By Dick Caro, CMC Associates, Acton, MA
There are plenty of wireless instruments for process automation available today, but they are all proprietary, do not conform to any standard, and generally are not interoperable. Each network constructed from these instruments generally uses a standards-based interface to the host system using OPC, Modbus, or similar protocols. Users would generally like to see a more integrated approach for host systems, but vendors tend to use such open protocols for connection capability to almost any host.
ISA100.11a is the name of the first standard being developed by the ISA SP-100 standards committee. The committee was officially chartered in 2005, with an editing team created in early 2007 to actually write the standard. Completion of the standard’s first draft is scheduled for November 2008, and it may be that this schedule will be met.
ISA100.11a is being targeted to serve the data acquisition and limited control needs of the process industries for this first release. Products designed to this specification are expected on the market shortly before the end of 2008 at the earliest, with compliance certified products available shortly after release of the standard in 2009.
Like all the HART specifications before it, WirelessHART is targeted to serve the data acquisition and control needs of the process industry. HART 7.0 planning began in early 2005, the specification containing WirelessHART was approved by HCF members in June 2007 and released by the HCF Board of Directors in September 2007. Among other capabilities, WirelessHART provides a new wireless communications path to capture the digital data contained in more than 26 million currently installed HART instruments without a direct digital connection to their control system. Products designed to the official HART 7.0 specification have been informally announced for sale and shipment, and compliance certified products are likely to be on the market by Third Quarter 2008, or sooner. At least one company has announced an OEM kit for other vendors to assist them in producing their own WirelessHART products.
Most observers question why there are two competing wireless standards for the same small industry. Many users involved in the development of ISA100.11a are questioning why these two very similar standards that are both based on the same radio (IEEE 802.15.4:2006) operating at the same frequencies cannot be united into one standard. These are excellent questions and deserve an answer that we will try to provide here.
The two standards are the result of the free-market. In November 2004, HART Communication Foundation (HCF) members explored the synergy of extending the HART protocol to enable wireless process measurement and control devices based on emerging wireless technologies (IEEE 802.15.4 and sensor mesh networking). HCF members endorsed the establishment of a Working Group to investigate and develop the appropriate extensions to the HART Protocol standards for wireless communication. The new HART 7.0 standard including WirelessHART resulted from this effort and is a natural evolutionary enhancement to the HART Communication technology. The HART 7.0 specification was produced by a small, focused, and dedicated group that was able to move much faster than the ISA 100 committee considering the vastly stated broader scope and purpose of ISA100. The HART 7.0 standard has been reviewed and approved by the 200+ members of the HART Communication Foundation, who are the largest group of vendors who manufacture field devices so it is well supported by process automation suppliers.
Interestingly, enough, many of the companies that supplied people to work on the HART 7.0 standard are the same companies that are also working on the ISA100.11a standard. The two standards are targeted at the same process applications, but are basically incompatible. WirelessHART is an extension to the proven, well established and well supported HART Communication Protocol standard. ISA100.11a is working to define an entirely new communication protocol targeted at the same process measurement and control applications. The HART Communication Foundation is primarily a vendor organization, and most of the people writing the ISA100.11a standard also work for the same vendors that are members of that Foundation. This is confusing for end users and causes them to wonder why these same vendors cannot cooperate with themselves.
Many end users and some suppliers have questioned the haste in producing a draft ISA100.11a standard that has not integrated the WirelessHART protocol. They also have questioned the value of the ISA100.11a standard when the WirelessHART standard is already available. As with the previous questions, these too are valid alternatives and should be examined. First, the haste in producing the ISA standard comes from the ISA’s sense of competition from WirelessHART itself and the general desire to complete a comprehensive work, once begun, as soon as possible.
ISA100 and HART Joint Analysis Team
The question remains: why can’t the WirelessHART specifications just be embedded into ISA100.11a? There are a number of reasons, mostly technical, but the main reason is that the HCF did not offer the wireless technology embedded in the HART 7.0 standard to be used in the ISA100.11a standard until the HART 7.0 standard was completed and released for publication by the HCF Board of Directors. Unfortunately, this was well after the writing of ISA100.11a had begun. Immediately following publication, the HART Communication Foundation provided ISA with the complete HART 7.0 standard to evaluate and consider for incorporation into the ISA100 standards including the offer of user rights should ISA want to adopt the WirelessHART portion of the specification.
