Methodological considerations regarding motorcycle naturalistic riding investigations based on the use of g-g diagrams for rider profile detection

Abstract

In recent decades, methods to realize naturalistic driving studies have been established to investigate the driver behavior under realistic conditions. Compared to the research on passenger cars, the naturalistic behavior of motorcycle riders is still rather unclear from a scientific point of view. However, understanding this behavior can contribute to the development of accident countermeasures. To this end, naturalistic driving methods can be applied to motorcycling. The objective of this study was to present a methodological approach to conduct a naturalistic riding investigation. The methodological considerations were supported by sample results on the rider profile assessment by using g-g diagrams (combination of longitudinal and lateral accelerations). The focus was on observable and measurable vehicle dynamics as a result of the rider input. A three-step approach was adopted, consisting of a pilot study, controlled main study, and naturalistic riding study. The first phase aimed at defining a measurement setup and investigating the general feasibility of using g-g diagrams for rider profile detection. The second phase involved experimental techniques. A heterogenous rider panel was observed in a quasi-controlled setup to better interpret the variations in data patterns resulting from the individual rider behavior opposed to, e.g., road type. The third phase involved the conduct of the naturalistic riding study. Three rider profiles were identified by considering the combinations of lateral and longitudinal accelerations. As a relevant indicator, the qualitatively assessed shape of the g-g diagrams developed rapidly and remained stable over time.