Data collection and processing tools for naturalistic study of powered two-wheelers users' behaviours.

Instrumented vehicles are key tools for in-depth understanding of drivers' behaviours, thus for the design of scientifically based countermeasures to reduce fatalities and injuries. The instrumentation of Powered Two-Wheelers (PTW) has been less widely implemented that for vehicles, in part due to the technical challenges involved. The last decade has seen the development in Europe of several tools and methodologies to study motorcycle riders' behaviours and motorcycle dynamics for a range of situations, including crash events involving falls. Thanks to these tools, a broad-ranging research programme has been conducted, from the design and tuning of real-time falls detection to the study of riding training systems, as well as studies focusing on naturalistic riding situations such as filtering and line splitting. The methodology designed for the in-depth study of riders' behaviours in naturalistic situations can be based upon the combination of several sources of data such as: PTW sensors, context-based video retrieval system, Global Positioning System (GPS) and verbal data on the riders' decisions making process. The goals of this paper are: (1) to present the methodological tools developed and used by INRETS-MSIS (now Ifsttar-TS2/Simu) in the last decade for the study of riders' behaviours in real-world environment as well as on track for situations up to falls, (2) to illustrate the kind of results that can be gained from the conducted studies, (3) to identify the advantages and limitations of the proposed methodology to conduct large scale naturalistic riding studies, and (4) to highlight how the knowledge gained from this approach will fill many of the knowledge gaps about PTW-riders' behaviours and risk factors.