SP50 times two?

"THE LAST thing that the end users want is another SP50,” says Marty Zielinski of Emerson Process Management, who is co-convenor of ISA’s SP100 Wireless Standard committee. I completely agree. Unfortunately, all of the signs point to the possibility that we’re going to have the same problems with SP100 that we did with the SP50 Fieldbus Standard committee.

Now that the deadline has passed for submitting statements of intent in response to SP100’s Call for Proposals, we find there are more than 20 companies and organizations from all over the world, which intend to propose that their solution be the chosen one for the standard.

One of the things I found most interesting about the Call for Proposals document itself is that it spends almost all its time talking about the “physical layer” of the specification, and virtually no time talking about what end users care about: the application layer. This is dangerous because the committee could get sidetracked for several years worrying about the physical layer.

I don’t believe that the physical layer is as important as the application layer. But, even if I’m wrong, there is a way around the “my chip is better than your chip” posturing.

Jose Gutierrez, formerly of Eaton and now Emerson’s technology planning director, was instrumental in the IEEE 802.15.4 standard and has made a suggestion I find has terrific merit. I’m passing it on because even though Jose has only been at Emerson a very short while, he’s a party to the SP100 situation. While I am a member of SP100, I don’t have a corporate axe to grind, sharpen, or stick in anybody. In fact, I tend to be an equal opportunity offender. Sometimes, people stop sticking knives in each other, and everybody gets mad at me.

What we ought to do is to set up some testing for the proposed silicon in the same way that IEEE 802.15.4 tested its silicon. Perhaps Wayne Manges, co-chair of SP100, could use the testing labs of his own Oak Ridge National Lab, and maybe we could get one of the other third-party labs, like UL or FM, or one of the other national labs to be the other test lab. Then, allow each test lab to run performance testing on each proposed chip or methodology, so the committee could actually have data to analyze, instead of marketing posturing. This way, the best silicon wins. That takes care of the physical layer.

Now comes the important part.

One of the things that SP100 lacks, and it’s certainly not from lack of trying to acquire it, is an acceptable number of user cases and application documentation. As this is written, the committee is trying to get more end users involved in the committee’s work, but I doubt that we’ll get many more users than we already have, and there has been a run of vendors trying to join the “end-user” committee. Many companies don’t want their end users to be that involved in standards activities to the detriment of the time they can spend doing their own jobs, and many vendors believe they know what the end users already want—what the vendor has to sell them. This can be just as dangerous as concentrating too much on the physical layer.

I’ve offered Control’s help in collecting more data. We need to do more than pay lip service to the idea of user input.