Pharma BI » Case Studies http://pharma-bi.com Business Intelligence Blog Mon, 26 Jul 2010 22:40:06 +0000 http://wordpress.org/?v=2.9.2 en hourly 1 Volcanoes, Airplanes and Quantifying Risk http://pharma-bi.com/2010/04/volcanoes-airplanes-and-quantifying-risk/ http://pharma-bi.com/2010/04/volcanoes-airplanes-and-quantifying-risk/#comments Thu, 22 Apr 2010 19:45:18 +0000 Christine Muser http://pharma-bi.com/?p=636 On April 19th, 2010, IATA, the chief trade group representing airlines, issued a press release which states in part: “IATA criticized Europe’s unique methodology of closing airspace based on theoretical modeling of the ash cloud. ‘This means that governments have not taken their responsibility to make clear decisions based on facts …’ said [Giovanni] Bisignani [...]]]> On April 19th, 2010, IATA, the chief trade group representing airlines, issued a press release which states in part: “IATA criticized Europe’s unique methodology of closing airspace based on theoretical modeling of the ash cloud. ‘This means that governments have not taken their responsibility to make clear decisions based on facts …’ said [Giovanni] Bisignani [IATA’s Director General and CEO].”

This statement leaves the unfortunate impression that decisions based on theoretical models are somehow suspect and of little value.  The better, but less soundbite-friendly, question should have been whether the model adequately predicts the risk. To make useful predictions, a model needs to do two things:

  • account for all key factors that influence the outcome
  • quantify how each factor influences the outcome.

To the frustration of passengers, airlines and businesses around the world, the eruption of the Eyjafjallajokull volcano in Iceland presented several challenges for modelers.  Some of the key factors playing a role here include:

  • Ash composition: how damaging is it?
  • Extent of the cloud: how high, how far and where?
  • Aircraft capability: how much ash exposure can airplanes handle?
  • The costs of staying grounded vs. the costs of catastrophic failure

These four bullets of course represent just the tip of the ice berg: each of them summarizes a long list of related factors.  For example, while current data is unavailable, scientists have to rely on past experience when estimating the size and weight of ash particles, their chemical make-up and the density of the ash.  To confound matters, predictions regarding wind speed, direction and turbulence need to be considered as well.

So the real question needs to be: how well does the model represent the current situation? How well do we understand what the actual key factors are?  How well can we measure them?  Are we relying purely on past insight or can we refine our knowledge with data from the current situation?  To what extent can current satellite images, air samples and meteorological measurements improve our ability to predict the risk to life and well-being of people and property?

This interactive map on the BBC we site shows one example of how output from such modeling looks by mapping the extent of the plume over several days along with normal flight routes across the Atlantic.  This post in Business Week sheds some light on our spotty knowledge regarding the real risks of volcanic ash.

The other critical question revolves around acceptable risk.  Risk not only originates from ash clouds. It also comes from “playing it safe.”  Billions of dollars in lost revenues and productivity put the livelihood of hundreds of thousands of people at risk.  Every day we accept the risks of driving our car – so at what point do ash clouds represent a higher risk than driving a car?  Even if we could come up with an exact numeric value for these risks, how does the value of human life fit into these equations?

Therein lies the apparent disconnect between statistical models and real life: intangible values sometimes outweigh what can be measured.  The decisions we make depend on how we actually perceive the risk.  Yet, in order to put our perceptions into perspective, we need to have good numbers to guide us – and getting good numbers requires a good model of reality.

So, before we start talking about law suits, we need to accept that risk is inherent to anything we do.  Blaming people for doing the best they can to balance public safety with economic considerations wastes resources that would be better spent on improving our ability to assess and manage risks.  We send unmanned drones to gather combat intelligence, why not modify them to collect air samples?  Why not fund research to create better models for volcanic plumes? Especially if history should repeat itself and Eyjafjallajokull continues to sputter for the next year or two.

Additional Reading

From Eruptions, a blog dedicated to volcanism:

Airlines lobby to reopen European airspace closed by Eyjafjallajökull
Posted on: April 18, 2010 2:30 PM, by Erik Klemetti

http://scienceblogs.com/eruptions/2010/04/airlines_lobby_to_reopen_europ.php

Eyjafjallajökull flight cancellations: How the right decision is being made to look wrong

Posted on: April 22, 2010 9:40 AM, by Erik Klemetti

http://scienceblogs.com/eruptions/2010/04/eyjafjallajokull_flight_cancel.php

Research Gap Left Airlines Exposed to Volcano’s Blast (Update1)

April 22, 2010, 9:05 AM EDT

http://www.businessweek.com/news/2010-04-22/research-gap-left-airlines-vulnerable-to-volcano-s-fire-and-ice.html

Is driving more dangerous than flying through ash?

