Unplugged: Developing Standards for Wireless Automation

The great disadvantages of new technology are that its commercialization involves a high degree of risk and its future holds an even higher level of unpredictability. Wireless industrial automation has been slowly embraced by some petrochemical and chemical processing companies. Meanwhile, the highly conservative pharmaceutical industry has so far, and not surprisingly, taken a wait-and-see position—and for good reason.

To build the most robust and secure wireless devices, suppliers build according to an approved standard. Wireless Internet communications, for example, all comply to the familiar IEEE 802.11 (b, g, and n) standards. Currently, both the HART Communication Foundation (HCF, Austin, TX) and ISA (Research Triangle Park, NC) are involved in putting forth an industrial wireless standard at the basic "bits and bytes" level (i.e., sensors) (see sidebar, "Which wireless?"). HCF released its standard, WirelessHART (HART generation 7), in September 2007. To date, ISA is continuing to develop a draft of its standard, officially noted as ISA 100.11a (often referred to as SP 100).

Which wireless?

Both organizations aim for global acceptance, including by the pharmaceutical industry, which thanks to the move toward quality by design, is preparing for increased in-process monitoring. Both HCF and ISA agree that a wireless standard is the right step to advance process automation and communication. Both are aware they are competing to achieve the same objective. And of course both would argue that their standard is the most beneficial, robust, and secure.

Brief history

Dual standards are not new to process automation. For example, the original wired HART digital standard was developed more than 15 years ago to be carried on top of the existing 4-20 mA analog signal. At that time, HART was an emerging communications protocol. "If you look back 10 years, HART wasn't supposed to be where it is now. I recently looked at the articles that were written and they said HART was an interim technology and by 2005 it was supposed to be dead or dying from a technology standpoint," says Ed T. Ladd, Jr., director of technology programs at HCF. HART is now worldwide and is the most used protocol for field devices in the process industries. HCF built upon that technology when developing its WirelessHART standard, which was reviewed and approved by its membership and ratified by its board for release on Sept. 7, 2007.

Emerson Process Management is currently the only company that has announced the availability of WirelessHART devices (see Figure 1). Although none have been installed at this time, prestandard wireless devices continue to be in use. "To date, we know that member companies are actively developing WirelessHART products to be offered for sale by the end of the calendar year," says Ladd. "Major vendors, including ABB, Emerson, Endress +Hauser, and Siemens, have all committed to support the technology."

At the time WirelessHART was released, ISA had been working on ISA 100.11a. "ISA began to work on a wireless standard in 2004. I'm not sure what the motivation was on the part of the people who decided they would not wait for ISA 100 but would build on available technology and develop WirelessHART," says Richard Caro, certified automation professional, CEO of CMC Associates, and chair of a user-working group for ISA. "ISA 100.11a is beginning to develop rather quickly and at this point is almost complete. The standard is scheduled for voting and approval by the end of 2008."

A matter of timing

As a member company of HCF, Emerson attributes the speedy approval of WirelessHART to the organization's size and approval process. HCF is a small organization comprising the major vendors and users of wireless devices. "It's consortium of companies and customers that have a vested interest in making something happen, so they tend to be more efficient at making things, and it shows up in a standard that goes to market much quicker," says Bob Karschnia, vice-president of Wireless, Emerson Process Management.

ISA is a worldwide standards-making body that operates under American National Standards Institute (ANSI) rules. Following ANSI rules means that the document must undergo several rounds of comments, resolution, and voting to become a US national standard and a cycle for becoming an international standard and approval through IEC. So far, the 100.11a prestandards document has undergone a round of comments and resolution, and, according to Caro, the final standards document is being drafted. "Right now, we are on schedule and the next seven months or so will be devoted to voting and comment resolution."

Testing

Designing a standard starts with conducting a number of "use cases." Use cases help determine whether prototypes can solve specific real-world industrial applications such as monitoring temperature in a tank or controlling the flow of steam into heating and cooling jackets. Both ISA and HCF have conducted use cases. According to Karschnia, HCF went through all of the use cases that had been developed for 100.11a. "We found that every one of them is met by WirelessHART."

Use cases differ from validated confirmation tests, which prove that an application meets an accepted standard. This is one difference between WirelessHART and ISA 100.11a. "As a matter of fact, WirelessHART is still preparing its confirmation test sweep," says Caro. "With ISA 100, the production of a validation sweep is going along in parallel with the standard. When the standard is issued, it will be just a few weeks afterward that the validation sweep is complete."

