From OBD to connected diagnostics: a game changer at fleet, vehicle and component level
Carlos Guardiola, Christian Vigild, Frederik De Smet, Klaus Schusteritz
Abstract
Early on-board diagnostics (OBD) standards were enforced in 1988 and, by the beginning of the XXI century, all major automotive markets require some sort of OBD. Over the years, the diagnostics software layer has grown in complexity, yet robust fault detection remains a challenging task: insufficient memory and computation power, suboptimal calibration, and the lack of sufficient real-life operation data for model development are some of the limiting factors. The connected vehicle paradigm allows a complete reshaping of the vehicle diagnostics: real-life data feeds provide operation data of the vehicle fleet; artificial intelligence assists data clustering and model development; and over-the-air update tools allow the deployment of new software components and optimized calibration. A smart combination of embedded and cloud components seems to be a major step forward for the next generation of vehicles, allowing the determination of the in-service emissions at vehicle and fleet level. While many challenges are still to be solved, connectivity offers a giant leap in the area of vehicular diagnostics.