Accepting Change
By Brian Whitehead
April 2000
Aphrase often heard in the rulemaking business is, "If it ain't broke, don't fix it." When it comes to maintenance, that's not a bad rule to live by. AMEs have learned the hard way that machinery works best when you don't mess with it. As a result, they also tend to be fairly conservative regarding the rules that govern their profession. However, when dealing with regulations, modification provides a better metaphor than repair. Repair, at best, can only restore a product to its original condition; modification aims to improve it.
Years of improvement
We know about product improvement, because aviation hardware changes constantly. Between 1950 and 1970, the number of air transport accidents
per 100,000 departures fell 90 percent due to technological changes such as jet engines and solid-state electronics. We thought the trend would continue. It seemed that every new model would be better, faster, stronger. Unfortunately, the law of diminishing returns was at work. Over the past couple of decades, the accident rate has stabilized. We have reached a point where each new safety innovation increases the complexity of the product a little more, adding its own set of failure modes.
Technological improvement hasn't come to an end, just moved to other areas, as in range, payload, and economy. Those successes have created such growth in air transportation that, despite the low rate of accidents per departures, total annual hull losses are now three to four times higher than in 1960. If the commercial fleet continues to grow as forecasted, with today's accident rate, we will be seeing twice as many losses in 10 years. If the status quo is unacceptable, then change is inevitable. We need only decide what form the changes will take. The chances of getting further safety returns from technology are now pretty slim. We may have more luck on the human side, which explains the current interest in human factors studies. The maintenance industry has taken to the idea of human factors training fairly well and understands the benefits at the individual level. We must now take things to the next level - identify and eliminate the potential causes of "organizational" accidents.
Organizational accidents
Accidents seldom result from a single incident. Often they're the final occurrence in a long chain of events. The sequence of errors may involve
peer-to-peer relationships, but can run vertically or diagonally through the organization. Often, the chain of errors begins on mahogany row. Senior managers must understand that decisions made far from the flight line can create latent failures. These problems can remain dormant for years, only to bite when the right set of compounding circumstances occurs.
Clear communication is one obvious cause of organizational accidents; a rational accountability structure is another. Each person must understand clearly just what his or her responsibilities are, and those responsibilities must be realistic. There is no point in making someone accountable for what they cannot control. That is why the existing CARs go to such lengths in trying to distinguish between the roles of the AME and the approved organization.
There are other circumstances that suggest a need for change: outsourcing of maintenance, international aircraft leases and partnerships for aeronautical product manufacturing, and a mobile workforce.
Notwithstanding all these factors, many of the practices and assumptions that underlie our maintenance systems are still based on old models. We need to know which work and which do not. Every proposal must be assessed on its merits and, if we are to reject an idea, let's reject it for cause, not merely because it's new.
Let's use the right metaphor. We're not fixing something that isn't broken. We're trying to make a good thing better.