Ladd questions this approach. "You have to write the standard, then build to the standard, then test to the standard. How do you build a device to a standard that doesn't exist? To me, that doesn't make sense. Once the standard is complete, then you can begin to develop your product to validate the standard, you make modifications to the standard if you find discrepancies, that is part of continuous process improvement."

According to Caro, the intent for ISA 100 is that before the standard is published, a number of companies will have already used the draft standards document to build product, and therefore they will get the actual use of the standards document and will have a chance to make corrections to it during the comment period. "The HART organization has the freedom of being independent to choose to issue their standard without that kind of comment-resolution period, which they have done. That shortcuts the timing requirement, but it doesn't necessarily make the standard available for production devices any sooner because they will have to go through the next revision of their document. It is now HART 7.0. Before that standard is completed it will certainly be at the 7.1 phase, reflecting the comments of the companies who have tried to build products to that standard."

Ladd explains that HCF provides 60 days for review and comment and a 30–60 day resolution period before the board of directors vote on ratification of any member-approved changes.

Weighing the differences

The most important question is whether there are any differences between ISA 100.11a and WirelessHART. The answer depends on who is asked.

According to Caro, "The main differences are very technical and not easy for users to identify. Some users have an immediate need for wireless protocol, so that is one benefit for devices conforming to WirelessHART, which will be on the market some number of months in advance of products that conform to ISA 100.11a. Users needing to augment safety systems, for example, are installing whatever wireless they can put their hands on for many of these applications simply because they want the functionality and they are not willing to wait for the standard with the idea that perhaps someday they may go back through those installations and replace them with something that conforms to their plant standard. Many users, particularly those on the ISA committee, say that to them the early availability of wireless devices is not an advantage. Being first to market doesn't necessarily mean that it's the one that is going to prevail." In addition, says Caro, "the time and effort being devoted to ISA 100 has benefits in the information-security area that exceed the requirements for WirelessHART."

Karschnia disagrees. "There are no differences that are important to solving the customers problems, and this is key. It is like one person saying something in English and another saying it in French. They use the same underlying basics, except that SP100 has deviated from an IEEE standard [802.5.14], and WirelessHART used the IEEE standard as is."

Ladd says the radios are the same and the frequencies are the same, but there is a difference in how they communicate and how passwords are established. "WirelessHART only allows join keys or passwords to be keyed in via hard wire, whereas [ISA] intends to transmit them via wireless. From our standpoint, this could be a security risk. Our standard requires a maintenance port that is secure and backward compatible to all handheld communicators and PC-based applications that are in the field today."

Emerson donated a lot of intellectual property to HCF and ISA. "It's the same intellectual property for security, time-synchronized mesh protocols, and 802.15.4 radios. Both of those standards have adopted those pieces. If you peel back all the noise in the system, you'll find that they look exactly the same," says Karschnia.

Double standards

ISA 100.11a and WirelessHART are built using the same chip (i.e., the physical layer of the two standards is identical). They operate at the same frequencies and use the frequencies in approximately the same way. However because of differences in the upper-layer protocols, they are not compatible. But compatibility is not the same as the ability to coexist. "The biggest questions that users ask are, 'We know they are not compatible, but can they coexist? Can I have a network of WirelessHART instruments that I've installed this year and then I go back later on in the same area of the plant and install another network of ISA 100.11a instruments?" says Caro.

Because the two standards use the same frequency, Caro says they have the potential to interfere with each other. "However, because we have recognized that the 2.4-GHz spectrum is busy, we have to provide ways for devices to share that spectrum and to recover from any collisions that may occur. And we have built in the recover mechanisms so that the user generally will not be required to intervene at all; that is, they will self recover, and the two networks will coexist peacefully within that same domain. We believe that to be true and we have additional adjustments that can be made to ISA 100, although it is not clear that any of those things can be done to WirelessHART."

Caro explains that technically, both protocols use a pattern of frequency hopping, and the patterns of frequency hopping are different. But occasionally they will hop to the same frequency at the same time. The result of that will be collision, an error message, and then a retry. The retry will occur at a different frequency and because the hopping patterns are different, then the retry in both cases will be at different frequencies. So there is an automatic recovery mechanism that will enable the two systems to coexist without the need for elegant software tools to adjust the networks.

"There is just a difference in philosophy of how the systems work," says Ladd. "We're using mesh networking, where all of the devices are routers. We're using 802.15.4 radios, and that is the same radios ZigBee [ZigBee Alliance, San Ramon, CA] uses. We added the frequency-hopping technology, additional security measures, and a number of things that will make the WirelessHART network much more robust and backward compatible to 26 million devices already installed."

