This paper describes an open and vital vehicle routing research opportunity.  We describe the pervasive vehicle routing problems (VRPs) automakers use for three different links in their supply chains.  Our research methodology was to analyze an automaker’s logistics problems to identify the most relevant VRP characteristics and search the VRP literature for contributions addressing those characteristics. Arguably the most challenging automotive VRP is parts collection for which we detail the key problem characteristics of capacitated trucks, periodic schedules and stochastic volumes.  We provide industrial data on the routes, supplier volumes and locations.  This shows the large scale of automotive VRPs and characterizes the randomness in the daily shipping volumes.  We observe that the daily volumes are stochastic with generally positively skewed distributions and positive autocorrelation.  Since the large-scale automotive VRPs would be computationally intractable for the existing methods on capacitated VRPs with stochastic volumes, research applied to the automotive industry typically assumes deterministic volumes.  Since we observed meaningful randomness in the supplier volumes, this identifies a need for research on periodic capacitated VRPs that are both stochastic and large-scale.  This paper’s data will serve as a test-bed for future research on this important problem facing the automotive industry.