Lean Thinking in a Cloud of Dust

A comment from Goran Gustaffson, a professor at Chalmers University in Gothenburg, Sweden who was scheduled to attend LPPDE this week:

An interesting observation is that the decision to stop air traffic in Europe is very un-lean in that it rests on assumptions rather than on facts. The only reason for grounding all these planes seems to be early computer simulations that showed alarming results of what volcanic ash could to do an airplane.

However, real tests now during the weekend with numerous planes from British Airways, Air France, Dutch KLM and German Lufthansa which have been flown through the ash clouds at different altitudes have not confirmed any of this. On the contrary; they didn't show any problems or damage at all to the planes.

My guess is that the computer simulations were carried out with data for the ash which is relevant for harder and more abrasive particles than are produced by this Icelandic volcano. Safety always comes first of course, and the decision to stop air traffic was probably a wise one in the beginning when we didn't have any data.

However, given the enormous economic losses that we have already suffered and which are easily estimated it is remarkable that the air traffic safety authorities didn't commence test flights and physical examination of the ash right away in order to verify their simulations. It is very worrying in itself, I think, if they trust computer simulations to the extent that they don't feel they need to justify them by physical tests, facts and observations. So a lesson from this chaos is perhaps that it can be VERY expensive not to be lean ...

Well, our history with airliners flying through volcanic ash has been a series of harrowing near-misses: planes that lost thousands of feet in altitude before regaining their engines. I think the authorities based their first reactions on the best combination of practical experience and data that they had. If I had gathered that data in my first LOOK, ASK and MODEL steps, I wouldn't volunteer to be on the first test flight during DISCUSS and ACT!

But LAMDA is a cycle, and waiting for the ash to disperse on its own is probably not a good idea - history also shows that this volcano could erupt for a year or more. I agree with Professor Gustaffson that they could have begun to develop physical models and conduct test flights much sooner. They needed to confirm their assumptions that the ash from this specific volcano was harmful to the engines. From the media reports, it seems like the airlines have undertaken test flights on their own in exasperation - it would be better done as a coordinated effort with standard protocols for the test flights to make it easier to correlate the data across airlines.

I can see a series of LAMDA cycles to gradually increase the risk as the authorities gather more evidence: fly test planes that they then dismantle to check for damage, then commence with air freight traffic if that goes well, and then add passenger traffic if there have been no incidents after a reasonable trial period. As a passenger, I would want to see those kinds of precautions in place before I was willing to risk my loved ones or myself. Safety is a key driver of customer value for the airlines.

Still, that's easy for me to say from my ocean-view suite on Hilton Head Island. If I had been stranded in a faraway place, running out of money and patience, I may be willing to take on more risk to get home.