< Back to
< Back to all News
Student, Professor Collaborate on Mathematical Modeling Solution for Australian Company
October 27, 2014
Creative Problem Solvers: Brent Austgen (left), a 2014 Rose-Hulman graduate, worked nine months with Mathematics Professor Allen Holder, PhD, to develop a cost-saving production solution for an Australian company. (Photo by Dale Long)
A Rose-Hulman student and professor used their operations research and computational science skills to help a leading global manufacturer determine if optimization technology could improve operations at a foundry in Australia.
The nine-month process produced optimized schedules that indicate a possible one-third reduction in manufacturing time, substantial cost savings, and gains over other quality manufacturing paradigms, like Lean and Six Sigma.
This is another example of how Rose-Hulman students are well prepared to tackle real-world solutions, and ready to make a difference when entering their science, engineering, and mathematics careers.
Pentair Water Solutions, a leading global manufacturer and supplier of water and environmental solutions, manufactures a range of irrigation-related products on a metal casting production line. This process involves pouring molten metal into a series of molds across two pouring lines—involving as many as 1,300 molds being produced at different metal temperatures and cooling schedules from the same original ladle that pours the molten metal.
Brent Austgen, a 2014 Rose-Hulman graduate with a triple major in mathematics, scientific computing, and electrical engineering, was asked to improve the process—with never personally observing the plant’s operations. Dana Hofheins, Pentair’s operations director, provided eight large spreadsheets, a few diagrams, and an hour long video of the foundry's operations.
Mathematics Professor Allen Holder, PhD, assisted Austgen on the project, which started in late September, 2013.
The first task for the duo was to model the problem in a way that greatly reduced the number of variables and only considered those possibilities that satisfied all the manufacturing restrictions. That was addressed by March, 2014, and Austgen asked for more data to validate the software.
“Mathematics is often a difficult language to understand and use. However, if we come equipped with ingenuity and intent, we gain what is arguably the most verbose language, and with it we can describe the world,” says Austgen. “At first glance, the model we presented may seem excessively complicated for such a simple process. The beauty though is that if you take the time to examine the details, you can see that we describe the process exactly—nothing more and nothing less.”
However, a couple of additional complicating factors, found late in the process, nearly sidetracked the project.
"We had made an incorrect assumption in one of the up-stream processes, and it was exactly this assumption that Brent and I leveraged to solve the problem,” states Holder. “For a while it appeared that we would be unable to adapt our prior success to our new realization of the problem."
Determined to solve the problem, Augsten and Holder started meeting several hours each day throughout the spring. A new model, designed by mid-May, brought forth a possible one-third reduction in manufacturing time for much of the week, bringing substantial savings for a process that costs $1,400 per hour.
Final approval on the project came before Augsten received his degrees and started work with Intel in Urbana-Champaign, Illinois. The project has been featured in international manufacturing publications.
"The best outcome when you look at the results of something of this magnitude, is when it appears to be exactly what we needed,” says Hofheins, a 1983 electrical engineering alumnus.
Hofheins adds that projects like this are important to the manufacturing sector because they provide gains over other quality paradigms, like Lean and Six Sigma.
"We must be clever and use the tools we have to our advantage to retain any manufacturing in Australia or the United States,” he states. “This problem is way beyond Toyota Lean and GE Six Sigma, and it really opens up the next level of manufacturing improvement which Lean and Six Sigma cannot."
The project showcases the close academic relationships that develop between Rose-Hulman professors and students throughout their undergraduate careers. It also highlights how Augsten was well prepared for the project through his coursework in operations research and computational science.
“Working with Brent was an educational highlight,” says Holder. “So many of Rose-Hulman's students are talented and motivated, but Brent's success with this project has been special. He overcame much and deserves several fold more than the eight hours of academic credit that he has received. He became a close colleague and I look forward to following his career success.”
Augsten adds, “I attribute a large part of the joy this project has brought me to Dr. Holder’s professionalism and friendliness. He has a knack for teaching. Professors like him are the reasons why Rose-Hulman is such a special place for the students.”