Smart Buildings

Winter 1996 in the beautiful land of Singapore, I was excitedly lapping up all I could learn of Structured Cabling.  That was when I first heard about Intelligent Buildings.  I was told that an Intelligent building is one which integrates multiple systems such as Office Automation (read Data network), Telecommunication (read Voice network) and Building Automation seamlessly.  The objective was to maximize performance, save on investment and operating cost and be flexible to accommodate change.  My imagination took flight when I was told that CMC Ltd., (earlier Computer Maintenance Corporation of India) had built an Intelligent Building in Mumbai to accommodate their headquarters.  I was determined to learn more about Intelligent Buildings.

The word 'Building Automation' had caught my immediate attention; I already knew a little about voice and data systems.  I was told that it included electronic and electro-mechanical devices and control systems to provide building services such as HVAC, FLS, SAC and EMS.  Oh! more jargons!  I got to understand that HVAC was Heating, Ventiallation and Air Conditioning; FLS stood for Fire and Life Safety, SAC was short for Security and Access Control and EMS expanded to Energy Management Systems.  I didn't know that someone in IT needed to be bothered about so many things besides computers, telephones and networks.  I slowly learned that Building Automation Systems too like Data and Voice network involved content such as data, voice, image or video carried over wires. The trick was in making each system respond to the data that was being carried from any other, so they all function in an integrated fashion to provide comfort and safety at optimal cost.

Over time, I also came to learn about some of the salient features of the CMC Building:

Lighting: The on/off and dimming of the artificial general lighting was controlled by a Microprocessor that responded to the available daylight. The Microprocessor controlled motors that operated venetian blinds on the upper half of the windows. These were reflective (mirror-coated) blinds that were tilted at pre-determined angles in response to angle of incidence of sunlight on different faces of the building.

Air-conditioning: The AC plant involved four independently controlled Air Handling Units (AHUs) located on separate floors allowing flexibility in use.  The Microprocessor Control system allowed sections being used to be cooled, when the building was not fully occupied. It also ensured automatic load management when building was to be run on standby power and full plant operation was not possible. It also ensured that AHUs will be switched off, ventillation fans will be turned on and lift lobby will be pressurized in the event of a fire.

Communication: Computers informed the EPABX of the location of the field worker, whose seat was allocated on a day-to-day basis to optimize seating capacity.

Johnson Controls Metasys systems (introduced in 1990), VingCard Security System, Honeywell DeltaNet and Excel Sytems and STEFA Control systems (later acquired by Siemens in 1998) were some early players who started with few domains.  HVAC involved Sensors, TouchScreens Thermostats and Controllers that control Air Handler Units, Unit Ventillators and Fan Coil Units.  FLS involved Heat/ Smoke Detectors, Fire Alarms, Sprinkler Valves & pumps, Anunciators and Controllers.   SAC involved Proximity Readers, Glass Break Detectors, Hall-Effect Door Switches and Controllers.

Today, with the change in environment, additional domains have come up; To name a few, Closed Circuit TV (CCTV) & Digital Video and Asset Tracking.  The former involves Cameras, Pan/Tilt/Zoom Dooms, Video Recorders and Video Manager Servers with features to detect Camera tampering, blinding, blurring and changed field of view.  The latter involves Radio-Frequency/ Infra-Red Tags and badges, Readers, Repeaters and Servers with features to track motion and deter thefts.  Integration and communication too has progressed by leaps and bounds.  Earlier vendors used one of the protocols such as BACnet (Building Automation and Controls network), Modbus, LonTalk, etc.,  Modern systems use an open interface to integrate devices from multiple vendors, that speak any of these protocols and connect over any of the physical layer technologies such as Arcnet, Ethernet, LonWorks and so on.

CMC House, Mumbai was a technological marvel, when it was commissioned way back in 1992.  Today's, state-of-art demonstration is Building 14 of Cisco's Bangalore India Campus, named BANYAN.  Banyan holds LEED (Leadership in Energy and Environmental Design) Platinum certification.  Its Operation Center integrates Safety and Security, Transportation, Utilities, Building Management, and energy systems such that they are managed together. 

Community Connect Services such as Office Resource Management provide a realtime update of all the services available. These functions provide intelligent systems that deploy user needed resources such as projectors, enable phone connections and supply power to electric outlets.  Digital Smart Signage systems save vital time and fuel by guiding people to the nearest parking spot. Energy management is obtained with realtime intelligent information driving consumption, switching off unneeded equipment for sustaining necessary services like Safety lights in public corridors during power outages. Daylight harvesting and optimal energy consumption reduces carbon footprint and cost. HVAC react and adjust based on occupancy.