Foundation Fieldbus and Profibus are two organizations that build their communications protocols on top of either ISA 100.11a or WirelessHART. "To the extent they have simple networks, then WirelessHART could serve as a base for Foundation Fieldbus. But to the extent that their customers want to do more like control any of the field devices, then they will find that the additional flexibility of ISA 100.11a will allow their networks to work better for control reasons," predicts Caro. "Profibus appears to be less interested in the additional security and the other requirements that are given by ISA100 and appears to be leaning toward founding the wireless version of Profibus on top of the WirelessHART protocol. That of course could change as the users of Profibus change. For example, Profibus is largely influenced by the German chemical industry, which is the largest industrial category in Germany and includes companies such as Bayer. To the extent that Bayer recognizes the additional reliability and security added by ISA, it could change the product specifications that are currently occurring within the Profibus organization. But since our standard is not yet complete, there is no basis for them to do that at this point."

Collaborations

On Sept. 27, 2007, HCF announced a collaboration with Foundation Fieldbus and Profibus to work on wireless technologies for process industries. The release mentions that "the organizations have agreed to base their work on WirelessHART technology ... and the emerging ISA SP 100.11a standard."

On the same day, ISA announced that HCF and ISA had collaborated to "develop a single wireless standard for process measurement and control applications." According to this statement, each organization had granted copyright license to the other such that HCF would have access to ISA100 documents at all stages of development and ISA would have access to HART protocol specification, including WirelessHART. ISA announced the formation of a joint technical committee "to assess WirelessHART technology for meeting ISA objectives and recommend possibilities for incorporation into the ISA100 family of standards." Whether the incorporation of WirelessHART into ISA standards will or can actually take place remains to be seen.

Says Caro, "For a long time, I have been a very strong advocate of trying to get the WirelessHART group to abandon what they have done and simply use those parts of ISA 100 that appear most like WirelessHART. That could be done without real technical compromise, but they seem to be unwilling to do that. We have an investigation to look at WirelessHART and make it a proper subset of ISA100, but that doesn't seem to be happening."

According to Ladd, "WirelessHART is not going away. WirelessHART incorporation into ISA 100 is up to ISA at this point. We have talked about it. There was a ballot posted on the SP 100 website to begin a WirelessHART subcommittee, so there are things definitely moving forward to incorporate WirelessHART into SP100."

"[In April], there was a motion raised within SP 100 to make WirelessHART as it exists, with no changes, an SP 100 standard, and I think that will be approved," says Karschnia. "If it doesn't, you end up having competing standards that are in the industry and that will be an issue because it will just confuse customers and less people will buy in general."

So is it political? "I don't know," says Caro, "It's more economic. And it's not economic in serving the end user, it is economic in serving the customers of the HART Communications Foundation who are the vendors."

Profibus, Profibus International, Foundation Fieldbus, and HART have formed a working group to develop a specification that will give common interface using a wireless gateway. "We want to give users some assurance that whatever they chose [Foundation Fieldbus, Profibus, or Profinet], they are going to be able to use it going forward," says Carl Henning, deputy director at PTO, Profibus, and Profinet North America. According to Henning, the group will look at both WirelessHART and the ISA standard, although because WirelessHART is complete, Profibus has been working mainly with that standard. "I've been told that we will probably see the Profibus gateway by the end of the year. This should help users because it will result in a common interface so they won't be stuck deciding which wireless sensor network to choose."

Says Karschnia, "It seems it would be worthwhile for SP 100 to change its focus from the bit-and-byte communication sensors to bigger problem of gateways." (In a network, a sensor transmits data, the data hops through a time-synchronized mesh protocol network, and arrives at a gateway.) "ISA should focus on how does that gateway will interface to the rest of the process control systems that exists—to the DCS, to the historians, to the business systems, and to the diagnostics systems. Once someone defines that, then the industry will adopt wireless more readily. To date, Emerson has its solution, Honeywell have its solution, and everyone will have a different solution. That is where ISA can add value."

Henning offers this perspective: "The thing to remember is that it doesn't matter who puts out the standard, it's what the users adopt that is going to be important. From that standpoint, I like the idea of having competing standards. The users are winners by having competitions because the standards organizations are spurred on to do better by that competition. There is a limit, but I think with two or three competing technologies, it's a positive."