Hello again, here is a little about myself that might shed light on this BMS issue.
I began with Johnson Controls, Sacramento California office, designing pneumatic controls systems for large buildings and industry just as the very first electronic interface controls were coming on the scene. 1972 approximately.
The first IO alone was massive... since then things have gotten smaller but not less complex... indeed they have gotten more complex.
I am 100% in favor of controlling all functions of a remote machine through the BACnet controls system... but ONLY if does not get too complex with 90 smaller systems attached so that one failure can affect the entire building, and rocket scientists have to be flown in to sort it out at $250 an hour...with a two day wait time.
Meantime 90 tenants will be fuming at the designer and the building owners and management. These will not so gleefully pick up the horn and let you know that are, err a few little problems.. which are costing them tenants.
Accordingly I attempt to follow the KISS principle (Keep It Simple Stupid) whenever possible, especially when it comes to controlling small complex systems remotely. Something like a Cadillac with 16 power ashtrays... too much is too much if the ashtrays all start opening and closing every time you turn the windshield wipers on.
A while back I was called in to debug and provide analysis on the Denver Air Ports now infamous luggage handling system that had cost the air port 20 million in lost revenue, as luggage was slung off of the conveyor system and out onto the work crews from high over head.
Baggage jams were occurring twice a day (I did the math on that for them, related to their bragging rights that they had over 10,000 interface devices and relays all with a 99.99% reliability rate)
That was my first clue.
You can do the math on 10,000 of those devices with a 0.01% failure rate, and see plainly where the two disasters a day was coming from.
Long story short, I prefer the simplest possible relay connection to the peripheral units wherever possible, even at the expense of some monitoring bells and whistles, so that when a unit fails... diagnosis is not complicated beyond belief, and so that a failure of the primary controller will not shut the peripherals down with no means of bypass.
Lacking that, the buildings operating engineers (who are not electronics guru's as a rule) have the pleasure of wading though this 138 page manual... my view is that such is not viable.
Good luck with your project, if you have trouble with it later I will be available to remediate if needed, Meantime I will opt out in the hopes that you can find a more accomplished BACnet guy.