Peter WelanderPillar to Post: Peter Welander's Blog


News and comment from Control Engineering process industries editor, Peter Welander

Backward compatibility vs. future proofing

March 31, 2010


We posted a news story earlier today containing some suggestions from Emerson Process Management on the topic of how WirelessHART and ISA100 can converge. While it may be an oversimplification, the gist of the matter seems to be that Emerson is suggesting that WiHART should become the working device-level protocol for ISA100. Adopt WiHART and be done with it.

This is a valid suggestion, and to many people, I’m sure it seems like a very logical approach. WiHART works, the number of devices available is growing, and there is a strong element of interoperability within the platform. Case closed.

Others aren’t so sure. You may have read Patrick Schweitzer’s comments on why he’s holding out for ISA100.11a. But there are at least two other points that need to be considered.

The first is HART itself. While there is every reason to believe that WiHART works exactly as advertised, just as it does in the wired world, the question you have to ask is if HART is the protocol you want to use for all your field device level communication. To many companies, using HART would be a major advance over plain analog, and it offers many capabilities beyond what they’re using now.

Still there are some who want a more advanced approach for one reason or another. For example, recently I attended the Fieldbus Foundation General Assembly and heard some interesting discussions of how that device networking method is being applied. While there are many reasons to use a fieldbus topology, arguably the most compelling one in this day and age is its ability to support sophisticated asset management. That was a major component of the discussion from Reliance Industries.

Herman Storey led a session specifically on asset management and explained how critical FF can be to an effective AMS. He made the point that you can use HART to support a program, but it is far more difficult to work with. Using a fieldbus network makes device diagnostic information much easier to get and it is better at helping preserve device configuration settings. If you’re using this kind of capability with your wired devices, you wouldn’t want to give it up for wireless.

The second point is more speculative but seems just as compelling. Over the years, many manufacturers with products that evolve try to maintain “backward compatibility.” You know, that’s when the new version of your spreadsheet software can still open files done in the earlier release. HART, like many in the same general space, has worked to make sure that devices built to its earlier protocols will still work with the new ones. Old devices may not gain the functionality improvements of the new platform, but they still operate as well as they always did. Nothing is forced into obsolescence.

It seems like there are opportunities for wireless devices to move from backward compatibility into a new realm of future proofing. I’m not sure of WiHART’s capabilities in this respect, but proponents of 11a say that most of what makes the devices work is embedded in software. This means that if there is an improvement in the protocol, the software can be replaced and the device moves forward rather than being stuck on the older release. Honeywell says it has done this, and as long as users are content with the 802.15.4 radio, the individual unit can be upgraded without hardware changes. How far this can go ultimately remains to be seen, but you can see the potential.

Let’s be realistic. As well thought out as these protocols are, they are going to pass through some iterations going forward. That’s a fact of life. The more you can build upgradability into the products, the better.