EDDL: The Technology Behind Interoperability
EDDL now is being enhanced to extend the concept of interoperability to the HMI and diagnostic data.
The FDT/DTM vs. EDDL battle is one of those Layer 2/3 battles mentioned earlier, in which it is once again North America and fieldbus led by Emerson Process Management vs. fortress Europe with ABB and Endress+Hauser at the helm. However, this battle seems more a tempest in a teapot because both groups are working to release an OPC-UA compatible version of their standard in two to three years and the entire system will be using the same technology as its backbone, thus making integration and any necessary work-arounds much easier.
The greatest risk of having another skirmish on this level of the enterprise is that if it is not soon resolved, and a way to store data on a host in a retrievable manner regardless of supplier is not found, a web-based solution using XML and HTML with perhaps a small web server will sneak in the back door and make the entire discussion moot.
EDDL and FDT/DTM are the tools with which to present the data required so as to make best use of the information in a modern control system. What is also required is a standard or series of standards to define the information itself and how to move it from one system to another. ISAs S-95 is one group leading this effort as are the NAMUR and WIB groups in Europe. Fortunately, efforts are also underway to have these groups work together with member-led ISA as the global instrumentation, systems and automation society taking the role of catalyst in these activities.
This leads to another current battleground in Fieldbus War II: the Industrial Ethernet juggernaut.
Industrial EthernetPanacea or Problem in Waiting?
As all CONTROL magazines readers likely know, Ethernet describes the Physical Layer of the OSI seven-layer model and, since most industrial protocols also use Internet Protocol (IP) and User Datagram Protocol (UDP), the Data Link Layer is also fairly common. The biggest differentiator is the User Layer, which of course is not part of the OSI model.
The question here, especially in the non-wet industries, is not if Ethernet will arrive, but when and how low can it go? Brian Oulton of Rockwell Automation indicates that traditional RS-232 communications are all now migrating to some form of Ethernet, especially since a number of chip vendors are offering dual stack silicon with, for example, Modbus/TCP and Ethernet on the same chip. Oulton also points out that manufacturers are still using proprietary protocols on Ethernet, which is not the best way to go since there is broader acceptance of solutions if an open solution is selected and the emphasis is placed on differentiating features.
This is consistent with what end users are saying as well. They would rather be able to pick the features they need from the manufacturer they need instead of spending their time mapping protocols and building gateways.
With the merger of companies both on the supplier and end user side continuing, it is becoming necessary for engineers to support multiple control environments with different hosts. It would be much simpler and efficient if it was possible to integrate the existing equipment with a new control solution by means of a common network, likely some form of Ethernet, through a single gateway supplied by the host manufacturer. This is done today when a manufacturer upgrades its control system to a new platform. If standards for integration and upgrade of multiple manufacturer devices existed, the same could be done across manufacturers as well. A definite win for all involved since the host manufacturer will still get the revenue to support the legacy system and the end user will be able to continue using his or her investment in hardware and software as long as possible.
History has shown that Emerson Process Managements release of DD technology into the public domain has been a boon for the development and integration at Layer 0 and 1 of the enterprise, along with the associated benefits identified above; the same could be expected at each of the other layers as well. Sure, users could mix and match components from many suppliers in a single system. It is unlikely they will, because they will then have to carry parts for multiple manufacturers, learn about each manufacturers equipment, and manage all the interfaces themselves when they could simply expect a single source to provide a complete solution instead. What may happen is that users will continue picking a single supplier to provide a function, such as safety system, basic controller, maintenance solution, etc., and know that they will integrate into a complete solution.
Fieldbus SafetySafety in a Spec
The final topic that is garnering the attention of the fieldbus specification developers is that of applying safety in fieldbus systems. As mentioned above, ISA is doing its part to make this possible while both Profibus and Fieldbus Foundation are spending considerable effort on their standards as well. Unfortunately, that is all they are doing at present and it seems that it could be at least another year before we see product available.
It is also unfortunate that it does not appear that any of the standards under development will address all the requirements being identified by the ISA-84 subcommittee, at least on the first release. What they are doing however, is using what is called a black channel to take advantage of the existing Profibus PA or Fieldbus H1 standards to handle the communications portion of the network and then adding the safety functionality and checks on top of this infrastructure.
An end user in Europe is looking forward to the availability of fieldbus safety because, Not only do we feel it will help increase the reliability of the existing standards, it will also allow for the use of diagnostic capabilities on the application level. As result the loop diagnostic coverage factor will increase and manual testing to prove functionality of the safety system will be reduced. On the control level the diagnostics will reduce the demand rate on a safety loop so, compared to todays situation, scenarios will be classified to a lower SIL level.
With respect to the infrastructure there are two philosophies: (1) Control and safety functions should not be integrated in the chemical industry, and (2) Control and safety functions are integrated on a segment.
Its likely that in the next few years, fieldbus safety will be implemented and if the technology is proven and end users are familiar with the technology, the functionality of mixing control and safety functions on the segment will be implemented. In this situation the use of a separate PLC system will be eliminated.
The Final ChallengeThe Enemy Within
The challenge with all networking technologies that will not be solved by the protocols, including safety buses, is that the equipment must first be installed correctly. From my personal experience, which has been confirmed by many others, installation problems are the cause of 90% of network problems. Getting the physical connections and terminations done properly is critical to the successful implementation of a project.
Bill Gates recently indicated that people are often too optimistic when forecasting two years into the future, and too pessimistic when looking 10 years ahead, so only time will tell if the fieldbus wars will be over at the end of the decade or simply moved on to another layer of the enterprise. It is my hope as well as that of many others that the wars will be over long before then.
Ian Verhappen is Director at ICE-Pros, Inc. an independent instrument and control engineering consulting firm specializing in fieldbus, oil sands automation, and process analyzer sample systems. Ian can be reached at Ian.Verhappen@ICE-Pros.com or through his web site www.ICE-Pros.com.