By Dan Daley

Like other areas of technology, building automation is experiencing both a convergence of systems and applications on one hand and a revolution in usage on the other. Integrated whole-building automation systems, like some tweaky Holy Grail, are the goal of systems designers looking to put security, access, HVAC, fire suppression, emergency systems operation, elevators, and other systems onto a single digital network: That’s the convergence. The usage dynamic is allowing greater yet more secure access to that integrated network via wireless controllers.

Ultimately, building control will be networked, using LANs, with wireless nodes built in and positioned in a security-based hierarchy. However, given the complex and costly nature of each system - HVAC for a large building can easily top the $10 million mark by itself - individual systems will almost certainly be standalone for the foreseeable future. It’s just not realistic to expect companies to develop core competencies across the spectrum of control needs.

BACnet and LonWorks
The integration solution lies in the networks themselves. Two systems, BACnet and LonWorks, can act as the unified protocols that link the control of these disparate function systems. BACnet was developed by a consortium of standards organizations - ASHRAE, ANSI, and ISO. LonWorks, on the other hand, is a proprietary protocol, put forth by Echelon Corp. (www.echelon.com) and promoted via its Open Systems Alliance (OSA), an opt-in pool of systems manufacturers and integrators that subscribe to and make systems that are compatible with the protocol.

Both BACnet and LonWorks are open-architecture systems that can operate over a LAN and use standard interface connections, including ARCNET, Ethernet, RS-232, and RS-485. The choice between them is like picking PC versus Mac; both offer control coordination that brings the same benefits of networked operation: reduced redundancy of cabling, fewer control points, less systems overlay, installation cost savings, and improved operating efficiency.

But the future will have speed bumps. Not every type of building function lends itself easily to networked operation. For instance, security systems by nature tend to resist simple network integration - proprietary programming code language greatly contributes to making them secure in the first place. However, systems can be adapted to at least bring them partially to a central control model.

Then there’s the issue of so-called legacy systems in the case of facility renovations and infrastructure rebuilds. Analog and digital control systems are generally not a good mix and become less so going forward, because the essence of digital is the ability to upgrade software to increase overall system performance and the essence of analog is hardware. Thus, renovation designs should plan for the contingencies of both integrating older systems with new ones or just ripping out analog systems and replacing them with fully integrated ones.

Another reason to go with an all-digital integrated network is that the consistency of the GUI (graphical user interface) can be extended to wireless devices such as PDAs and pagers. Building systems operators want to see remote GUIs that can transfer the same amount of data as the stationary control panels. Using layer-page GUI design, which uses links on a main operations page to call up full-page versions of more detailed operations, they can. The architectural imperative here is to acknowledge that cabling to create a web of reliable nodes to feed these wireless devices will require extensive wiring, particularly around natural barriers to wireless signals like the shafts of elevator banks.

Because of the complexity of each type of building function system, the need for integration will drive the necessity of cooperation during the planning and design stages. The sooner decisions about type of systems can be made, the sooner compatibility inquiries by the systems integrator can be made and decisions about which protocol to use can be settled. This is where an experienced systems integrator can be the architect’s best friend. It’s also why we’re seeing more reliance on a design/install approach - one entity to specify, design, and install the system, reducing the risk of multiple links causing the chain that is the network to be less than what it should be. .