Page last updated at 10:51 GMT, Wednesday, 21 April 2010 11:51 UK

http://news.bbc.co.uk/2/hi/uk_news/magazine/8633484.stm

Could aircraft dodge the volcanic ash cloud?

Page last updated at 14:06 GMT, Tuesday, 20 April 2010 15:06 UK

By Stephen Mulvey

BBC News

http://news.bbc.co.uk/2/hi/in_depth/8632583.stm

Recriminations erupt in ash-fueled aviation crisis

AP 4/21/2010

By Arthur Max, Associated Press Writer

http://news.yahoo.com/s/ap/eu_iceland_volcano

How volcanic ash could ground your next flight

By Larry Dignan | Apr 15, 2010

http://www.smartplanet.com/business/blog/smart-takes/how-volcanic-ash-could-ground-your-next-flight/5989/?tag=content;col1

How Volcanic Ash Can Kill An Airplane

Apr 15, 2010 09:00 AM

Ray Wert

Story & pictures of KLM Flight 867

http://jalopnik.com/5517775/how-volcanic-ash-can-kill-an-airplane

Iceland Volcano Vs Mt. St. Helens And Airspace

Pilot’s discussion forum

http://www.airliners.net/aviation-forums/general_aviation/read.main/4783270/

Volcanic Ash Contingency Plan

ICAO (International Civil Aviation Organization)

http://www.paris.icao.int/documents_open/files.php?subcategory_id=63

http://www.paris.icao.int/documents_open/show_file.php?id=274

MD F.E. Bustillo Iii MD F.E. Bustillo is a 3rd degree contact

Owner, Kansas City Safety Alliance and Research

Location
Kansas City, Missouri Area
Current
  • Owner at Kansas City Safety Alliance and Research
Connections
9 connections

Public Profile
http://www.linkedin.com/pub/md-f-e-bustillo-iii/20/824/b55

Experience

Owner

Kansas City Safety Alliance and Research

Currently holds this position

]]> http://pharma-bi.com/2010/04/volcanoes-airplanes-and-quantifying-risk/feed/ 2 Why We Need Good Data http://pharma-bi.com/2009/12/why-we-need-good-data/ http://pharma-bi.com/2009/12/why-we-need-good-data/#comments Thu, 10 Dec 2009 17:54:59 +0000 Christine Muser http://pharma-bi.com/?p=497 Recently, while working on input for a decision tree, I ran into a scenario that reminded me of the fact that we cannot improve a decision simply by applying a tool or technique. We also need good data.

Here is a hypothetical example: Let us assume we are a contractor who is evaluating a fixed bid [...]]]> Recently, while working on input for a decision tree, I ran into a scenario that reminded me of the fact that we cannot improve a decision simply by applying a tool or technique. We also need good data.

Here is a hypothetical example: Let us assume we are a contractor who is evaluating a fixed bid contract.  This contract will pay $115,000 if we accept a clause for liquidated damages of $50,000 in the event we do not meet some project conditions.  We can remove this clause from the contract, but in that case it only pays $100,000.

From past experience we know that our project costs will fall somewhere between $80,000 and $90,000 and that the likelihood of coming in at the lower cost estimate is around 20%.  This leaves an 80% chance that our costs will come in around $90,000.  Looking at our current capabilities we estimate that we have a 90% chance of being able to meet all conditions and thus avoid having to pay damages.

Putting all of this into the decision tree pictured below, we conclude that accepting the liquidated damages clause is the better business decision.

Decision Tree: 90% Probability of Avoiding Damages

Decision Tree showing the EMV of two contract options

But how good is our estimate for avoiding damages?  Can we really trust it?  What data do we have to back it up?  Have we really considered all the factors that can influence our estimate?  After all, as the image below shows, if we are off by only 20 percentage points, the decision becomes a toss up.

Decision Tree: 70% Probability of Avoiding Damages

A decision tree showing what happens when we lower the assumption for avoiding damages from 90% to 70%

In a decision tree each chance node acts as a weighting factor, so it is worthwhile to pay special attention to events that are estimated to have a very high or very low chance of occurring.  We want to be sure that we have good data to back up these optimistic (or pessimistic) numbers.

Of course it is not always feasible to gather all the data we need.  Sometimes the data is too expensive given what is at stake, sometimes it is unavailable and sometimes the quality of the data is too unreliable for a given purpose.  In that case, experience and judgment need to fill in the data holes.  We also call this “making assumptions.”

When making assumptions, we should clearly identify them and decide what to do when one or more of them has to change.  We need to

  • identify which factors influence our assumptions
  • determine how these factors influence the result
  • be able to recognize when a significant change in our assumptions is needed
  • have a process in place to handle these changes when they do occur.

No one can predict the future with certainty.  But the more we understand the probabilities, the better prepared we are.

]]> http://pharma-bi.com/2009/12/why-we-need-good-data/feed/ 0