Tag: Module E

Teaching Tip: Using Knee Surgery to Illustrate Learning Curves

One of my favorite OM topics is learning curves (covered in Module E).  It takes only an hour of class time to cover well and it’s also a subject that is very motivational. When I tell students that understanding learning curves may save their lives one day, it catches their attention.

While  learning curves originated in the aircraft industry– and continue to drive production rates at Boeing today– an equally important application is in surgical procedures. I usually mention my friend the urologist, who has done thousands of kidney transplants. Who would you rather have operating on you?  A rookie on his or her 3rd or 6th transplant–or someone who is well down the experience curve?

In the text, we discuss one -year death rates for heart transplant patients, which follow  a 79% learning rate. Yesterday’s Wall Street Journal (Feb.14, 2012) provides a second medical example, with an article titled “Study Shows Knee Surgeons Have a Learning Curve”.  Here we find that if a patient’s ACL (anterior cruciate ligament) surgery was among the first 10 such cases of a surgeon’s career, the patient had 5 times the risk of having another ACL repair within a year as a patient whose doctor had performed more than 150 of the operations.

While it isn’t surprising that there is a learning curve, “it was striking to see the figures shown in such a dramatic way,” says the orthopedic professor who did the study.  ACL surgery, while routine, he adds, “is fairly complex.”  Potential pitfalls include incorrectly placing the graft, not fixing it solidly, and not dealing with other damaged ligaments in the area. Given that the learning curve is inevitable, what can be done? Two solutions are: have an experienced doctor supervise surgeries early in a new surgeon’s career, or use medical simulators in training.

Meanwhile, my 1st question is always: “How many times have you done this procedure?”

Teaching Tip: Learning Curves and the Boeing 787

I first heard of the importance of learning curves when working at McDonnell Douglas right out of college during the peak of the Viet Nam War. While I toiled designing the wing of the DC-10 jumbo jet in the basement of the St. Louis plant, over 30 F-4 Phantom jets were rolling off the assembly line on the ground floor per month. Each one took less time than the one before it–and we could compute to the hour how long completion would take, as the learning curve effect is well-known in the airplane industry (see Table E.1 in Module E of OM, 10th ed.).

But the F-4 is old news and we want to provide you with a more current example. Businessweek (Aug.23-30, 2011) tells the story of Boeing’s Dreamliner 787, the world’s fastest selling jet, racking up more than 800 orders before  it even flew. The planes have an average “catalog price” of $202 million, and Boeing plans to assemble 10 a month by 2013– a record for wide-body jets.

But here comes the bad news. As you know, Boeing is running 3 years behind schedule because of supply chain problems (that we have blogged about several times). The company has amassed $16.2 billion worth of inventory in the form of 35 almost-finished  jets  scattered at parking spaces from Washington to Texas. Some are waiting for seats, some lavatories, and others engines. Boeing has spent an average of over $250 million to build each of the 44 planes it has “completed” so far.  The 45th plane will cost $184 million. To reach a break-even point at 1,000 planes, the unit cost must drop to $113 million. And this can only be accomplished with a very aggressive learning curve of 76%.

Based on the plane (no.45) currently being completed, the rate sits at 82.5%, not far from the 84% learning curve the Boeing 777  jumbo jet  has followed. At this rate, the loss will be $4 billion per year through 2015. Is the learning curve critical to Boeing? Absolutely!