Can we Estimate Truck Accident Risk from Telemetric Data using Machine Learning?

TL;DR

This study explores the potential of machine learning models to predict truck accident risk from telemetric data but finds that current approaches do not succeed, highlighting challenges in using such data for risk estimation.

Contribution

The paper evaluates two machine learning methods for accident risk prediction from telemetric data and discusses the difficulties leading to negative results.

Findings

Neither approach successfully predicted accident risk.

Challenges in using telemetric data for risk estimation are discussed.

Methodological attempts did not improve prediction accuracy.

Abstract

Road accidents have a high societal cost that could be reduced through improved risk predictions using machine learning. This study investigates whether telemetric data collected on long-distance trucks can be used to predict the risk of accidents associated with a driver. We use a dataset provided by a truck transportation company containing the driving data of 1,141 drivers for 18 months. We evaluate two different machine learning approaches to perform this task. In the first approach, features are extracted from the time series data using the FRESH algorithm and then used to estimate the risk using Random Forests. In the second approach, we use a convolutional neural network to directly estimate the risk from the time-series data. We find that neither approach is able to successfully estimate the risk of accidents on this dataset, in spite of many methodological attempts. We discuss the difficulties of using telemetric data for the estimation of the risk of accidents that could explain this negative result.