 |
 |
       |
 |
Home | Site Map | Blog | Contact |
 |
|
|
Research Content / Research Articles / Planning, Execution & Project Management / The Real-Time Enterprise
Background & Context: The Real-Time Enterprise
‘Real-Time’ is becoming a primary design point for the Enterprise Architectures of many Global 2000 organizations as well as increasing portion of the public sector. The purpose of this article is to provide background to EA teams to better understand the origins and context of real-time systems.
From the late 1950s through the end of the 1980s, most professionals believed that core business processes had a ‘lifecycle’ on average of about every seven (7) years before a fundamental change to that business process was made. (In fact professionals in some industries believed that certain underlying core processes would be unlikely to change for as long as 20 years – such as customer billing by telecommunications firms, healthcare claims processing, etc.)
By the early to mid-1990s, most professionals had come to the realization that the rate of change in business processes was accelerating as a result of competitive pressures and technology advancements. For many, that acceleration was dramatic. By the end of the 1990s, it was not uncommon for a business executive to describe the acceleration of the rate of change in core business processes as having gone from every 7 years to as little as every 18 months. In some industries, the lifecycle of core business processes was even shorter. And if a CEO was asked how fast they wanted a major change to be made, most replied “yesterday”.
The acceleration of the rate of change in business processes is also described as the compression of business process lifecycles. It is illustrated below:
By the late 1990s, ‘Business/IT Alignment’ had become a hot issue. CEOs perceived IT as not aligned with core business objectives and strategies. CEOs were striving for corporate agility but viewed IT as generally unresponsive.
The view that IT was unresponsive was not surprising as:
The reason for the misalignment of business and IT is simple, the rate of change of business processes is significantly faster than IT’s capacity to deliver (solutions to newly emerging/changing business requirements). In the early-90s, Larry DeBoever argued that what was required were highly adaptive information systems that did not inhibit an organization’s ability to change processes. He developed methods for creating an adaptive enterprise architecture which, in turn, would enable the adaptive enterprise that he argued was created to be competitive in the then coming decade.
Many companies in the late 1990s embraced the concepts of business process reengineering (BPR) to address, in part, the challenges created by the acceleration of change in processes. BPR’s reputation is tarnished primarily due to the unrealistic goals (‘big bang’) that many BPR adopters often established for their re-engineering projects. However, the objectives of streamlining processes and driving efficiencies are still top of mind with executive leadership.
Since the demise of BPR, numerous companies have embraced the more incremental approaches of Six Sigma and Total Quality Management (TQM), and now Business Process Management (BPM), to address the need to streamline processes and drive efficiencies in response to the accelerating rate of change. Unlike BPR, Six Sigma, TQM and BPM efforts have continued to extend their reach in the companies that have adopted them.
BPR, Six Sigma, TQM and BPM projects consistently recommend that interfaces between disparate information systems be automated rather than having the output of one information system be re-entered into another information system (e.g. the output of an order entry system be entered into a sales force automation system; the output of a financial system be entered into a budgeting system). Three benefits are typically achieved by automating the interfaces between systems:
The rate of business change slowed somewhat from 2000 through 2003 as a result of several factors – burst of the dotcom bubble, Y2K hangover, 9/11, economic sluggishness, etc.
Today, a new, more intense wave of business transformation is underway. The driving forces include:
Global competition – the rise of China, India, Vietnam are just a few examples
Product innovation – for example, Apple’s iPod is in its 6th generation as a product even though it was introduced less than 6 years ago; the food sector estimates that over 50% of the products on grocery stores were introduced in the last 5 years
Cost pressures – due to competition, demands of large customers, pricing transparency (via Internet), etc.
Business morphing – such as the evolution of 7-Eleven from convenience stores to food service outlets
Emergence of new business models – such as e-tailing and virtual organizations
Emergence of new technologies – including the Internet, RFID, Wi-Fi, SOA, VoIP, etc.
Customer expectations – such as 7x24, on-line access to their information
And none of the above forces are abating.
