Recent changes in petroleum supplies have sharpened interest in developing fuel-saving lubricants, alternate fuels, and fuel-efficient vehicles. These problems typically require engineers to identify and measure small changes in fuel economy and vehicle performance in the presence of considerable noise.
This poses new challenges for both engineers and statisticians. Now more than ever, they must work together in all aspects of these problems: from developing quantitative models and designing experiments to analyzing and interpreting test results.
Using examples from actual fuel-saving engine oil tests, this paper discusses statistical aspects of the productive collaboration that should occur between engineers and statisticians.