| |||||||||||||||||||||
For NASA, which originally intended to fly space shuttles several times a month, the Operations Research and Industrial Engineering professors created analytical models incorporating stochastic random processes to estimate maintenance needs and unscheduled repairs of such expensive avionics components as circuit boards for navigation.
In the semiconductor industry, too, expense is a major concern, and industry leaders turned to Roundy to help decide when to invest in the machine tools to make a new line of processing chip and to optimize production capacity during the transition.
By law, auto makers manufacture replacement parts for seven years, and estimating end-of-life production quantities is a challenge to them because some owners buy generic parts after their warranties run out. Then, too, there are the random engineering and manufacturing failures that throw a monkey wrench into the calculations. “Product recalls are just a nightmare for everyone concerned,” Muckstadt says, “for the manufacturer who has to adjust the production line to make the new part, for the owner who has to contact the dealer, and for the dealer who has to schedule the repairs.” |
Professors 
“Does it makes sense to keep a small inventory of fully assembled modules on hand, to manufacture them as needed, or to stock component parts to be able to assemble them as needed? Those are the kinds of questions we asked,” Jackson says.
Car and truck parts are another story. “With 75 or 80 million newmodel and old-model General Motors vehicles on the road, registered dealers need to have access to hundreds of thousands of service parts, from large side panels to very small fuses,” says Muckstadt. He and Jackson have conducted research for GM for the past 15 years. “You want 16 distribution centers to be able to ship electro-mechanical parts anywhere in the country within a day and virtually any other part within two or three days,” he says.