Descriptive and conceptual structure of naturalistic driving study research: A computational literature review

Naturalistic driving studies (NDS) are an emerging method of collecting driving data from drivers in instrumented vehicles undertaking everyday trips without experimental control. A computational literature review was performed to assess the NDS research domain that aimed to quantitatively describe the extent and structure of existing applications of NDS data. A corpus of 1120 documents was analysed using the methods of scientometrics and text mining to identify prominent contributors and topics. NDS research saw particular prominence in the US and China, however, international collaboration was limited compared to other disciplines. Network mapping of documents and words showed a high degree of overlap in the data sources, types, and analysis methodologies across NDS research. In the context of a safe system approach to road safety, driver-centred behaviours and characteristics such as distraction, risk, and older age were most relevant in terms of number and occurrence, in contrast to relatively underrepresented aspects of road infrastructure and vehicles. • Computational literature review aligned to scoping-style conceptual goals. • Scientometric and network mapping of NDS documents, authors, and nations. • NDS is an interconnected field with overlapping data, analysis, and applications. • Application to drivers and their behaviour more prominent than environment or vehicle. • International collaboration is limited in NDS publications.