In the last post, I started to resolve the trends around this field, but to find the most relevant and current papers, I tried to explore information in the last 5 years and focus on the literature reviews that bring me the comparison, analysis and study results to find the patterns between them.

Understanding and modelling driving behaviour is a difficult study topic today. It is generally acknowledged that different drivers’ driving behaviours vary depending on a range of factors, such as driving history, emotions, population, and environmental conditions. In addition, even the same driver’s driving style can change from trip to trip or context to context. Understanding driving behaviour is crucial, particularly when attempting to identify the circumstances in which a driver demonstrates risky driving behaviour. This research is based on three main questions as suggested by the subject facilitator to reduce and focus the research.

1. What are the trends in smartphone use to capture driving behaviour using machine learning techniques?
2. What metrics tactics are used to define the driving behaviours predicted by the data collected from smartphone sensors?
3. What machine learning algorithms and metrics of performance are used for driving behaviour analysis

Therefore, the annotated bibliography was collected based to respond to those questions, then, the objective of Q1 is to find trends of use of smartphone sensor data to analyse driving behaviour, it was resolved using a methodological search in Google Scholar using keywords related to this specific area, different queries and filtering papers over the last 10 years were count to find the patterns correlated to the queries. The goal of Q2 and Q3 is to present an overview of the most recent techniques and methods used for driving behaviour analysis, where machine learning and metrics performance are used to identify the type of driving behaviour using as inputs metrics tactics. The selected paper was selected reading titles, abstracts, introduction and jumping to the conclusion to identify the usefulness. To find recent literature over the past 5 years, it was analysed the related work of some documents that guide authors to their research, this guided this research to resent citations associated with the question to resolve.

The papers were gathered to make it simple to find data relevant to the literature and the question to be answered. This helped to structure the research’s final report and findings. On the other hand, it enables the creation of a literature review that connects the literature findings to the subject of the study.