How Am I Driving? Using Genetic Programming to Generate Scoring Functions for Urban Driving Behavior

Roberto López,Perla Juárez-Smith,Olanda Prieto,Luis González Gurrola,Leonardo Trujillo,Antonio Posada,Graciela Ramírez,Leticia Méndez

doi:10.3390/mca23020019

Abstract

Road traffic injuries are a serious concern in emerging economies. Their death toll and economic impact are shocking, with 9 out of 10 deaths occurring in low or middle-income countries; and road traffic crashes representing 3% of their gross domestic product. One way to mitigate these issues is to develop technology to effectively assist the driver, perhaps making him more aware about how her (his) decisions influence safety. Following this idea, in this paper we evaluate computational models that can score the behavior of a driver based on a risky-safety scale. Potential applications of these models include car rental agencies, insurance companies or transportation service providers. In a previous work, we showed that Genetic Programming (GP) was a successful methodology to evolve mathematical functions with the ability to learn how people subjectively score a road trip. The input to this model was a vector of frequencies of risky maneuvers, which were supposed to be detected in a sensor layer. Moreover, GP was shown, even with statistical significance, to be better than six other Machine Learning strategies, including Neural Networks, Support Vector Regression and a Fuzzy Inference system, among others. A pending task, since then, was to evaluate if a more detailed comparison of different strategies based on GP could improve upon the best GP model. In this work, we evaluate, side by side, scoring functions evolved by three different variants of GP. In the end, the results suggest that two of these strategies are very competitive in terms of accuracy and simplicity, both generating models that could be implemented in current technology that seeks to assist the driver in real-world scenarios.

Highlights

It is a well known problem that reckless driving affects society in a variety of ways, with noteworthy impacts on health, economic and social issues
We present the results of the execution of each Genetic Programming (GP) strategy applied to the problem of finding suitable scoring functions for driving trips
The format for all comparison tables present the performance on each fold, showing the best training RMSE, the RMSE of the best solution found, the size of the best individual given in number of nodes, and the average size of the population

Summary

Introduction

It is a well known problem that reckless driving affects society in a variety of ways, with noteworthy impacts on health, economic and social issues. Insurance companies could work on coverage plans designed according to a driver’s profile which could, at least partially, be based on automatic tools that exploit useful information obtained from regular driving trips [1,2]. For these reasons, there has been a growing amount of interest in the development of approaches that can extract information from sensors embedded in almost every modern mobile device, such as GPS, gyroscope, magnetometer and accelerometer. Other authors have suggested to expand the amount of sensors used to score driving behaviors, such as using real time data about weather, road conditions and traffic density, among others [6]

Objectives

Methods

Results

Conclusion