Premium On Safety - Issue 16, 2015 - 6

Wrapping...

(continued from page 5)

Done thoughtfully, data analysis and monitoring can offer diagnosis

style. You might go back and
forth, especially if it takes
several layers of analysis
and treatment for operational shortcomings that could prevent your
to reach the conclusion.
Remember the characterflight department from attaining the margin of safety you expect.
istics that make a really
persuasive case study: an accident chain of
you've identified involved one or both, that's
tial principles: Multiple searches using difsome complexity, especially urgent or unusual
probably an area that warrants closer scrutiferent criteria may be required to identify all
circumstances, and of course clear evidence
ny. How many would your current procedures
relevant cases, and unless those criteria are
of the error you're trying to prevent ("Case by
have failed to prevent?
very simple (e.g., make, model, and time periCase," Premium on Safety, Issue 14). A series
Alas, most of the time the results aren't
od), it's necessary to review every candidate
of quick thumbnail descriptions may be more
that clear-cut. But the goal isn't to attain the
to make sure it actually fits the bill ("Where
effective than one extended narrative.
standard of proof needed for publication in
the Wreckage Is Buried," Premium on Safety,
Take equal care in presenting numerical
a scientific journal; a plausible suggestion is
Issue 13). It helps to take notes during that
evidence ("Does It All Add Up?" Premium on
enough to merit some attention. Additional
review, checking off whatever characteristics
Safety, Issue 15). Present clean compariseem relevant: density altitude, contaminated analysis might be helpful. How many of
sons ("apples to apples") when possible, and
runway, IMC at night, etc. There's no shame in those slippery runways were shorter than
point out possible confounding factors that
4,000 feet? How many of those flights were
being old-fashioned and entering these varican't be controlled. Be careful with percentsingle-pilot?
ables into a spreadsheet (which can facilitate
ages-make sure they're computed from
Armed with those numbers, you can begin
computation)-or even in making tick marks
comparable groups. Use bar charts, not pie
building your case. Make it as compelling as
on a legal pad if that's more comfortable. It
charts, but don't get cute with the scales: 94
possible. Especially when you're changing
also makes sense to flag potential case studlooks a whole lot bigger than 91 if the vertiies as soon as you notice them-this can save established procedure, the most consistent
cal axis only extends from 90 to 95. Choose
adherence to the new ways will come from
lots of time later!
scales appropriate for the subject.
getting everyone involved to understand and
The resulting collection of individual
Figuring out what's happening is only half
believe in the reasons for the change. If the
accidents provides a stock of raw material:
the problem, often the easier half. Fixing it
boss needs to be convinced that increased
How to use it depends on the nature of the
often requires convincing other people that
safety justifies additional expense, the arguproblem. Those who've lived truly virtuous
you're right. Seize every advantage!
ment can't be too strong or the evidence
lives may find that taken together, the data
documented too thoroughly.
show a constellation of risk factors that can
David Jack Kenny is the Air Safety Institute's
Whether it's most effective to begin with
be compared directly to known measures
statistician. He is a fixed-wing ATP with pristatistics and reinforce them with examples
of exposure. If less than 20 percent of your
vate pilot privileges for helicopters and the
landings are made at night or on contaminat- or use case studies to motivate a more sysowner of a Piper Arrow.
tematic presentation is a matter of personal
ed runways but 45 percent of the overruns

SmS CoRNER

Six Ways to Lose money
There are six hotspots in aviation maintenance and ground handling that cost the
aviation industry tens of billions of dollars
per year. After exhaustive research into the
underlying causes of these errors, a new
program has emerged that directly targets
these high-risk, high-cost scenarios.
Convergent Performance's Aviation

By JAy DANKoff

Maintenance Never Events™ program is
laser-focused on six costly and avoidable
events; Wrong Part, Wrong Person, Wrong
Fluid, Improper Movement, Lost Tool, and
Unsecured Panels. The global costs associated with these are staggering.
According to multiple analyses, the
annual cost of FOD and ground move-

ment damage alone is as much as $15
to $16 billion (The Operators' Manual for
Human Factors in Aviation Maintenance,
Federal Aviation Administration, 2003;
The Economic Cost of FOD to Airlines,
McCreary, 2008 (Insight SRI White Paper).
Add in wrong part/installation errors,
the cost of underqualified maintenance,
(continued on page 7)

https://www.nxtbook.com/nxtbooks/aopa/premiumonsafety_2014issue14/#/7 https://www.nxtbook.com/nxtbooks/aopa/premiumonsafety_2014issue14/#/7 https://www.nxtbook.com/nxtbooks/aopa/premiumonsafety_2014issue13/#/6 https://www.nxtbook.com/nxtbooks/aopa/premiumonsafety_2014issue13/#/6 https://www.nxtbook.com/nxtbooks/aopa/premiumonsafety_2014issue13/#/6 https://www.nxtbook.com/nxtbooks/aopa/premiumonsafety_2014issue15/#/6 https://www.nxtbook.com/nxtbooks/aopa/premiumonsafety_2014issue15/#/6

Premium On Safety - Issue 16, 2015

Table of Contents for the Digital Edition of Premium On Safety - Issue 16, 2015