To successfully respond to the acceleration of business processes and changes in those processes, organizations must increase the velocity of information throughout the extended enterprise. To achieve this:
The goal of many companies is to provide end-to-end integration across systems for specific business processes. End-to-end integration can enable business events that are ‘transactions’ (e.g. an order) to be processed straight-through, without delay. This is referred to as Straight-Through Processing (STP). Examples include:
Financial Services – the SEC mandated that financial firms fully settle a trade within one day (‘T+1’) rather than the previous standard of five days (‘T+5)
Telecommunications – firms are implementing systems that process a customers order for service and, in real-time, provision the desired service at the switch, as well as initiate billing
Health Care Insurance Claims – providers and claims processing firms are working toward real-times ‘claims-to-cash’ solutions in which a patient that is seeing a doctor for a relatively routine treatment will have the payment of the claim authorized and payment to the doctor completed before the patient leaves the office
Health Care Delivery – providers are undergoing clinical transformations and implementing electronic medical records that are always current
Autonomic Logistics – are being implemented on next generation aircraft (e.g. F-35) which enable a plane to continually self-diagnose and automatically schedule specific maintenance activities, as well as order replacement parts, while in-flight
A specific example is the end-to-end business process integration and interoperability achieved by Wal-Mart and Dell with their suppliers. These strategies are widely recognized as a primary contributor to Wal-Mart’s and Dell’s leadership in their respective markets. Over the last 24 months we have seen numerous companies attempt to replicate the success of Wal-Mart and Dell by implementing demand-driven supply networks (DDSNs).
In all of the above examples, a business event automatically causes other events in the value chain to occur.
A variety of technologies have been employed or otherwise adapted by companies to deliver this breadth of end-to-end systems integration including middleware, message subsystems, triggers, stored procedures, MOM, EAI, ETL, BPM, EDI, VPN and so forth.
The acceleration of business change requires executives, line managers and staff to increase information velocity so that the can make significantly faster decisions and, to the extent possible, make significantly better decisions.
As a consequence of this acceleration, the mid- and late-1990s were characterized, in part, by the explosion in the demand for data warehouses and the adoption of business intelligence (BI) tools offered by vendors such as Business Objects, Cognos, Essbase/Hyperion, Information Builders, SAS Institute, etc. Enterprise software vendors such as Oracle and SAP also presented their own BI solutions. Not surprisingly, Microsoft began to invest heavily in developing business analytics capabilities built on top of Microsoft’s emerging BizTalk capabilities.
The types of solutions we have been describing are referred to as real-time (information) systems regardless of whether the systems are performing the straight-through processing of transactions in real-time, or the systems are providing information to decision-makers in real-time.
The genesis of sharing information (over long distances and in real-time) dates to the inventions of the telegraph, telephone and radio. The concept of developing information systems that operate in real-time emerged as the computer age was born in the 1940s.
Not surprisingly, most of the early research was driven by U.S. military requirements. The U.S. Navy, and later the U.S. Air Force, sponsored project Whirlwind in the late 1940s and early 1950s specifically to develop a real-time, defense early warning system.
Following the launch of the Sputnik satellite by the Soviet Union, the U.S. Department of Defense (DOD) established the Defense Advanced Research Projects Agency on February 7, 1958[3]. The Agency’s mission since 1958 is unchanged.
DARPA’s mission is to maintain the technological superiority of the U.S. military and prevent technological surprise from harming our national security by sponsoring revolutionary, high-payoff research that bridges the gap between fundamental discoveries and their military use.[4]
DARPA’s role in the development of the concepts and technologies that underlie real-time systems over the last 40-50 years cannot be underestimated.
DARPA accomplishes this by funding universities and research organizations to conduct fundamental and applied research across a breadth of disciplines including communications, computer languages, operating systems, database systems, spatial data management, distributed database, artificial intelligence, messaging systems, voice recognition, predictive modeling, etc.[5] The development of technologies such as the original Internet (ARPANET), packet-switching, GUIs, relational database, mass storage, etc. all have been funded by DARPA.
