards, mesh is often not needed in factory automation, he says, and may, in some applications, be detrimental. “For the most part, what we’re looking for is a short distance between two points, where you’re just replacing a wire for a moving part. The controller is right in the vicinity of the end device, so we don’t need to mesh that across hundreds of feet to the other side of the plant,” Read explains.

Unlike the process world, in which sensor and instrumentation data is often brought back to a central point, discrete manufacturing tends to have more distributed architectures, Read continues. “We’re not running as one big process. We’re running as multiple, small discrete processes.” In a robot work cell, for instance, multiple robots typically all have their own controllers that are having their own conversations with their own end effectors, or the programmable logic controllers that are managing them, Read notes.

This means that any factory wireless standard will need to accommodate these multiple transmissions within a relatively small, physical area while avoiding interference. Standards such as 802.11g offer ways to do this through use of nonoverlapping channels, says Read, while the new 802.11n version of the standard, published last October, will offer even more opportunities for channel separation.

Not waiting

Despite a variety of RF devices currently on the market that may potentially meet factory automation needs, they are not currently interoperable, so development of a standard is vital as a way to ensure multi-source availability, Read emphasizes. But even without a standard, Read says he has heard of some end-users who are already deploying wireless control for robot end-effectors. And if a standard can be put in place, he expects that discrete manufacturers would move quickly into broader use of wireless for factory control.

“Once we’re able to demonstrate that these are reliable devices, and that they’re deterministically providing updates, I don’t see a reason not to fully jump into it as an input and output media,” Read concludes.

Subscribe to Automation World's RSS Feeds for Feature Articles

More in Control