General Motors and Center for Business analytics team up on testing project
Auto giant General Motors turned to BGSU’s Center for Business Analytics for statistical consulting on an engine oil testing project. This project not only assisted GM, but also gave graduate level statistics students a high power dose of real-world learning experiences.
The two-phase statistical consulting project between GM and the Center for Business Analytics began three years ago and recently concluded. Andrew Buczynsky is a technical specialist, engine oil, at GM. He became interested in using statistics with a new oil testing project and met with the director of the Business Analytics Center, Dr. Arthur Yeh.
Buczynsky describes the project which brought the automaker to BGSU. “I was looking for statistical consulting on a project outside the scope of typical engineering work done internally. GM had introduced a new engine oil specification called dexos™, and a new engine test was being developed to measure oil oxidation as part of the specification.”
The GM technical specialist continues, “The statistical work was conducted in two phases to coincide with the test development schedule. First, calculations were done to determine the number of runs required to distinguish between oils of different oxidation propensity with a certain level of confidence. The answer allowed a budget to be set for prove-out testing. Second, once the test procedure was finalized and reference oils were run for prove-out at several laboratories, results were analyzed to: (1) determine whether the test, in fact, could differentiate between oils; (2) calculate the mean and standard deviation for each oil; (3) calculate the repeatability and reproducibility of the test. The mean and standard deviation were then used to establish control limits for each reference oil.”
As a result of his first experience working with the Center, Buczynsky was pleased with the results from Dr. Yeh and his graduate assistants. Deadlines were met and draft reports were well written, according to Buczynsky. “Responses to my requests for more analysis, explanation of the results, and revisions to the report were done promptly. I was very pleased with the work done by BGSU, and would not hesitate to employ their expertise on a future project. I found the collaboration beneficial in better understanding the problem having to explain it and better understanding the results having to review them.”
Dr. Yeh says the graduate students who worked on the project really benefitted from such an experience. “They were given the opportunity to utilize what they learned in class to analyze real data. GM also funded the stipend for the graduate students working on the project.”
Besides assisting GM and benefiting the students’ learning, the GM project also contributed to academic research. During the first phase, a statistical methodology was developed to carry out a critical calculation and the development of the statistical methodology was turned into a published journal article co-authored by Dr. Yeh and the graduate student working on the project.