Certainly we have not seen the best of how smarter buildings can get.  The future looks very exciting.  Cisco's 'Smart+Connected Community' offering is ever-evolving. IBM is another leader, talking about building a 'Smarter Planet'.  The technology is ever expanding, to include architecture, data-communication, energy management and many more to provide producitivity and efficiency for social, environmental and economic sustainability.  This is indeed THE space to watch!

Smart Diagnostics

It was winter of 1992. I had just begun to learn Computer Trouble-shooting. It was fascinating to see the kind of intelligence built into the ubiquitous Personal Computer. It had a way to tell the Power User or the Support Engineer what was wrong with it. It used to spit cryptic error codes like ‘201’ or ‘301’ to say if something was wrong with the Programmable Peripheral Interface on its motherboard or with the UART on the Serial Port adapter that attached to it through a Bus. But, how will it communicate if the PC was not functional enough to let the Power On Self Test (POST) use the display card? It had a way around this road block, too. It then used the simple speaker -- more appropriate to be called the beeper, given the polyphonic advancements with speakers now available on much smaller and cheaper devices. One beep indicated some failure and two indicated some other. A short beep indicated certain problem and a long beep quite another.

Over time, specific parts of the Personal Computer began to come equipped with embedded diagnostics. Hard drives were among the first subsystems to become intelligent. For instance, Compaq had come up with Intellisafe drives – drives that could think and talk! An Intellisafe drive monitored several aspects of its operation such as Spin Up time, Temperature, etc., to determine if it was moving away from health. It notified the user that the drive is about to fail, when it noticed any of the performance parameters falling below acceptable levels. Instead of having to respond to an emergency, the user now had ample time to backup his data and replace the drive under Pre-Failure Warranty. Subsequently, the standard came to be ratified as S.M.A.R.T. (Self-Monitoring, Analysis & Reporting Technology). IBM had been a fore-runner to Compaq with its PFA (Predictive Failure Analysis).

When heavily upgraded Personal Computers began to be used as Servers for File Storage or for Database applications with several other computers on the network dependent on them, ‘Availability’ became a critical requirement. Concepts such as RAID and Redundant Power Supply had advanced availability of individual parts by leaps and bounds. But what if the computing power itself suddenly became unavailable? Server Companies responded with Out-of-band management techniques, naming them Remote Insight (RI), integrated Lights Out (iLO) and so on. This allowed control over the Servers from outside the Local Area Network, when the servers stopped responding to input from within the network. When the Server hung and a virtual console application on one of the network’s desktop could not be used to restart the server, the only way to reboot the machine was to gain physical access to the Server Room. A very smart alternative for the administrator who was at home during off-hours or who was managing multiple sites remotely or had to manage servers in a hosted (and therefore physically secured environment) was to somehow gain access to the Server Hardware, as if he was physically next to it. A specialized card plugged into one of the Servers IO slots, with a dialup connection to it for the Administrator to access remotely from anywhere, allowed him such access.

Alongside Servers, networking devices such as the Modems, Switches and Routers brought about greater sophistication in embedded diagnostics. The LEDs on the Modem or the Switch indicated if the corresponding ports were working fine. A dumb terminal could be connected to the console port of the switch or the router to find out detailed information about how packets were being transmitted or dropped. With the emergence of networking, smart diagnostic technologies found their way from computing machines into connecting devices. However, the information churned by the networking device continued to be cryptic and required a highly trained professional to decrypt them and fix the issue.

The increasing complexity of telecommunication devices and the spiraling need for highly trained support engineers compounded several problems in the support industry. The cut-throat competition among System Integrators in an industry already dependent on expensive knowledge workers resulted in slender margins getting ever thinner. The dependence on skilled professionals meant that quality suffered whenever an engineer was lost through attrition. To address the Revenue-Cost dynamics and the demand-supply equation, today’s products come with very intelligent diagnostics embedded within. These not only provide meaningful information about the error conditions but also step-by-step instructions to resolve them.

Cisco’s Smart Call Home (SCH) is a case in point, providing higher availability through proactive, fast issue resolution. The built in diagnostics provide proactive, detailed and real-time alerts on core network devices and even remediation recommendations based on Cisco proven practices, to help identify and resolve issues quickly, conserving valuable staff time and improving network availability.

(Courtesy: http://www.cisco.com/en/US/services/ps2827/ps2978/ps7334/smart_call_home_overview.pdf) 

SCH provides increased operational efficiency through reduced troubleshooting time. When the problem is serious, it generates a Service Request to Cisco that is routed to the right team for the specific problem. Service requests to Cisco include relevant diagnostic and product information, so user does not have to repeat information when engaging with Cisco engineer. The figure above illustrates how a problem that earlier required 2 business days for resolution is now solved in under 6 hours.

The devices with SCH capability continuously monitor their own health using Generic OnLine Diagnostics (GOLD) technology and automatically notify user of potential issues even before a failure occurs, using secure transmissions. SCH gives the user the option of receiving proactive notifications of problems that are likely to be emerging issues, such as high temperature independent of any fan failure or accumulating single-bit memory errors.

