The Rise of Smart Buildings
Roland Piquepaille writes "In a very well-documented article, Computerworld describes the current status of building automation systems (BAS) that control heat, air conditioning or lighting and how these systems are merging with traditional IT infrastructures. Computerworld writes that they're not enough standards in this industry and asks a fundamental question: who will administer these building networks, IT or facilities managers? Take for example Yale University which wants to connect 210 campus buildings, but also wishes "to integrate the BAS with the university's accounting system for billing and chargeback." Imagine the security risks involved with such an approach. This shorter summary contains selected excerpts of this must-read article."
The Rise of Smart Buildings Just add water!
THIS IS A ROLAND PIQUEPAILLE ARTICLE
Here is the "short summary":
The Rise of Smart Buildings
In a very well-documented article, Computerworld describes the current status of building automation systems (BAS) that control heat, air conditioning or lighting and how these systems are merging with traditional IT infrastructures. Computerworld writes that they're not enough standards in this industry and asks a fundamental question: who will administer these building networks, IT or facilities managers? Take for example Yale University which wants to connect 210 campus buildings, but also wishes "to integrate the BAS with the university's accounting system for billing and chargeback." Imagine the security risks involved with such an approach. Read more...
Let's start with a an assessment of the current situation.
As building automation systems (BAS) that control heat, air conditioning, lighting and other building systems get smarter, they're converging with traditional IT infrastructures. Emerging standards are enabling data sharing between building systems as well as with other business applications, improving efficiency and real-time control over building operating costs. Information security concerns, immature standards, the reluctance of vendors to give up proprietary technologies and ignorance among IT professionals of the convergence trend are all slowing the pace of this transformation, but it's gathering momentum.
But who will control such networks? And are there enough standards in this industry?
Open standards are just beginning to evolve and will likely break down the silos between building systems ranging from physical security to elevator controls. And the data from those systems is likely to be shared with other business applications such as the accounting system. This will allow for more-efficient buildings as applications are developed that can capitalize on newly converged data streams and real-time access to data.
[Right now,] standardization has started from the bottom up. Proprietary cabling systems in networks that link sensors and other devices to controllers on individual floors have given way in recent years to two competing, open protocols, BACnet and LonTalk, while floor controllers are migrating onto IP backbones.
Barry Haaser, executive director of LonMark International, says LonTalk and BACnet will prevail at the device level for technical and cost reasons. Others aren't so sure. "Instead of two guys running the IT and controls networks, why not one guy? I see IP going down to the individual device," says Anno Scholten, chief technology officer at BAS vendor Plexus Technology Ltd. in Irving, Texas.
IT infrastructure used in building automation systems This diagram shows how "building automation systems today rely on open, industry-specific protocols such as LonTalk (shown) or BACnet for device-level communications. But they increasingly leverage Ethernet and TCP/IP for home runs back to the control systems." (Credit for image and legend: Computerworld).
Let's take the example of Yale University to see how complex can be the merge between control systems and IT infrastructure.
But sharing the IP backbone raises security concerns among network administrators. Yale University is starting a project to consolidate its BAS onto an IP network that will link 210 campus buildings, and it plans to tie the BAS into a room-scheduling system that will automatically control energy usage based on room occupancy. For security reasons, Bill Daniels, manager of systems and technologies for the university's facilities group, has created an isolated, parallel network that's protected by firewalls and uses nonroutable IP addresses to keep data off the Internet.
Jerry Hill, director of systems engineering at Yale, says security is paramount. "We don't want a student to hack into our building management systems just beca
No, IT will NOT do it. You all might want to think that IT has claim to it, bit they won't.
The facilities management and Operating Engineers will be doing it. Just like they run building services now. I can imagine a bunch if CS/IT geeks going up against a bunch of good ole union boys about who has rights to this domain.
The unions already have been training and educating their members in 'modern' methods. This is really is the next step to inteligent controllers for things like chillers, boilers, elevators etc...
And to top it off, the integration will not be done in house, it will be outsourced to contractors.
Don't believe me: Elevators controllers are extensions of CNC controls. Modern chillers already have integrated controllers that include remote controls and monitoring.
Building engineers/Facilities ALREADY have CAD databases of newly installed equipment, piping, steam generating systems.
I'm an IT person. I've seen a couple of these systems. In general, they use networking simply to transport data between HVAC controllers.
In my experience, the customers have multi-building networks. Within each building, all the HVAC sensors and controls are all wired to a central control device, not over the network. The control device is typically some solid-state box bolted to the wall, not a PC.
All these boxes talk back to a central server (crummy PC with BAS software) over the WAN. The server then tells the boxes what to turn on and off and sends out alerts if something goes wrong. The alerting is basic, no SNMP or emails. A pager if you're lucky, but probably just a flashing message on the screen. My understanding is that there are some default settings the boxes can use if they should lose connection to the server.
As for this being an area for IT to take over, I don't see it. The vast majority of the work involved is with wiring HVAC sensors and systems back to the controllers and in programming the settings into the BAS software on the server. There is very little IT knowledge required. If you can program a cable modem router, you probably have enough IT knowledge to program the IT part of these things.
To me it seems strange that this article does not mention DALI (Digital Addressable Lighting Interface). It is a new standard for Lighting control that is sure to shake things up. Especially when you see DALI is currently being pushed by lighting manufacturers such as Osram, Atco, Helvar and Phillips to name a few. Dali places a fair bit of intelligence at the lighting fixture, and can be easily intergraded to TCP/IP networks using such systems as Atco's windim@net. This allows for remote monitoring and control. Better yet DALI can be wired using standard 240v insulated cabling and can be run next to the mains wiring (no segregation). It really is a smart step foward for lighting control. Check it out http://www.dali-ag.org/
Oh, come on, I can't be the only one that saw the RoboCop miniseries! Well, maybe I am.