The ISA100 technical committee has not yet taken HCF up on its offer of granting use rights to ISA for the wireless portions of the HART 7.0 specification, because ISA100 leadership did not adequately understand the need for a unified standard. Instead, a Joint Analysis Team (JAT) was formed with members from both the ISA100 standards committee, and members selected by the Board of Directors of the HCF to study the issues.
The JAT considered several alternatives and came to the conclusion that full integration of the two specifications would not be considered until a later release of the ISA100.11 standard, because no simple solution was offered, design of the integration could become very complicated, and would delay completion of the first release of ISA100.11a with no technical benefit. (Benefit of a unified standard to the end user was apparently not considered to be very important.) However, this committee did succeed in defining a host network interface to a dual-mode gateway that would be the same for both ISA100.11a and WirelessHART field networks, which is being written into the ISA100.11a draft standard.
This way, the DCS or other host system may be able to access information from a field device on either a WirelessHART or ISA100.11a network in basically the same way, from the dual-mode gateway, should the final ISA100.11a standard incorporate this feature. While the wireless field networks are separate, distinct, and not interoperable, they may coexist and might be joined at this common dual mode gateway.
Design of this common host level interface was made possible by incorporating the design for this interface already present in the HART 7.0 specification. So, while at the same time agreeing not to use WirelessHART, the ISA100.11a editorial team created the specification for a host level interface to the gateway in the ISA100.11a Release 1 draft standard that simply used the same core interface already in the HART 7.0 specification as the basis technology.
This host level interface should be very easy for the Wireless Cooperation Team (Fieldbus Foundation, HART Communication Foundation and Profibus Nutzerorganisation) to build upon as they define a common wireless gateway interface for WirelessHART onto Foundation Fieldbus, Profibus and Profinet. In so doing, they will have also created a common interface to ISA100.11a as well. These consortiums had previously agreed to specify a common interface to a WirelessHART gateway. The dual-mode gateway interface proposed by the JAT and embedded into the ISA100.11a Release 1 specification seems to accomplish most of this task with the added benefit that it allows that same interface to be used with ISA100.11a Release 1 networks.
Integration of the Standards
Now, let’s turn to the rest of the problem: integration of WirelessHART and ISA100.11a. There are several different ways in which these two standards may be integrated:
- ISA100 may adopt the entire wireless protocol specification of HART 7.0 replacing the current draft standards work of ISA100.11a,
- HCF may adopt the entire wireless protocol specification that is ISA100.11a replacing the current wireless protocol contained in the HART 7.0 specification, or
- Merge the WirelessHART protocol into ISA100.11a by inserting a protocol splitter just above the common MAC (Media Access Control) layer creating a dual-protocol stack. Then messages would be handled by the proper protocol stack for routing and servicing according to the network identity configured into the field device.
- Merge the two specifications by creation of a truly unified single specification with options for the complex features of ISA100.11a, but retaining the simpler features of WirelessHART as the set of defaults.
Option 1 questions the necessity for a totally new ISA100.11a protocol when the WirelessHART protocol is available. If all of the process control needs of industry could be fulfilled through the use of the WirelessHART protocol alone, then this suggestion would have merit. However, it is the opinion of many technical experts working on ISA100.11a and some end users directly involved with its development, that this is not the case.
- If the needs for wireless field communications were defined by the requirements to support only HART and Profibus-PA applications, then WirelessHART might be adequate. However, the HART 7.0 specification currently does not provide for direct peer-to-peer data paths that would be necessary for Foundation Fieldbus “control in field device” functionality. The draft ISA100.11a standard was designed to provide direct peer-to-peer data paths necessary to close control loops using only field instruments similar to Foundation Fieldbus H1 wired networks. Some observers question the value of control in wireless field devices, however.
The HART Communication Foundation expects that the next revision of the WirelessHART standard will also support Foundation Fieldbus, since there is a joint working group in which HART Communication Foundation and members of the Fieldbus Foundation are working to develop that capability. And like every other revision of the HART standard since version 1.0, it will be fully backward compatible with every other version of HART ever issued.
The other issue is the speed of a WirelessHART network when used to supply measurement data for control loops. In wired HART, the 4-20mA analog signal is usually used to supply process variable (PV) data for control purposes. Modifications to the current WirelessHART specification are currently being made to reduce the latencies to approximately those of ISA100.11a as indicated in the current draft specification. WirelessHART then will have a maximum latency of 100 ms, which appears to be quite fast enough for process control. However, at least at first, most wireless applications are expected to be asset management and process monitoring related and much slower update rates will be acceptable.