SCH also provides quick access to information. It provides recommendations using email notifications and SCH web portal. Security advisories, field notices and end-of-life notices are personalized for the user’s hardware and software inventory.

From the humble beginnings of POST and BIST (Built In Self Test) all the way to the Smart Call Home, the world of Embedded Diagnostics has come a long, long way!  Soon, it is going to be TOTAL Machine To Machine (M2M) -- diagnostics embedded locally or on a remote device not just recommending a fix but also implementing it through automation.  Such days are not far away!

Smart Support Web

When I had just entered my final year in College way back in 1991, Artificial Intelligence was the Buzzword. I had picked up a book on Artificial Intelligence by an American Publication, hoping to execute a project in that realm. I had also got to hear that LISP (standing for LISt Processing) was essential to programming for Artificial Intelligence, and was on the lookout for a Center that would provide training in LISP Coding. I neither got to learn LISP nor execute my Engineering School project in Artificial Intelligence. I had to settle for C programming in the area of Control Systems (a simulation of sorts of a Programmable Logic Controller). It was very early days for AI and understandably I did not find much help within the academic community.

Just the year after I entered college, I had heard about experiments to build a Computer System that could defeat the World Chess Champion. Deep Thought had been pitted against Gary Kasporov but had lost. About 7 years from then, somewhere in 1996, Gary Kasparov got to try his human intelligence against artificial intelligence once again over a game of Chess, this time with Deep Blue. Betraying a prolonged infancy, artificial intelligence could not hold fort against Kasparov’s genius. It also spawned a lot of debate about Man Vs Machine. Questions like “What would Deep Blue do, if it won over Kasparov? Would it take Deep Pink for a date?” emerged. The world was stunned when Kasporov lost the subsequent year 4-2 to a heavily upgraded Deep Blue. A machine had finally defeated a Reigning World Champion in a match under standard Chess Tournament time controls. AI had dawned in a big way.

Today, when I look back, it has been a quiet journey towards intelligence and smartness, for a host of products, without the bravado and hurrah that surrounded the early efforts. A case in point is the Support Web. For a long time, Product Companies have utilized a Support Website, enhancing them with useful information categories, from time to time. What started with contact information of Support Centers and online Product Manuals and troubleshooting guidelines, soon had software patches, driver upgrades and fancy applications that would deliver greater value with their products, available for download. But this is nothing compared to the development that has been demonstrated in recent years. Today’s Support Website belong to a different genre altogether.

Some of the Smart Support Websites allow you to register your purchase and use that to customize your default support page. This saves you time when you look of troubleshooting information that applies to the model that you own. More advanced Support Websites remember your clicks, profile you based on your preferences and web-usage pattern and use that to customize your default page. All these Smart Support websites leverage community participation to varied degrees in support content creation. Most of them have Discussion Forums and Wikis where Subject Matter Experts (SME) respond to User Queries. Some of them allow the user community to provide solutions too. “Crowd-sourcing” as it is called, results in creation of pertinent support content. This is so because when users provide troubleshooting tips, it addresses a problem that they have faced and is therefore likely to be faced by another user. It results in scratching where it itches. Any amount of intelligence from SMEs can not address the field requirement with such accuracy.

Today’s state-of-the-art Support websites have diagnostics that talk to embedded applications on Products. They receive diagnostic error messages from the devices connected to network, use complex rules to diagnose the problem with the product and provide remedial measures that the user can apply to solve the issue. Soon, web-borne diagnostics would be fixing issues on Products that connect to them over the network, without human intervention.

The advancement in Smart Support Websites and the competition among Product Companies to leverage the web technology for Customer Service & Support has spawned a host of recognitions and awards. Technology Services Industry Association (TSIA) announces STAR Award for Best Online Support. It recognizes Support Websites that leverages Web-based technologies and knowledge management processes to provide an effective Web self-service experience, including driving customer adoption of online support, personalizing the experience, using customer feedback for improvement, and measuring customer success for ongoing improvement. Another group, The Association of Support Professionals (ASP) rates 10 Support Websites each year as the “Ten Best Web Support Sites”. Service & Support Professionals Association (SSPA) is yet another group that announces Awards in various categories such as Best Online Communities, Best Online Support, etc., Localization Industry Standards Association (LISA) announces ‘Best International Web Support Sites Award’ recognizing customer support provided in multiple languages. The Honors list for each of these awards point to excellent Support Websites that are the smartest in the world today.

Dell, Intuit, Cisco and Xerox have some of the best in class Support Websites that have revolutionized the support industry. Web 2.0 has indeed become a reality over the last few years. We could very well be standing on the threshold of yet another internet wave – Web 3.0. Watch Out!