"Once we've identified and embraced our sickness, we'll have strength...and that's when we get dangerous." - John Waters
"Until then, it's a higher paid salary taking care of it."
Don't be so cocksure about that. I'm part of a group of buildings professionals that provide infrastructure support for a major telecommunications carrier, and I can tell you for certain, our pay grade is superior to that of most IT managers and engineers in the company.
With regard to BAS systems, they are far more complex than networking gear. The systems we support typically employ a wide variety and large number of sensors, communicating on different protocols. The code is customized for each facility, since it must interface with and control equipment from multiple vendors. These systems require maintenance and troubleshooting skills atypical of most IT professionals.
I shudder at the thought of IT maintaining our BAS systems. One thing most of the IT professionals I've assisted over the years fail to realize is that electrical and mechanical infrastructure are the underpinnings of the rest of the facility. When you're out of power or cooling, you're out of business, plain and simple.
This isn't the sig you're looking for... Move along.
I work within an industry that would supply some of the stuff to be used in so called "Smart structures" I've been to the conferences, I've talked to the people.
.1 Hz, if that doesn't work, give it up. I don't want everything else to be a similar way.
It all falls apart because of cost. I can control anyone of the different systems in your house/building, I can monitor any variable you want. But the cost point is much much higher than you would expect. Even for a small 2000 sq foot house to monitor each room, control lighting and so forth would cost tens of thousands of dollars installed.
Then it still wouldn't work well, because I can only monitor so many different variables, and there are too many exceptions.
For instance. I worked in a semi smart building. Part of it was that motion sensors turned the lights on and off. If it didn't sense movement for 5 minutes or so it shut off the lights.
All well and good until someone is in the bathroom alone for a legngthy constitutional. Then your trapped on the can in a dark room.
Yes a minor example, but just one of many thousands of issues that come up. Say heating. You want to lower heating in rooms that aren't used, so you lower the temprature. Then someone comes in to work for an hour or two. Well rooms don't heat up immediately, so while the system is trying to heat the room up, the person says, "It's to damn cold in here" and goes somewhere else.
Or the opposite, you turn off the air conditioning, then people avoid that room because it's hot and muggy, next thing you know you've got mold in the walls.
All for a system that costs tens of thousands of dollars for a small building, hundreds, or millions for a large building.
As to networked appliances. Who want's to update the software in their toilet so they can use their microwave.
Who wants to find out out their boiler has a bug that shuts it's down under certain conditions, only to be told.
"It's a known problem, it will be fixed in the next release."
Who wants to have to re-boot their stove.
For some things simple analog controls work fine, things like on/off switches, potentiometer based volume knobs, and tuning knobs.
I can't stand the current generation of car stereos because the volume goes in steps, either just a little bit too loud, or a little bit to quiet. In the old days I could fiddle with the tuning knob to get in a hard to reach signal. Now I can only go up or down
In the UK we call such systems BMS, Building Management Systems. It amounts to vaguely-intelligent way to manage building energy consumption; that is the sole remit. Realise that, while there may be ways to access the info remotely and thus expose the system to security risks
The real point is to monitor boiler firing cycles, and window-openings (night-purge cooling etc) remotely to minimise running cost.
;)
Yes, it's great. I can watch, in real-time , the window management of a school I designed two years ago, from a terminal 200 miles away. I can learn from it, in terms of how the building is really used, as opposed to how it was assumed to work. Can I over-ride choices? No, and neither can any one else by 'hacking' the system. The truth is, BSMS systems are dumb - they are pre-programmed and (at best) report. No-one (esp. the investors) is actually interested in spending for IP addresses for the windows on the Arts wing, the necesary actuators and so on. I can monitor these things only because the necessary sensing is already part of other systems - like the alarm systems.
To everyone who wants to set off the sprinklers at their High School: please realise that sprinkler heads are purely reactive and work solely on rate-of-rise of temperature; they are not remotely addressable. Smoke sensors, on the other hand, can be
No offense, but the majority of the posters so far have no CLUE what they're talking about. I work for one of the largest building automation companies in the US (and world) and write the logic for programmable controllers. First of all, there is no way you'd EVER want an IT department taking control of your HVAC system unless they've been trained VERY well and their building's controllers were programmed with an IT department in mind. I started off in the IT world, and thought moving to building automation would be a cinch, but let me assure you, there was a huge deal to learn. When you're dealing with Chillers that can blow a cap that costs $10,000+ to replace, just because you accidentally allowed a chilled water valve to open up while your return water was still too hot after the a switchover in your 2-pipe plant... well... let's just say you want a building maintenance guy dealing with these situations. As far as standards go, just because IT people don't know about them hardly means they don't exist. The most prevalent standards today are the ones mentioned by the article - BACNet and LonTalk. Both are fairly simple protocols that allow for efficient communication over a wide range of network media. They were designed with slower networks in mind, so that if your bandwidth is only 100k/sec, you're still going to be fine. Usually the controllers are on a slower copper wire network, and then routed through an ethernet network to the frontend computer. Personally, I'd like to see the business go towards using standard ethernet and tcp/ip the whole way through, because of the lower costs of standard routers, repeaters, bridges, etc. As far as security concerns go... If they're worried about someone hacking in to their HVAC system and harming things, then their system was progammed shoddily. A well programmed system always takes into account the stupidity of users. You place safeguard upon safeguard upon safeguard. Even guys that have been facility managers for years will try to do stupid things, so you plan ahead and only let them make non-harmful changes from the frontend.