Modelling and optimizing automotive assembly plant flash-programming time

Abstract

As automotive companies standardize their electronic control units (ECUs) on a common electrical architecture across all vehicle platforms, more and more software-enabled features will be delivered using assembly plant flash programming. Assembly plant flash programming is the process by which software or calibration data sets are downloaded to vehicle electronic control units using an off-board programming tool during the final vehicle assembly process, thus postponing vehicle product differentiation and customization, reducing the rate of growth in the number of unique ECU part numbers, and reducing manufacturing complexity. As vehicle electronic control modules and their software content or calibration data sets increase, the demand on manufacturing time and space increases significantly. This paper identifies and examines all factors that impact the assembly plant flash-programming time, develops simple and accurate analytical models, and describes an innovative optimization-based approach for determining the parallel flash-programming time for the in-vehicle communication buses that are capable of interleaving packets. The optimization approach can be used to identify the optimal scheme for the parallel flash programming of all vehicle ECUs. By following the suggested scheme, it can dramatically increase the amount of calibration data that can be flash programmed within the same process time constraint, thus increasing manufacturing flexibility and more fully utilizing limited manufacturing production capacity.