To minimize the latency of a mesh network, ISA100.11a is being written to allow configuration of a field backbone into the network. WirelessHART does not specify such a backbone network. The HART Communication Foundation states that it designed WirelessHART without a field backbone to permit it to be used with any commercially available backbone such as Foundation Fieldbus HSE, Profibus, Profinet, WiFi, etc, in much the same way that ISA100.11a, which is itself designed without a backbone, can be used with commercial backbones.
Option 3 certainly is one way to gain a single standard, which end users clearly want, based on all the research that has to date been done. However, since there is no work on this for the current draft of the ISA100.11a standard, it is unlikely that this will happen any time soon. It also seems clear that the earlier availability of WirelessHART devices may make this option unnecessary.
The purpose of Option 3 in integrating these protocols is to allow suppliers to build field devices that will be capable of operating in either a network configured to meet only the “original WirelessHART specifications,” or in a more general ISA100.11a network. Therefore, devices requiring the protocol splitter will only be required for installation in plant areas in which there are devices that are NOT capable of using ISA100.11a protocol (do not contain the dual-protocol stack.) This means that the merged protocol is only necessary only to support installations of new devices in wireless network segments that were originally constructed to use only the WirelessHART protocol.
Clearly, Option 4 is the one for which most end users wish. There are two problems with this option: it will take time to produce – more time than any other option; and it does not seem to solve the market dilemma in which there will be two incompatible product families being sold until the vendors support the unified standard, if they ever do. The only problem option 4 resolves is the end user dilemma.
Will there be a significant number of those purely WirelessHART networks to require merged protocols within the ISA100.11a standard? Almost certainly. Emerson has already sold significant numbers of SmartWireless devices that are extremely similar to WirelessHART, and that the company has pledged to convert to WirelessHART products. The WirelessHART version of the device that attaches to the 4-20ma wiring for an existing HART transmitter is called a “thumb” by Emerson/Rosemount who has already announced that it is currently taking orders with a shipping date in early May.
No supplier has yet made an announcement for an ISA100.11a equivalent device, but such products are likely to appear before the end of 2008 or in early 2009. This apparent time gap is driving all but one of the major automation vendors to build product to support WirelessHART without a corresponding commitment to support ISA100.11a.
If the promised launch of WirelessHART devices occurs in mid-2008 and those products are actually shipped later in 2008 in any quantity, then there may be hundreds to thousands of these devices installed. However, companies supporting ISA100.11a may not be ready to relinquish the market to WirelessHART. If, however unlikely this may seem based on the schedule issues necessary to develop a real issued standard, vendors release “pre-ISA100.11a” devices with a pledge to convert to compliant ISA100.11a in the future in mid-2008 and actually ship them later in 2008, there could be a reduced demand for pure WirelessHART devices.
We know that ISA100.11a can be configured to accomplish many of the WirelessHART functions, using a very similar application interface based on EDDL. Furthermore, the host network interface for both protocols will be the same. It also appears that there is little to no cost difference to support either protocol, but the cost of supporting two protocols is quite large for both vendors and users. Furthermore, vendors may at any time choose to support both WirelessHART and ISA100.11a protocols by instantiating the required protocol stack in a number of vendor-specific ways, such as downloading or activation if previously installed. Given that there will be little functional or selling price differential is it worth the effort to integrate these standards? Is it worth it for vendors to support both standards?
Will the effort to create a dual-protocol stack standard actually occur? Will it really provide any value and will it be worth the cost? It is certain that without strong end user pressure, the suppliers will not make any changes to WirelessHART with the result that some products will be installed in the field without the capability to ever be converted to ISA100.11a. Some users will buy WirelessHART because it will be available earlier than ISA100.11a instruments. Others will choose to wait for either devices that are ISA100.11a compliant, or are promised to be convertible to ISA100.11a. However, it is more likely that users will purchase neither in any quantity until the suppliers pick one standard or supply devices that can be configured or downloaded for either protocol. This confusion in the marketplace is just like HD DVD vs. Blue-Ray – users may decide to buy neither. The market always makes the “best” decision, not always the best technical decision.
Which “camp” will win this battle of the wireless protocols, sometimes called an “air-war?” The vendors seem to be aligned into these two camps, each with their own strategy. WirelessHART seems to have the support of device suppliers in the process automation space. The support of device suppliers for ISA100.11a though is not yet clear. Early adopter users must make a “bet” on their perception of the winner. Either way, if this division continues, the users will lose.