Real Innovation
Search:

Email This Page

Return To Previous Page
From: 
Please enter your email address above.
To: 
Please enter recipient's email address above.
Subject Line: 
Please enter a subject line for your email above.
Additional Notes or Comments: 
Anti-spam: 
2378

Please enter the number shown.

Page: 

Local Problems Lead to Ideal System Solutions

By Michael S. Slocum

Problem solving at the sub-system, local level can be inefficient. By defining a problem and attempting to solve it without regard for the entire system, there exists the risk of local optimization, with global non-optimization. This happens when the ideal solution of the existing problem does not contain elements of the ideal system. It is key to remember the objectives of developing the ideal final result for an existing system (ideality requirements):

  1. Removes original disadvantage(s)
  2. Preserves original advantage(s)
  3. Does not introduce new disadvantage(s)
  4. Minimizes (any) increase(s) in complexity

These objectives may be suspended for a particular generation, but not for the system model in its entirety over the total life cycle.

All too often problem-solvers are concerned with resolving the critical issue at hand, resulting in the ignoring of system considerations. Ideal resolution criteria are created that are associated with the micro-system that contains the problem. This helps to characterize a local solution, but this is independent of the system's ideality requirements. The ability to direct system evolution is lost when responding to problems in such a tactical fashion. The strategic and tactical implications of problem solving in the system at large need be considered.

Problem solving with little or no system consideration creates more problems for the solver; these problems typically are observed quickly after non-system solutions are implemented. A rush to fast solution generation actually inhibits the efficiency of the problem solver. Sometimes local vs. system requirements can contradict each other, but it is then possible to make intelligent decisions concerning solution selection. Consider an example:

Problem: Fuel in a hydrazine fuel cell vaporizes at very low temperature causing an inefficient utilization of fuel in space
System: Hydrazine fuel tank for in-space satellite propulsion
Ideality requirements:

  1. Increase fuel usage efficiency asymptotically to 100 percent
  2. Preserve the reliability of the existing system
  3. Minimize any solution-driven increase in system complexity
  4. Solution must be reliable

With this problem and these ideality requirements, solutions at the problem level can be formulated. Potential solutions include adding heating devices to the fuel tank to cause condensation of the fuel in cold conditions (in space). If this solution satisfies the ideality criteria (1-4) better than any other potential solutions, it would be implemented. The problem-solution decomposition looks like:

P: Fuel in a hydrazine fuel cell vaporizes at very low temperature causing an inefficient utilization of fuel in space
S: Add a heating device to the fuel tank to cause condensation of the fuel in cold conditions

There are new problems created by implementing solution, S:

P': A heating device requires power to function
P'': Feedback system is required to monitor fuel state in order to control operation of the heating device

Assume the secondary problems (S' and S'') are resolvable and that introducing the heating system solution is an effective resolution. Although this addresses the initial problem, S, how does it impact system level ideality requirements? The problem solver is not sure as these were not identified the solution to the problem, S, was generated in a strategic vacuum. The problem solver must look at the system ideality requirements and determine what impact, if any, will be experienced by ignoring them during problem resolution for problem S.

The hydrazine fuel cell sub-system is part of a satellite system that also includes (not an exhaustive list) a power supply, power generation system (solar panels), instruments for scientific study, radiation shielding, control moment gyros and other ancillary systems. The fuel cell is just one part of this complex system. A fuel cell problem resolution was critical in order for new missions not to suffer from the same failure modes as previous systems. Resolving the problem without consideration for system effects, however, may be detrimental. The system ideality criteria may look like the following (a selection):

    5. Power requirements must be reduced to increase an instrument's operation life cycle
    6. System heat generation must be minimized due to heat generated instrument noise

If solution S is implemented, it violates both ideality requirements, 5 and 6. In addition, the identified secondary problems can and should be expanded to include at least two additional problems:

P''': The heating system requires power for operation. This will reduce instrument operational life cycle.
P'''': The heating system creates thermal energy that will interfere with instrument operation

By looking at the secondary problems, P'-P'''', it is clear that the original solution, S, is unacceptable. If all the ideality requirements, 1-6, were included in the original problem-solving exercise, then solutions that met all of them would have been generated. This moves into a solution space that may include the solution, S, but would also include others that may satisfy the resolution of the problem, P, as well as the system level ideality requirements, 1-6.

Conclusion

In this example, considering only the ideality requirements of the sub-system (fuel cell) allows for the creating and selecting of solutions that do not evolve the system (satellite) toward idealness. Sacrificing system evolution may be required to resolve a problem quickly, but this should happen intentionally realizing all the secondary problems generated. When comparing a solution against all sub-system and system ideality criteria, it is possible to develop a more accurate perspective of the total impact the solution will have on the system when implemented. 

About the Author:

Michael S. Slocum, Ph.D., is the principal and chief executive officer of The Inventioneering Company. Contact Michael S. Slocum at michael (at) inventioneeringco.com or visit http://www.inventioneeringco.com.

Return To Previous Page