PI: Will Northrop, Mechanical Engineering
Co-PI: Shashi Shekhar, Computer Science & Engineering
Luying Liu, PhD student, Mechanical Engineering
The premise of this project is that by mining and analyzing in-use vehicle data, snowplow vehicle fuel use can be reduced while still providing a critical service for roadway safety and convenience. Currently, 600 MnDOT snowplows are instrumented with an AmeriTrak mobile computer integrated into the Maintenance Decision and Support System (MDSS). The vehicle computing systems currently provide high fidelity vehicle data to MnDOT that are underutilized due to the immense size of the database. Our proposed research builds from our existing work in vehicle powertrain and routing co-optimization (VPRO) to harness large quantities of vehicle data to improve vehicle fuel consumption and lower operating costs. Through a collaboration between experienced researchers in mechanical engineering and computer science, this project will analyze on-road vehicle data collected from snowplows in the MnDOT fleet along with historical route information using highly efficient algorithms. Geographic locations of high fuel use, or "hotspots" will be detected and correlated with exogenous parameters like road snow cover, elevation change and vehicle salt loading. Since high snow levels and drifts require more vehicle power, one outcome of the project will be to use fuel use hotspot analysis to recommend snow fence location. Another objective of the project will be to determine which routes are highest energy consumers and recommend routing changes to reduce fuel vehicle fuel consumption.