So much data

July 1, 1998

So Much Data

So little information

By Bill de Decker

July / August 1998

Bill de Decker is a Partner with Conklin & de Decker Associates, publishers of aircraft operating cost databases, MxManager® integrated maintenance management software, and consultants on cost analysis and fleet planning. He has over 35 years experience in fixed and rotary wing design, marketing, training, operation, and management. He also teaches a number of aviation management courses.

They say we live in the information age, but most of the time, all we get is reams of data that do not tell us much of anything. In fact, data is pretty useless from a management point of view, until someone has tabulated, organized, and analyzed it. Because, that's when data becomes information you can use to better manage your operation.

The maintenance of aircraft is an excellent example of this. Consider the drawers full of work orders, log book pages, part tags, purchase orders, inventory records, etc. that reside in any maintenance department. Usually, this data is carefully filed and never looked at again unless there is some specific maintenance problem or question. And so, the data sits in a filing cabinet, until moved to long term storage or thrown out. Obviously, it is very important to have all of this data available in case there is a problem, but it does not help one bit in managing the maintenance department. And yet É that same data contains a treasure-trove of information, waiting for the smart manager to mine and put to use.

One way good information can make you a smarter manager is by focusing attention on the things where extra effort will make a big difference. We've all suffered from examples where that focus was on the wrong thing. For example, one company I worked for had a policy, and I'm not making this up — that you couldn't get a new pencil from the supply room unless you turned in a pencil stub less than 2.5 inches long! The policy didn't last very long, but while it was in effect, I'm sure it saved a bunch of money on pencils. However, there was also a complete mismatch between the effort required and the impact on the company's bottom line.

There are a number of techniques used to match effort and impact. One frequently used technique divides a task into its sub-tasks, and then focuses on the five most expensive ones. Typically, these five most expensive sub-tasks will account for 60 to 90 percent of the total cost of the task. In short, this technique focuses attention on the high impact items.

A second technique to match effort and impact is to list the cost for every occurrence of a particular maintenance action for a specific component or task, such as the overhaul of a gearbox. This focuses attention on the variation in cost that often occurs from one overhaul to the next. The goal, of course, is to understand what causes these variations.

A third technique is to benchmark the various cost factors experienced by your operation against data from other operators, databases and manufacturers' published data. The idea here is to see how well you are doing when compared with others and to understand what is causing the differences.

Our company has analyzed work orders for tens of thousands of flight hours covering various fixed and rotary-wing aircraft. Some examples of the potential impact of this type of analysis follow:

One analysis we did of work order data for 4,600 flight hours in the life of an 8-seat small jet, focused on the cost of parts by ATA chapter. The five ATA chapters that consumed the most parts were as follows:

Cost Percent of Total Parts Cost Engines (Chapters 72) $262,247 24.6% Navigation (34) $154,885 14.5 Engine controls /accessories (73-83) $126,862 11.9 Landing Gear (32) $126,932 11.9 Air Conditioning (21) $ 77,211 7.3

These 5 items account for 70 percent of the total cost of parts for the aircraft and are clearly the area to concentrate on for cost savings. The listing quickly confirms that whatever effort this operator can spend (better negotiations, more training) on reducing engine overhaul and repair costs, is well worth it.

One surprise is the cost of parts for the navigation system. Further investigation showed a large part of the problem was the fact that the operator had installed 12 vertical gyros over the 4,600 flight hours, instead of the one or two the factory data would predict. Is there a cooling problem? How about some troubleshooting?

Analysis of the landing gear cost showed that brake and tire usage weresomewhat high compared to expected usage — possibly due to landing technique?

The analysis of brake overhaul costs also showed that the overhaul cost varied by a factor of eight from the lowest to thehighest cost of an overhaul. Which leads to the obvious question — what do we have to do to get low-cost overhauls every time? A good illustration of how this data can save money is the overhaul cost of the main transmission of a single-engine turbine helicopter. In an analysis of work order data for 50,000 flight hours, the average cost of overhaul was $15,000. However, the actual cost was either $5,000 to $10,000, or it was about $35,000 to $40,000! Further detailed analysis showed that the variation was always associated with a particular part. If it did not meet tolerances, it needed to be replaced at a cost of over $30,000 since there was no repair procedure involved.

Development of a repair procedure lowered the average repair costs and eliminated almost all $35,000 to $40,000 overhauls.

Benchmarking is a powerful way to measure the performance of your maintenance department from year to year, and is useful in a comparison against others. For example, one analysis we did of a small fleet of twin turboprop aircraft, showed that over a period of 5 years, this operator had averaged 1.01 labor hours per flight hour. This compared with a benchmark (from a published database) of 1.04 labor hours per flight hour — good confirmation that this company's efforts to control maintenance labor were headed in the right direction.

How do you get the data to accomplish this kind of analysis? Assuming you have the work orders, there are two ways. The first, and by far the best, is to use integrated maintenance management software that records work order data, and provides this type of in-depth analysis capability. The second is to develop your own database, and then enter the work order data into it. In either case, you'll take all that data and turn it into powerful information you can use to effect real savings.