- February 2, 2004
- 474 views
There is no argument that top executives should not be required to understand the same level of detail that engineers and first line managers must have about specific processes that are enabled by technology. However, in a world that is increasing facilitated by data flow among automated devices, it is important for senior executives to understand the overall process and information flow and the resulting systemic vulnerabilities that can cause catastrophic failure to the enterprise.
According to a recent article in the Wall Street Journal, Communications Failures Cited By Utility Officials for Blackout, September 3, 2003, page B2, by John J. Fialka, both computer-to-computer and human-to-human communications failures contributed to the August 14, 2003 Northeast blackout. Increasingly, in a world run by real-time connectivity whether it is online commerce or automated power grids, failures hit the bottom line, not only by the loss of immediate revenue, but potentially through damage to reputation with the subsequent increased competitive pressures, as well as possible fines and increased regulatory oversight.
Flattening the Organization?
Pundits, including this one, have long pushed the hypothesis that information technology, when integrated with business process change, flattens the organizational hierarchy and can speed decision-making. However, the opportunity to accomplish this is dependent on several non-technology and non-business process factors. Good management has a high level of trust in their workforce; however, it is difficult to commit capital funds to initiatives that management does not understand. Such is often the case with IT funding. We are rapidly reaching an era when executives will need to understand both the impact and unintended consequences of information flow in the firm. On the surface, it would appear that lack of investment and a focus on electricity marketing were factors in the recent blackout. How many of these same executives will know how to appropriately spend the reported $100 billion needed to upgrade the system? Data reduction up the chain of command, such as described in the Columbia report, is common procedure. Top-level executives are very busy and do not need the details. However, they do need to be able to authoritatively challenge hypotheses put forth by those selling their solution to a business issue and sort out the bet your company situation. This is their job! In the final analysis, rapid decision-making is a function of smart, competent agents reacting quickly to changing events. When online systems are involved, the decision-making process involves complex data handling and control systems coupled with human intervention. As we all know now, major process failures in these environments can evolve rapidly.
For now, we must address this type of problem using todays capabilities, however, new approaches are evolving that may better facilitate decisions in the future. Innovation is the hallmark of the knowledge age, and despite its current problems, once again we look to the skies. Futuristic Decision Making
The space industry is working on sophisticated decision support models that incorporate procedural and declarative execution systems into spacecraft operations, such as A Hybrid Procedural/Deductive Executive for Autonomous Spacecraft, http://www.ai.mit.edu/people/williams/papers/agents98-hybrid.pdf. Spacecraft are inherently complex and operate in hostile environments. Decisions are made against documented procedures, which we have previously seen may have limitations, using automated systems to manage physical processes that may not be readily observed. As with many earthbound complex real-time systems, the default automated decision is usually the off position or shut down. The authors suggest that there is a hybrid method for dealing with uncertainty. Ambiguity Management allows planners to construct worst-case models that address a number of most likely scenarios against a most likely state of the world not just automatically shutdown; safety and environmental concerns need to be incorporated into such a system. Clearly, this decision support model requires substantial context integration with robust information handling systems.
While this may sound futuristic and even a fantasy, we may be closer that we think. Todays software is workflow enabled. Business rules (procedures) are routinely incorporated into information flow and data handling systems. The hard part is keeping procedures fresh or updated. The next step will be to determine what options may be available other than shut-down, when an automated system encounters a failure. Real options theory may have a role, particularly if the automated systems can provide real-time scenario adjustments to the humans making the ultimate decisions. Challenges to be Met
For most of its life, information technology has largely been thought of as a back office functions. Instruments, sensors, process control systems, and SCADA were part of the engineering and manufacturing process. Recently, the online society is integrating these systems as well as customer facing processes.
While there is much work to be done to realize futuristic decision support systems, unless senior executives recognize the need to get involved rather than deferring to technocrats, they put their organizations, and possibly supply chain partners and customers that integrate with them, electronically at risk. The Failure is Not an Option mantra, like all buzz words, is largely overdone. However, currently, NASA is reevaluating the Shuttle program, and Northeast electricity providers are undergoing additional scrutiny, facing potential fines and significant system wide upgrade capital expenditures. Corporate governance must extend into the field. Online systems upon which the organizations cash flow depend, fall under Sarbanes-Oxley. Like Y2K before, management can be held accountable for material adverse impact on the organization if online systems fail. The nervous system of modern business must remain robust and vital if the firm is to remain financially healthy. Management failure to understand the ramifications of their information systems on their business is not an option, but rather a recipe for destruction of significant shareholder value.
Once an organization goes online or real-time with its revenue producing and asset management systems, it is by default, and perhaps by law, obligating it to take good care of its nervous system. Good IT health begins at the top, is institutionalized throughout the organization, and continues throughout the organizational life cycle.