High Reliability Organizing in High Risk Industries Banner - Artwork by E. Steele, for the Wildland Fire Lessons Learned Center

Menu:

Updates:

Feb 22, 2009:
Set up new page

Other Models

Models Useful for High Reliability Organizing
Models and programs can help bring structure for thinking and acting.  In stable environments these adjuncts can help organizations succeed as variation may endanger the program.  Several dangers exist with use of developed programs.  We can become lost in the metaphor or the concrete presentation of abstract thought.  This is when rules become reductive.  Or we may lose the nuance that led to the model’s success in other environments.

When models and programs become products and replace adaptive thinking they create rigidity or a reliance on non-adaptive structures.  Consequently, complexity or dynamic environments will lead to crisis if not catastrophe from which the organization may not recover. 

The following are models for quality, safety, and reliability that interact with or contribute to high reliability.  The descriptions are only to introduce the concepts.  The reader is referred to primary sources for more information. 

Crew Resource Management (CRM)
In 1978 a commercial jet with a well-trained crew ran out of fuel and crashed killing 10 of the 189 people on board.  The NTSB identified a recurring problem in air crashes, “a breakdown in cockpit management and teamwork during a situation involving malfunctions of aircraft systems in flight .… because their attention was directed almost entirely toward diagnosing the landing gear problem.”  Also contributing were the failure of the captain to respond to the crewmembers’ advisories and the failure of the other two flight crewmembers to successfully communicate their concern to the captain.   

NASA then studied the question of how an experienced, well-trained crew can run out of fuel in flight.  This occurred, basically, because the pilot refused to take recommendations from junior officers and narrowed his focus to a single problem when under stress.  The corrective program NASA developed became Cockpit Resource Management.  Because of problems identified in the 1990s with communication between crew on the flight deck and passenger areas, it became Crew Resource Management. 

CRM is designed to avoid, trap, and mitigate error through communication, leadership, and decision making.  The original problem involved extreme authority gradient and coning of attention under duress for which CRM works well.  Its success derives from a change in values, beliefs, and behaviors.

Because changes required for CRM cannot be imposed on people it will fail when the professional does not accept and internalize the approach.  The officer can perform well in training and testing but revert to maladaptive behaviors in flight or under stress.

Others who use CRM inadvertently reduce it to methods of communication (almost by checklist) and change in a few behaviors.  This misses the methods onerous leaders have of maintaining authority gradient such as intimidation by countenance or tone of voice.  It also ignores the difficulty of changing behaviors that have major social ramifications.  These subtle or nuanced culture changes may be the source of CRM’s success. 

Recognition-Primed Decision Making
This is a model described and codified by Gary Klein to explain how people make quick and effective decisions.  The decision maker generates actions based on past experience or training then selects one most likely to succeed.  This can occur rapidly and below the level of conscious thought.  Unusual or complex situations, however, can lead to error.  This model describes the difference in speed of decision making between experienced and novice decision makers as well as the need to understand experience.

Six Sigma
Believing that reliability comes from consistency, Motorola developed this approach for internal quality of control and to increase their return on investment.  They found that, because of the complexity of their products, the level of 3.4 defects per million products was a small enough number of defects that this exceptional performance dramatically reduced complaints about their products.  GE then adopted and successfully used the system in the 1990s.

Six Sigma (6σ) is a statistical term that refers to 3.4 occurrences per one million events.  Statisticians use sigma (Greek letter σ, represents a standard deviation) to measure how a sample deviates around the mean (average or norm).  Through a mathematical equation the distance between sigmas (standard deviations) on the X-axis is the same but the area under the curve changes.  The area under the curve represents the number in that portion of the sample.  When graphed it forms a curve called a “normal curve” (the technical name for this is the Normal Gaussian Curve or Linear Distribution, see graph below).

  Normal Curve

Six Sigma describes the area of the curve that is six standard deviations from the mean or 0.000 34%.  Because of the nature of the curve, while six sigma is 3.4 occurrences per one million events, five sigma is 233 occurrences. 

It has the assumption that risk and errors develop from variation in processes which will propagate through the system to reduce the quality of the product.  The organization will then benefit by identifying and controlling variation in those processes that most affect performance and profits.

In mechanical processes inputs, transforming processes, and outputs are related in a deterministic manner with each directly determining the outcome.  Cause and effect are direct and have great importance in identifying where to intervene.  If you determine the cause of the problem you can fix it.  The cause will be from variation (deviation from the expected) in the process that results from error.

In Six Sigma, variation, or error, is to be identified and mitigated.  It is represented on the normal curve as wider than desired curve.  A wide curve represents a larger number of products or processes in the sample that vary from the expected.  This variation prevents parts of the system from working with each other just as you would see if people came to work with different languages, training, clothing, and expectations. 

To identify and reduce variation, Six Sigma programs will analyze root cause and develop corrective action with a program called for Define-Measure-Analyze-Improve-Control. 

Lean (Toyota Lean)
During recovery from World War II, the automobile manufacturer Toyota developed the Toyota Production System, a system that describes a more effective and efficient organization over time compared to similar organizations.  This would produce a lean organization (absence of fat) which became Toyota Lean or Lean in the 1980s.  Their primary orientation is to give value to the customers and build sustaining relationships.       

Value is defined by the customer; Toyota would study and provide that value.  An organization that uses Lean understands that it is selling value and must add value to the product or process.

A Lean organization considers people as more valuable than machines.  This respect for people builds relationships inside and outside the organization.  The organization relies on the resulting highly motivated people within their company because such employees will more likely identify a problem and speak up to fix it.  This is comparable to the HRO’s deference to expertise. 

Lean organizations constantly look for ways to improve, commonly by maintaining vigilance for problems much as an HRO would with pre-occupation with error.  The Lean organization uses flow for its processes rather than batch steps, a source of the Just-In-Time delivery method that replaces warehousing or stockpiling materials.  Identifying obstructions to flow will uncover sources of waste and identify areas for improvement.

Kaizen, a term used in Lean, refers to a Japanese philosophy of improvement through all aspects of life.  Kaizen is the process Lean uses to improve the system and give value.  Value, relationships, motivation, and flow make a Lean organization.

Human Factors
Human Factors has a very broad meaning that refers to an interdisciplinary profession that applies a blend of engineering and behavioral sciences to solve complex human – machine interface, human – complex systems control, and human to human communications.  It covers individual, team, and only lately organizational performance.