The first large-scale deployments of real-time information systems in the commercial world were airline reservation systems, credit card systems, and ATMs.
Today, examples of real-time systems abound. There are ‘day traders’ who rely on automated logarithmic trading, GPS systems that track the locations of trucks, and Internet-based lending systems.
Cell phones and instant messaging have created a real-time communications environment for the masses.
In Fantasy Football leagues the moment an NFL player runs the ball, catches a pass, kicks a field goal in a game, etc. a Fantasy Football member who has that NFL player on their Fantasy team roster can immediately see the points that they have gained for their Fantasy team over the Internet (see for example the Fantasy Football leagues supported by CBS Sportsline[6]).
It is no surprise then that many people – from business executives to Generation ‘X’ consumers – wonder aloud ‘why can’t everything be real-time’?
There is no information system that operates in absolute real-time. There is always some latency between the business event and the subsequent action (e.g. spawning a transaction for the next business process, sending an alert, updating a management dashboard). The latency might be measured in increments as small as milliseconds or even nanoseconds but latency occurs. Information cannot be moved faster than the speed of electricity (which is the speed of light).
Said differently, real-time systems are those information systems which appear to the user to execute in ‘real-time’. In the 1990s, we published research at the META Group that described this as ‘near real-time’ (NRT). When an executive is looking at information that is ‘up to the moment’ they are happy, for most applications, if the information is current as of the last several minutes. Most real-time systems require a few seconds to complete all of their processing end-to-end, such as an ATM.
Real-time systems are over-hyped in the press and by industry analysts. Real-time systems are oversold by consultants, vendors, and well-intended IT staff.
The simple fact is that not all information systems will be real-time (at least for the foreseeable future) simply because they don’t need to be. For example, there is no demand for real-time payroll systems that constantly deposit payroll funds every second or every minute while employees are working. Similarly, there is no demand for asset tracking systems that track the location of all of a company’s physical assets in real-time. (A company might want to track inventory in real-time with RFID and locate their trucks in real-time with GPS, but they are not interested in tracking the location of every desk and every whiteboard and every stapler…in real-time.)
Clearly, not all systems need to be or will be real-time. At the same time, we believe that within five years (2011) most of the mission critical information systems of medium and large companies will be real-time.
The real-time enterprise is an organization that has the majority of their mission critical information systems operating in real-time. Today, there are few real-time enterprises outside of financial services and banking (and Fantasy leagues!).
We have identified several forces that are driving the demand for real-time systems including:
Two additional factors will compound the adoption of real-time systems over the next five years:
As a consequence of the above, we believe that most businesses are just exiting what is best described as the ‘Stone Age’ of real-time systems, and that within five years (2011) most medium and large enterprises will be real-time.
Conclusion
The need for increased information velocity and more event-driven real-time systems has an impact on not only the technology infrastructure and software needed to enable these capabilities, but also changes in the business processes and information environments themselves – exactly the type of problem that EA is intended to solve.
Directions: EA teams looking for more justification of their program, resources and investment should identify the need for increased information velocity, and more event-driven, real-time systems as a major capability requiring enterprise architecture to be successful on an enterprise-wide, sustainable and significant scale.
[1]See, for example, the writings of Dr. Steven Spewak, Dr. James Martin, and Colin White
[2]The first use of the term “Information Velocity” and related research was published by the META Group (now Gartner) in the 1990s. This research was performed by Larry R. DeBoever.
[3]The Agency’s original name was Advanced Research Projects Agency (ARPA). The name was changed to DARPA in 1972. President Clinton later changed the name back to ARPA in 1993, only to change it back to DARPA in 1996.
[4]June 23, 2006 http://www.darpa.mil/body/mission.html
[5]Organizations receiving DARPA funds over the years include MIT, MITRE, SRI, Xerox Park, IBM Labs, Computer Corporation of America, BBN, Stanford, Carnegie-Mellon, the University of California, CalTech, etc.
|
.jpg)
.jpg)
 |
|
 |