Few students get a chance to present their research project to the senior leadership team of a multi-million-dollar company and work with its business intelligence team to turn that project into one of its frequently used decision support tools.
But that kind of opportunity is available to industrial and systems engineering students in Ananth Krishnamurthy’s group at the University of Wisconsin-Madison, thanks to a multi-year partnership among the global manufacturing company National Oilwell Varco (NOV), UW-Madison, Penn State University, and Texas A&M University.
“This provides industry-driven projects that are outstanding in terms of complexity, accountability and challenge,” says Krishnamurthy, a professor of industrial and systems engineering at UW-Madison who has spearheaded this partnership since 2011. “There is no way that I could replicate this kind of experience in a classroom.”
Other companies and one of the field’s largest professional societies have taken note.
At the annual meeting of the Institute of Industrial and Systems Engineers in May 2017, the academic-industry partnership received the I5 Award for “implementing ways to improve their organizations with ideas learned from the previous year’s IISE conference.” It was the first time for UW-Madison to be a co-recipient of this innovation award.
The partnership grew out of the Center for Quick Response Manufacturing that Krishnamurthy directs at UW-Madison. NOV, which was already a center member, makes highly specialized oil drilling equipment. One of its land or offshore rigs, which contains pretty much every single tool for drilling oil, may cost up to $1 billion.
“This is a very challenging industry for several reasons,” Krishnamurthy explains. “Its business model requires highly customized manufacturing where the knowledge you have gained from one site—say Brazil—doesn’t transfer well to another site, say Norway. The equipment NOV builds is very large, so transportation is another formidable task, and any kind of failure is extremely expensive.”
Repeated demand and large volumes typically reduce lead times and improve the efficiency of manufacturing processes. Lacking these features, NOV’s supply chain is already more complicated than that of many other industries, but layered on top of that are the oil market’s notorious up-and-down swings.
To cope with these swings, the company often needs to decide if, when and with whom to subcontract. This means weighing the logistics of outsourcing a job and working with a new contractor against the cost of purchasing its own machines—not knowing if and when that capital investment may be recovered because of the equipment’s highly intermittent use.
Since 2011, more than a dozen graduate students and several undergraduates have worked on solving these kinds of problems for NOV, while also benefiting the Center for Quick Response Manufacturing more broadly. The “quick” in its name refers to reduced lead times and a faster response to business cycles, all while keeping costs low.
At the start of a new NOV project, students and faculty visit a company facility to understand the nature of the problem; next, they develop a research-based solution for that problem and test it at other facilities; and finally, they propose a tool for improved supply chain management that can be used at multiple locations.
A successful example is an insourcing/outsourcing optimization tool that former PhD student Ashesh Kumar Sinha—now an assistant professor of industrial and manufacturing systems engineering at Kansas State University—developed with former master’s student Thomas Davich, who now works for internet technology giant Google. It is actively being used by several NOV facilities around the world.
“Our first task was to find the best mix of in-house and outsourced work, which is a mixed integer optimization problem,” Krishnamurthy explains. “The second task was to estimate the impact of that mix on in-house equipment use and lead time, which is a queueing problem. Since the result of the first question is the input for the second, and the result of the second then revises the solution of the first, our novel contribution was an iterative mathematical algorithm.”
For Krishnamurthy, the NOV partnership—which will continue through 2020—is a win-win for all sides.
The company has access to academic expertise, obtains a customized suite of tools and methodologies, and may hire students as already well-trained employees; the professors receive a rich set of supply chain problems that invigorates their research programs and provides endless teaching material; and the students get to explore careers of interest, including opportunities for internships and global travel, and experience the satisfaction of knowing that the products they helped develop are being used to solve real-world problems.
“Seeing your students succeed is the best measure of your own success as a faculty member,” Krishnamurthy says. “I believe that custom-engineered manufacturing has a bright future in the United States and that this kind of training will serve our students very well in many different job sectors.”
Author: Silke Schmidt