The Great BACnet debate

I've watched the postings on the BACnet list server for the past two years and have not seen much debate - until recently. It seems that the ugly truth for all protocols is now reaching the discussion stage. The question is: how are proprietary features in different vendor's BACnet products to be handled with regard to interoperability? Why are there proprietary features at all? How can a voluntary protocol management process succeed in getting agreement among vendors?

Well it's tough. Just look at the internal debate now going on. It's always been fasinating to watch various vendors jockey for position in the HVAC and PLC product world. Back in the late '80s General Motors decreed that all of their plants will go with MAP (Manufacturing Automation Protocol) and their suppliers had better do the same - because this is a "new" world and GM is leading the charge. Well MAP flopped. It was expensive. It was clumsey. It was built on a broadband backbone that was very inappropriate. OK, back in the '80s there really wasn't another approach. Ethernet also was clumsy. Big coax cable. Small coax cable. And there was very little 10BaseT around. Prices were relatively high for whatever option you chose.

But MAP was never to happen. It was an imposed solution. Clearly there was a problem in the GM, Ford and Chrysler plants. But the problem was not of the same magnitude in their supplier plants (except for what is now Delphi and Visteon). But that is not the case for BACnet.

BACnet has a chance to lay out a communication scheme that provides true interoperability. I doubt, however, that it will ever happen. Too many cooks in the kitchen.

The most recent proposal by David Fisher to provide a type of XML file to describe customer properties is on target. Maybe David can lead the charge and make it all happen. He certainly is trying.

The problem is that the building automation industry is chasing a technical problem (which, granted, may be the only type of problem they can chase) while customers don't care. That's right. They really don't care. So what that a few proprietary variables or properties cannot be seen through the standard protocol description. Virtually 99% of all customers don't care. They can barely manage their buildings with the tools they have now! They are overloaded with management issues and people details.

No. The BACnet discussions are needed but not sufficient. This topic needs to be explored more.